Managing Digital Transformation Med Omslag PDF

M ANAGING
DIGITAL
TR ANSFOR M ATION
Per Andersson, Staffan Movin,

Magnus Mähring, Robin Teigland,
and Karl Wennberg (eds.)
Managing Digital Transformation
Per Andersson, Staffan Movin, Magnus Mähring,
Robin Teigland, and Karl Wennberg (eds.)
Karyn McGettigan, Language Editor

Stockholm School of Economics Institute for Research (SIR)
is an independent research foundation established in 2010. The
Institute’s purpose is to conduct high quality academic research in the
economic sciences with an ambition to combine scientific rigor with
empirical relevance. Its board includes professors and other represen-
tatives of the faculty at the Stockholm School of Economics. The Institute
encourages and assists its researchers to communicate their research
findings to both the scientific community and society at large.
Chair: Professor Richard Wahlund

Director: Johan Söderholm
Address:
Box 6501, SE-113 83 Stockholm, Sweden
Visiting address: Sveavägen 65, Stockholm City
Phone: +46(0)8-736 90 00
www.hhs.se/en/Research/Institutes/institute-for-research/
[email protected]
Keywords: digital innovation, organizational transformation,
digitalization trends, customers, business models, platforms,
eco-systems, analytics, information technology, change
management, Internet of Things

ISBN: 978-91-86797-31-7
First edition
© SIR and the authors, 2018
Art Direction and Design: Petra Lundin, Manifesto
Production: Manifesto, www.manifesto.se
Cover photo: Westend61/Getty Images
Distributed by:
Printed by: BrandFactory, Göteborg, 2018
PROJECT SPONSOR
Stiftelsen Marknadstekniskt centrum, MTC

The Foundation MTC promotes value-creating interaction
and learning between business and research in the areas
of market, service development, digitalization and
ecosystem development. The foundation was established
by the Royal Swedish Academy of Engineering Sciences
(IVA) and the foundation of the Swedish Institute of
Management (IFL) in 1974. MTC is a non-profit organiza-
tion, thus the projects are financed primarily by major
corporations and government agencies.
In his central role at the Wallenberg Foundations,
Peter Wallenberg Jr has furthered a broad range of important research
and research-led education initiatives at the Stockholm School of Economics
(SSE) and its Institute for Research (SIR). This indispensable work has also
helped create a fertile ground for research on digital innovation and
transformation: a phenomenon currently experienced, shaped, and
managed in and between organisations and throughout society.
This is the topic of this book, which we dedicate to him.

Contents
Acknowledgements 10
Introduction 12
Digitalization: Different Perspectives

1. Strategic Challenges of Digital Innovation and Transformation 17
Per Andersson and Christopher Rosenqvist
2. Reaping Value From Digitalization in Swedish Manufacturing Firms:

Untapped Opportunities? 41
Magnus Mähring, Karl Wennberg, and Robert Demir
3. Digital Platforms: A Critical Review of the Core Concepts 65

Henrik Glimstedt
The Digital Customer

4. Catering to the Digital Consumer:
From Multichannel to Omnichannel Retailing 97
Sara Rosengren, Fredrik Lange, Mikael Hernant, and Angelica Blom
5. Digital Trace Data: Which Data Should we Collect

and What Should we do Once we Have it? 115
Claire Ingram Bogusz
6. Managing Digital Media Investments 133

Erik Modig and Martin Söndergaard
Re-Organisation in Order to
Bridge the Gap to Digital Customers
7. Digitalization of Professional Services:
The Case of Value Creation in Virtual Law Firms 155
Tale Skjølsvik, Karl Joachim Breunig, and Frida Pemer
8. Robotisation of Accounting in Multi-National Companies:

Early Challenges and Links to Strategy 175
Martin Carlsson-Wall and Torkel Strömsten
9. Uncertainty and Complexity in Predictions From Big Data:
Why Managerial Heuristics Will Survive Datafication 189
Gustav Almqvist
10. Explaining the Behaviour of News Consumption 203

Adam Åbonde
11. Digital Transformation Supporting Public Service Innovation:

Business Model Challenges and Sustainable
Development Opportunities 217
Per Andersson and Lars-Gunnar Mattsson
Business Models and Ecosystems

12. The Role and Potential of IoT in Different Ecosystems 243
Jan Markendahl, Stefan Lundberg, and Staffan Movin
13. Digitalization, Collective Intelligence,

and Entrepreneurship in the Care Sector 265
Erik Lakomaa
14. AgTech and the City: The Case of Vertical Farming

and Shaping a Market for Urban-Produced Food 281
Maria J. Bustamante
Future Outlook
15. Future Outlook on Digitalization 301
Robin Teigland, Claire Ingram Bogusz, and Anna Felländer
About the Authors 333

An Assortment of Our Latest Publications 341
Acknowledgements
Every year since 1992, the Stockholm School of Economics Institute for
Research (SIR) has published an Annual Research Anthology, and this year
SIR is publishing the book in cooperation with MTC (Stiftelsen Marknad-
stekniskt Centrum). The purpose of the SIR Annual Research publication is
to enable managers and practitioners better understand and address strategi-
cally important challenges by showcasing SSE research on a selected topic of
importance for both business and society.
This year’s book, Managing Digital Transformation, features authors from
academic areas across SSE together with representatives outside the institution.
The book’s eighteen chapters show the strength and breadth of SSE’s research
within the area of digitalization and reflect the importance that SSE places
upon closely linking research to practice and on investigating the leadership
challenges and their implications in order to support value creation in society.
Participating in the many ongoing research projects at SSE and the multi-
tude of aspects of digital transformation addressed in the various chapters has
been very rewarding for the editors. We would like to thank all the authors for
their hard work and cooperation throughout the project. In finalising this
book, we have relied upon the expert work of Karyn McGettigan for language
editing, Petra Lundin for layout and graphic design, and Marie Wahlström for
digital access to the book. We are, indeed, most grateful for their excellent and
diligent work.
The Director of SIR, Johan Söderholm, and the Chair of SIR, Professor
Richard Wahlund, have provided important support, for which we are deeply
grateful. We would also like to thank Vinnova for its financial support for the
research project Progressive Digital Development: Pre-Requisites for Success of which
this book is part. Hopefully, the book will become a reference for future
research and funding areas.
Finally, we would like to thank all the companies and organisations for
sharing their challenges and engaging in dialogue and research collabora-
tions with us so that we can produce more solid and relevant research to help
better our society.
10
The authors and the editorial team would like to express their gratitude to
the following for generously contributing to this valuable research:
Ü&VSPQFBO6OJPOÕT)PSJ[POSFTFBSDIBOEJOOPWBUJPOQSPHSBNNF
Grant Number 688670
Ü'PSUF(SBOU/VNCFS
Ü)BLPO4XFOTPO'PVOEBUJPO
Ü**45IF*OUFSOFU'PVOEBUJPOJO4XFEFO
Ü*OGJOB'PVOEBUJPO
Ü*P54WFSJHF*OUFSOFUPG5IJOHT4XFEFO
Ü+BO8BMMBOEFSBOE5PN)FEFMJVT'PVOEBUJPO
Ü.BSJBOOFBOE.BSDVT8BMMFOCFSH'PVOEBUJPO
Ü1FUFS8BMMFOCFSHGPVOEBUJPO
Ü4.)*4WFSJHFTNFUFPSPMPHJTLBPDIIZESPMPHJTLBJOTUJUVU
Ü4WFSJHFT3JLTCBOL
Ü5PSF#SPXBMEI'PVOEBUJPO
Ü5PSTUFO4´EFSCFSH'PVOEBUJPO
Ü7JOOPWB
And to the Swedish retailers, executives and other research participants, and
all others who kindly helped make this book a reality.
Stockholm, January 2018

Per Andersson, Staffan Movin, Magnus Mähring, Robin Teigland, Karl Wennberg
11
Introduction
One of the hottest research topics lately is digitalization. Many research proj-
ects are focusing upon different perspectives. Gone are the days when digital-
ization or business implications of ICT were just about increasing efficiency.
Instead, the ripple effect of digital development can now be felt wider and
deeper than ever before. The way in which business is conducted and how it
creates value, as well as how corporations can become more efficient and
sustainable, are all implications of digitalization. Adapting to new demands
and taking advantage of the plethora of possibilities, however, is not always
easy.
Managing digitalization and the transformation of business always involves
new challenges. The novelty and complexity of the digital age has led to an
increased academic interest in the area of digital transformation and a call
from companies that seek support in this process.
We take a look at digitalization from the perspective of business research.
This creates a better understanding of the challenges that today’s businesses
are facing. We believe this anthology will serve as a tool to help businesses
better understand the force that is digitalization and support these corpora-
tions in their digital transformation.
The idea behind this anthology grew as Marknadstekniskt Centrum was
taking part in several interesting research projects. Companies were asking
MTC to facilitate contact with scholars and supply them with academic
insight. Vinnova came on board, by supporting the project Progressiv digital
utveckling förutsättningar för framgång (Progressive Digital Development: Pre-Requi-
sites for Success) of which this book is a part: its aim to stimulate business to
become more progressive in digital change. At last, this book and the website
www.digitalchange.com have become a reality.
This joint venture between Marknadstekniskt Centrum and The Stock-
holm School of Economics Institute for Research follows the SIR tradition of
publishing an annual yearbook to showcase its vital research contributions.
The book begins with an overview of digitalization, then moves to under-
standing the new digital customer, and ends by exploring re-organisational
effects, business models, and ecosystems. We hope this year’s anthology will
be useful for managers by facilitating their digitalization processes.
12
PART 1: DIGITALIZ ATION – DIFFERENT PERSPECTIVES
The role of digital technology in business and society is rapidly shifting from
being a driver of marginal efficiency to an enabler of fundamental innovation
and disruption in many industrial sectors, such as media, information and
communication industries, and many more. The economic, societal, and
business implications of digitalization are contested and raise serious ques-
tions about the wider impact of digital transformation. Digitalization affects
all private and public operations, as well as the internal and external work-
ings of any operation. Digitalization is the major driving force behind sweep-
ing large-scale transformations in a multitude of industries. Part 1 includes
various perspectives on digitalization and digital transformation.
PART 2: THE NE W DIGITAL CUSTOMER

Digitalization has resulted in more user-centric business and user-centric sys-
tems. The changing behaviour of the digital consumer/customer is discussed
here as it connects to new forms of customer involvement and engagement, as
well as analysis models of what creates customer value in this digital context.
PART 3: THE RE-ORGANISATION IN ORDER

TO CONNECT WITH THE DIGITAL CUSTOMER
How can companies connect with digitalized consumers and non-digitalized
customers? This is a central issue in managing digital transformation, as it
draws attention to the emerging intra-organisational, marketing, and cus-
tomer interaction challenges associated with digitalization: for both the con-
sumer and the supplier. Another aspect of this is the internal handling of new
forms of organizational ambidexterity; that is to say, companies and organi-
zations engaged in digitalization processes often require an internal re-organ-
isation in order to handle the demands that digitalization brings, and to
explore new digital opportunities while promoting their existing business and
operations.
PART 4: BUSINESS MODELS AND ECOSYSTEMS

How do companies change, adapt, and innovate their business models? Given
that digitalization leads to a convergence of previously unconnected or loosely
connected markets, the digitalizing company and organisation is analysed in
its systemic and dynamic context. This part draws attention to business models
13
and business model innovation. Incumbent firms need to adapt and change
business models while competing with digital start-ups based upon new scalable
business models, accessible ventures, and rapid processes of intermediating.
These chapters discuss completely new co-operative business models: processes
that need to be developed as companies shift from products to digitally based
services.
The Ecosystem places digitalizing organisations and companies into their
broader and systemic context. This includes discussions on digital disruption,
industrial convergence processes, and shifting patterns of competition and
cooperation. Digital technologies cause markets to converge in many new
and sometimes unexpected ways. The result is the emergence of new roles
and market positions of technical platforms.
Staffan Movin, Stiftelsen Marknadstekniskt Centrum
14
Digitalization:
Different Perspectives
CHAP TER 1
Strategic Challenges of Digital

Innovation and Transformation
PER ANDERSSON AND CHRISTOPHER ROSENQVIST
Introduction
Digitalization has reached all industries and all sectors of society. Companies
and industries are currently facing challenging transition processes; the
future appears to be less predictable for many, which threatens existing com-
petitive position. Meanwhile, digitalization opens up for many new options,
thus, shifting companies’ and organisations’ opportunities to re-position their
business and operations. Digitalizing incumbents and new digital start-ups
both face a number of strategic challenges associated with digital transforma-
tion and digital innovation processes. Our focus in this chapter is on these
strategic challenges. We will extract, present, and discuss a set of common
strategic challenges that are associated with digital transformation and inno-
vation processes, while drawing upon insights from cases in fifteen different
sectors and digitalizing arenas. Larger incumbents are under pressure to
transform their business while acting in a short-sighted quarter-to-quarter
perspective. The rapid pace with which new start-ups are creating strong
market positions sometimes leaves the incumbents with no choice but to col-
laborate – or be left behind. One major strategic challenge stemming from
digitalization is the development of business models based upon new forms of
cooperation and partnerships. One report expresses this new competition and
the resulting strategic challenges:
“The current generation of entrepreneurs is unlike any other. Empowered by

digital technologies and unencumbered by legacy structures, they are unleash-
ing fundamentally new business practices at a pace that was almost unthinkable
just a couple of decades ago... inventing new business models and monetization
strategies all along the way.” (Source: Ericsson Digital Disruption Report)
17
PER A N DERSSON A N D CH R ISTOPH ER ROSENQV IST
Digitalization poses strategic challenges for both incumbents and for the
new digital disruptors. Major challenges concern business development,
including business model changes. Building upon a broad set of case studies
in a variety of industries, the purpose of this chapter is to provide an empiri-
cal overview and a discussion of the recurrent strategic challenges that are
associated with digital transformation. To set the scene for this discussion, we
introduce three short introductory illustrations from three completely differ-
ent sectors.
Our first example comes from the automotive industry. The increased
demand on behalf of car users for “car access as a service” requires new
business models and a strategic shift from product to service orientation in the
automotive industry. Meanwhile, the digitalization associated with the
launching of connected vehicle strategies enacts another strategic challenge,
which appears to be part of the digitalization process: the creation of new
cooperative business models across previously weakly connected business
networks. When launching various types of “connected vehicle” concepts,
companies in the automotive and the ICT industries report similar experi-
ences. When Ericsson presented its connectivity platform for connecting the
car, thus, directing attention to business model challenges, the ITC company
argued that business modelling and new forms of partnerships could be an
area for innovation: “It… enables new profitable innovative business models
for the industry where new actors – government and third party players – are
able to share revenues. Several stakeholders share the growing interest in
being connected on the road: Governments want to enhance road safety and
collect road tolls and congestion charges. Insurance companies want to be
able to offer insurance based on how you drive; media and content companies
want to be present in the vehicle.” (Ericsson: Connected Vehicle Cloud). This
leads to challenges in establishing completely new business models across
industries. One of the anticipated obstacles relates to the vast differences
between the automotive industry and the mobile communication industry, as
expressed by one mobile operator: ”The automotive and mobile industries
have been drawn together by the unstoppable rise of the Connected Car. As
with any partnership, there will inevitably be teething problems, but both
sides are aware of the importance of making the relationship work as the
demand for connectivity in cars grows.” (Telefonica: Connected Car Industry
Report 2013, p.9)
18
ST R AT EGIC CH A LLENGES OF DIGI TA L I N NOVAT ION A N D T R A NSFOR M AT ION
Our second example is from the medical technology industry. A business

unit manager in a global medical technology company was interviewed about
his and the company’s main challenges with regard to directing attention and
resources to digitalization:
”The biggest challenge is to grasp the complexity of the strategic challenges that
we are facing when taking the next steps into digitalization. We had an idea of
what to start with, but soon we experienced that there were many strategic issues
that required our attention. And they seemed to be strongly connected. We
experienced some difficult delimitation problems quite early when starting to
work with this…”
The company thought it could do something about the masses of machine

and patient data that was being generated from the company’s machines
when used globally in intensive care rooms. Thus, big data issues captured
the company’s early attention. A first mover advantage was expected in rela-
tion to competing global medical technology suppliers that had (connected
and complementary) intensive care machines placed in the same rooms in
different parts of the world. However, creating a new business around the
analyses of big data aimed for intensive care professionals and their hospitals
would require a shift in business orientation: from selling products, associated
hardware, and services to also becoming a service and consultancy provider.
Intensive care staff would have limited use of data analyses from only one
supplier; much more value would be created if analysed data from comple-
mentary medical technology suppliers were included in the service package.
This would require some radical re-thinking on behalf of the medical technol-
ogy suppliers of their present, competition-based business models, as well as
technical integration. New cooperative business models would be required
with patients as the focus of value creation. The initial and delimited big data
was difficult to separate from other main strategic issues: moving from prod-
ucts to services, shifting toward more user-centric business operation, and
handling completely new cooperative business models.
Our third example is from the education sector and the emerging educa-
tion technology industry. Sensavis is a Swedish company that offers high
quality 3D visualisation software for the K12 education sector. Sensavis’
vision is to improve learning outcomes by making it easier to understand
complex and abstract phenomena, particularly within STEM: science,
19
technology, engineering, and mathematics. The small company’s mission is

to provide educators all over the world with tools that enable them to reach
and include all students in the learning process. Headquartered in Sweden,
the company operates in 36 countries and serves over 550 schools, thus,
reaching more than 220,000 students. Sensavis and the education technology
industry face a unique challenge: the public education sector is characterised
by fairly high levels of bureaucracy, risk-aversion, and slow technology adop-
tion (2014 OECD Conference: Innovating the Public Sector: From Ideas to Impact).
This often stands in stark contrast to the innovation-driven education tech-
nology industry, which presents multiple barriers to entry for education
technology companies. In addition, investor impatience to show returns puts
more pressure on education technology companies, such as Sensavis.
Although education technology companies have the opportunity to transform
learning, the potential impact of their digitalized services often depends upon
their ability to develop a viable business model. In the case of Sensavis, as the
small education technology start-up successively learned more about the
customers, the users and the other stakeholders in the local networks, it
altered and adapted its business model four times during its first years of
existence: “We had the ability to spark interest and got the traction of people
who were early adopters and visionaries, but we lacked the understanding of
what people really needed...” And considering this was a situation where
digitalization was difficult due to existing structures and institutions, “we are
trying to change 100 years of tradition in schools and that is the biggest chal-
lenge.”
The three examples indicate that digitalization processes often enact radi-
cal new forms of business and industry change. The same complexity
emerges regardless of the digitalizing industry or sector. During these digital-
ization processes, previously unconnected or weakly connected industry
networks become connected (for example, automotive and telecom indus-
tries). As with these first two examples, many incumbents in different indus-
tries have begun their digitalization journeys; one over-arching experience is
that digital transformation processes often embrace big strategic challenges.
And, one strategic challenge does not come alone; it is often connected to
other bigger issues. Hence, the guiding question is the following:
What are the general, strategic, and managerial challenges associated with digitaliza-
tion?
20
Companies report on a set of major strategic challenges when engaging in

digital transformation. This seems to be general in the sense that they appear
in many different business (and public) contexts. When starting the process
of digital transformation, these strategic challenges do not come alone. For
example, business model challenges can often become intertwined with tech-
nical challenges associated with the introduction of a new technical platform.
Companies show different behavioural patterns in terms of how they deal
with this inter-connectedness of digitalization challenges. We will describe
and elaborate upon these general strategic challenges. First, let us say a few
words about how they have been extracted.
About the Underlying Research

The digital transformation challenges are collected from a longitudinal com-
parative research project focusing upon in-depth studies of companies in 15
different contexts engaged in digital transformation processes (see Table 1.1).
The discussion here builds upon a broad set of ongoing case studies, and
secondary sources on digitalization. The strategic managerial challenges
were extracted from ongoing cases of digitalization processes listed in the
table. The cases all share the fact that the effects of digitalization are apparent/
considerable (a transformation) and are of central strategic importance for the
organisations involved; this includes the role of various technical platforms.
In many of the represented sectors, digitalization is associated with an over-
lapping between sectors and networks, which sometimes includes tension
between different “industrial logics”. Thus, the processes in most cases open
up for considerable changes and re-positioning of companies: where business
development and business model innovation become part of the digitalization
changes. The strategic challenges discussed in this chapter provide a snapshot
of ongoing digitalization processes that have been extracted from both sec-
ondary sources (digitalization reports from different industries, government
reports, companies’ annual reports, and business press) and from a set of
ongoing case studies that occurred between 2015–2017 (see Table 1.1.) The
main case studies (from areas 2, 5, 7, 8, 11, and 15 in the table) are based upon
interviews in the case organisations and workshop sessions. These focus upon
internal and external organisational challenges associated with ongoing and
future steps in the case companies’ digitalization processes. In broad terms,
the digitalization challenges were initially discussed along a set of different
21
themes, which included the following issues: technical, customer exchange

and value, external partnership and business ecosystem, economic and finan-
cial, and other business models. With these as guidelines, a set of ten general
strategic issues was finally extracted from the cases and the secondary
sources from the fifteen digitalization areas. (The research is ongoing as of
2017; therefore, no claim is made that the list is exhaustive or that all issues
are equally relevant for companies in all fifteen areas.)
Table 1.1: Cases of Digital Transformation Processes: Studied in 15 Different Areas
Digitalization Area Start of Focal Sector (and Main Ongoing Case Studies and/or
Empirical Business and Societal Empirical Collection During
Studies Issues) 2016–2017
1 “The Smart Home” 2016 Building construction Mainly ongoing data collection
and digital homes from secondary sources
2. “The Connected 2014 Automotive industry Volvo-Ericsson: The connected vehicle case
Vehicle” (reported In: Andersson & Mattsson 2015)
3. “The Big Media 2014 Media and Mainly ongoing data collection from
Event” entertainment industries secondary sources, including case of a
major sporting event
4. “The Mobile 2010 Administration and Mainly ongoing data collection
Enterprise” office operations from secondary sources
5. “The Remotely 2014 Healthcare sectors Getinge: ongoing case study of integrated
Monitored Patient” patient care and big data challenges
6. “The Beyond-the- 2016 Pharmaceuticals Mainly ongoing data collection from
Pill Solution” and healthcare secondary sources, including reports on
pharmaceutical companies buying health
care digital support companies
7. “The Connected 2015 Farming and food industries Vertical Farming: ongoing case studies of
Farm” the digitalizing of supervision in farming,
including vertical farming
8. “The Networked 2016 Public and private sector Sensavis: ongoing case study of a new
University” education start-up company in the education
technology sector
9. “The Smart City” 2015 City planning and Mainly ongoing data collection from
sustainability issues secondary sources, including reports from
Stockholm and Dubai
10. “The Monitored 2015 Sustainability and issues of Mainly ongoing data collection from
Environment” environmental monitoring secondary sources
22
11. “The ‘Mobility 2015 Transportation and logistics Nobina: ongoing case study of public
-as-a- Service’ transportation supplier (“Public
System” transportation as ‘mobility-as-a service”)
12. “The Automated 2016 Manufacturing, industry Mainly ongoing data collection from
Production System” automation and production secondary sources
13. “Digitalizing 2017 Banking, finance Mainly ongoing data collection from
Finance” institutions and new secondary sources
FinTech start-ups
14. “The Networked 2017 Public safety, social welfare Mainly ongoing data collection from
Public Society” organisations, and elderly secondary sources
care
15. “E-com 3.0: 2015 Retail sectors: music, food, Universal Music Sweden: ongoing case
Consumer Centric and fashion study of management of big data in music
Retailing” consumption
Ten Digitalization Challenges

What are the major managerial challenges experienced by managers in dif-
ferent sectors? We have extracted ten general strategic challenges taken from
discussions and meetings with both global suppliers of digital technology
solutions for transforming sectors and with companies engaged in digital
transformation. We can still say that while the composition of problems under
each of the ten strategic issues can differ, it appears as though several issues
are common, based upon a first exploratory step of analysis and the given
empirical limitations. In addition, the centrality and importance of certain
challenges can differ between industries and sectors, partly depending upon
the stage of digitalization in which a company or sector finds itself. For exam-
ple, the issues facing media companies is often at the forefront of digitalization
– such as, in the music industry – yet they do not fully mirror those of compa-
nies in other more traditional types of manufacturing and construction
industries. We will now summarise these companies by giving short descrip-
tions of them. We begin by addressing the challenges associated with the
technologies, explore the challenges of increased user-centrism, and then
move toward business model issues; , lastly, we will end with more systemic
challenges. We focus upon the broader strategic issues, and end the list with
linking these to important internal organisational challenges.
23
1. M ANAGING THE IMPORTANT ROLES AND

M ARKET POSITIONS OF TECHNICAL PL ATFOR MS
Platform management research in general – and more specifically, informa-
tion management – tends to draw attention to the growing importance of
technical platforms, as well as their organisational and technical complexity.
The emergence of industry-wide technical platforms sparks profound changes
– and challenges – in industrial structures. This blurs industry boundaries,
reshapes markets, and impacts firms’ strategies, structures, and management
processes (Yoo et al. 2010). This transformative impact of digitalization is
often connected to the growing importance of technical platforms. As Gawer
& Cusamano (2014) argue, the concept “platform” has become almost ubiqui-
tous; their definition of “external platforms”, however, draws attention to the
external strategic dimensions: “We define external (industry) platforms as
products, services, or technologies that are similar in some ways to the for-
mer, but provide the foundation upon which outside firms (organized as a
“business ecosystem”) can develop their own complementary products, tech-
nologies, or services…” (ibid, p.418)
A technical platform’s eco-system can refer to one or several platform own-
ers and a more or less explicit platform leader. The way in which technical
platforms emerge and how their ecosystems evolve is of great importance in
management. Platforms are “manageable objects” (Gawer & Cusumano 2014),
purposefully managed to bring multiple parties together: primarily users and
‘complementors’. Platform leaders and their competitors, as well as suppliers,
complementors, and users are involved in both competitive and cooperative
interaction: that is to say, in co-opetition. This creates new managerial chal-
lenges. In digitalization processes, a complementor might become a platform
leader that is in competition with the incumbent leader. Moreover, from the
perspective of layered modular technology, there appears to be many loosely
coupled layers of devices, networks, services, and contents when it comes to
technical platforms that are central in many digitalization processes. Yoo et
al (2010) describe the strategic challenges of this: “Because of the dynamic
nature of the layered modular architecture, the same firms can compete on
one layer and peacefully coexist on other layers” (p.729). Many of the identi-
fied management implications come back to the fact that digitalization and
the new technical platforms upset old industry structures and require new
types of “frameworks” for analysing and understanding competition. Digital
24
technologies and platforms require actors to create new “meanings” to prod-

ucts, competition, customers, etc. (Yoo et al 2010, p.729).
Part of the managerial issue concerns the important role of technical plat-
forms as new intermediaries in service innovation. That is, the technical platforms
take on and/or are given a central role between a set of complementary suppliers
(complementors) and a set of customers. The intermediating role seems to be
associated with new complex patterns of cooperation and competition, thus,
creating important strategic challenges for involved companies, including plat-
form leaders. Intermediation in the digitalization processes seems to be associ-
ated with dynamic network processes and changes in network connections.1
2. M ANAGING BIG DATA:

CRE ATING NE W VALUE BASED SERVICES FROM DIGITALIZ ATION
Quotes from three managers interviewed in 2016 in an ongoing research
project2 on digitalization processes in different sectors of the networked soci-
ety, illustrate the problem area. This is the focus of our second strategic
challenge connected to digital transformation:
“We are working hard right now to try to figure out a strategy for big data
– one important question is if we should build our own internal big data
analytical capabilities, like some leading companies that have created specific
units and business focused on this, or if we should outsource parts of big data
operations to partners.” (Interview with product manager at engineering
company engaged in IT and industry automation, Jan 2016.)
“One of the strategic issues that we are struggling with in our ongoing
digitalization concerns our future approach to big data. There is a business
opportunity in this, but one of the tricky issues concerns the fact that our
products are strongly connected to a number of competing companies’ prod-
ucts. Big data analyses for the end user of these systems is of less value if the
analyses do not involve data from all products. This is a new situation of
competition and cooperation emerging….” (Interview with marketing man-
ager at high-tech medical device company, Dec 2015.)
1 In an article published within our ongoing research, Andersson & Mattsson (2015), argue: “We find it
useful to introduce elements from methodology, specifically including material objects as actors and to
acknowledge the performative role of technology for overlapping and intermediating in industrial
networks.” (p.92)
2 A three-year research project on digitalization processes ending in 2017: Renewal of the Service Society
(“Det mogna tjänstesamhällets förnyelse”, Wallanderstiftelserna)
25
“The first issue is of course to understand if and how our big data analyses
can be actually turned into value creating services for our customers and
other partners. The second issue concerns our own role in this and what our
business model should be when creating new business value for ourselves.”
(Interview with marketing manager at a media company, Feb 2016.)
Recurrent questions in the contacts with managers concern the experienced
challenges associated with big data (BD), now and in the future. As the above
quotes indicate, these challenges concern both strategic and practical opera-
tional issues around BD: What role should the company and other organisa-
tions have in BD, including specialised BD analytic companies? What value
can actually be developed from BD, and for whom in these stakeholder net-
works? And what are the various practical challenges when taking a step into
BD operations: translating accessed, structured, and analysed data into value
creating services? A number of emerging question marks concerning data
sharing, privacy, and ethics around BD are also added to this. Practitioners’
concerns regarding BD, thus, circle around three broad issues: 1) strategic,
which includes the (external) distribution of work and control over various BD
related activities; 2) value and business models, regarding the actual output
value from BD analytics and associated business models; 3) operations and
practices, including the processes of translating BD analyses into value creat-
ing services, and how to internally re-organise in order to manage BD-related
operations. Edelmann & Singer (2015) are in line with the last point; they
describe how companies draw upon large amounts of customer data in order
to analyse and build effective “customer journeys” that, in turn, require new
internal organisational structures and innovative types of management.
3. COPING WITH USER-CENTRIC SYSTEMS:

MEETING THE POWER OF CONSUMER NET WORKS
One common observation in digitalization reports is that the 4th Industrial
Revolution leads to increased user-centric orientation in companies (World
Economic Forum). A consequence of this shift, for example, is that marketing
and its bridging role are placed right at the centre of this development. Con-
sumers buy and use products and services in new ways, sometimes sharing
products and services as part of what collectively has been labelled “the
sharing economy” (Frenken & Schor, 2017). Companies’ existing product and
product sales-based business models become challenged when large networks
26
of consumers or organisational customers start sharing products and begin

demanding “access to product use”, rather than ownership, which is often
based upon digital support systems.
This general shift has also been associated with changes in connectivity,
advances in analytics and artificial intelligence, and the growing profusion of
smart devices and sensors. The experience concept, customer value, and
outcomes receive special attention as consumers’ set of daily digital activities
grows: they search and buy, download, stream, access, connect and create, do
peer reviews, and so on. Since the growing digital ecosystems are becoming
increasingly digital and more user-centric, companies strive to meet the num-
ber of new challenges. One such challenge is how companies should connect
in the best way possible to these digitalizing, and sometimes cooperating,
consumers: directly, via platforms, in aggregated marketplaces, via connected
objects, products, and services, in digital information systems, and so on. A
wide range of new digital marketing technologies, systems, and solutions are
implemented to support back-office analyses of customers’ digital journeys
and new digital behaviour, and also to support new digitalized interactions
with customers. Together, these new marketing technologies are beginning to
shape newly emerging marketing platforms, thus, connecting the digitalizing
company with its digitalizing customers. The challenges of this digitalization
for companies’ marketing operations include the speed and pace of the
responses to, and reactions from, customer actions. Interaction processes are
becoming more continuous with ongoing customer contacts; deep and on-
going insight into customer journeys, and marketing planning processes are
being radically affected. For example, long-term product life cycles and mar-
keting planning is changing, as digitalization enables continuous experimen-
tation and product testing. In turn, this also affects and creates challenges for
the strategies as well as the total business of companies.
4. GO-TO-M ARKET AND SCAL ABILIT Y:

GOING FROM DIGITAL PILOTS TO THE SHAPING OF NE W M ARKET
Digitalization processes include the development, production, implementa-
tion, and use of new digital solutions and offerings. A recurrent managerial
issue in many business sectors is how digital solutions in pilot tests can be
translated into scalable business models to target larger markets. Going from
single successful pilot projects to market launch of new targeted digital offer-
27
ings is often experienced as being challenging due to the uncertainties regard-

ing which parts of the associated business model can support scalability. An
example is a global ICT supplier and incumbent in its ICT business, which
considered a new business opportunity. The company tested new digital
infrastructures with sensors and Internet of Things technology, which also
included mobile phone applications aimed for the audience and spectators at
a world sport championship. The managerial issue in focus was how to turn
a successful digitalization pilot into the next commercial step by creating a
viable business model that could support the sales of the system on the global
market for “big events” (sports, music festivals, and so on.) Which parts of the
business model for this particular “digital solution” could be stabilised, if any,
and which parts needed to be kept agile and adaptable? The business model
scalability issue turned out to be essential. In order for the company to find a
way to go to market and continue providing these types of solutions, it needed
to find a scalable way to transform the ideas tested during the pilot project
into new revenue streams as part of the business model. One of the first issues
to solve became which central partners to target in the (global) stakeholder
network as part of the business model, as well as finding out also the value in
which different stakeholders were seeking from the large-scale digital solu-
tion. A central issue in the digitalization project then became determining
how to transform the value created by new digital big event solution into real
cash flows based upon a new revenue model. Similar large-scale digital infra-
structure projects in other industries indicate similar scalability challenges.
5. M ANAGING DIGITALIZ ATION AND SERVICE TR ANSFOR M ATION:

SHIF TING FROM PRODUCT TO SERVICE-BASED BUSINESS MODELS
In many sectors, digitalization has become connected to transitions from prod-
uct to service provision or, in the words of Vargo & Lusch (2004), shifting
toward a service dominant logic. This has also become one of the major man-
agerial challenges in digitalization processes: one that often requires new
organisational principles, structures, and customer interaction processes.
Business models change from transaction to relationship-oriented, which
means that new sets of capabilities will have to be developed: such as organi-
sation structures, metrics, marketing and sales incentives, and more. Three
possible service innovations can be seen in the following examples: Philips
launching its concept of “lighting as a service”; Volvo marketing its “connected
28
vehicle” based upon new digital infrastructures; and, public transportation

company Nobina moving toward “mobility-as-a-service” for city travellers
instead of bus operations. These innovations involve many knowledge areas,
new digital technologies and platforms, and actors from several industries in
new partnerships. Such digital service innovations may be new, in terms of
how individual services are connected to each other (bundled or unbundled),
the role of different actors (including the users), the organisation and the dis-
tribution of services (as well as the price and payment for services). We have
seen and reported (Andersson & Mattsson 2015) how cooperative and compet-
itive relationships in business practice change when different knowledge areas
and industries are involved in such digital service innovations. Shifting from
product to service-based business models when going through a digital trans-
formation means that uncertainty and complexity need to be acknowledged.
Moreover, the value of the new digitally-based service for a user may be more
or less difficult to perceive and evaluate during the transformation.
6. SHAPING NE W CO-OPER ATIVE BUSINESS MODELS

Digitalization opens up for network interdependencies that cross industry
borders. An individual company in such a context cannot independently
develop and implement a sustainable digital transformation, including busi-
ness model change (Berman 2012; Westerman & Bonnet 2015; Ehret & Wirtz
2017) For example, implementing the ideas of “smart cities” based upon new
digital infrastructures, means that different private as well as public actors
may prefer other designs of the model. Conflicts between actors with different
business models need to be addressed. For a digital service innovation of this
magnitude, many uncertainties will emerge, thus, challenging traditional
roles and positions. And, these are challenges that both young digitalizing
firms and established businesses face when they seek to collaborate with one
another in order to engage in digital transformation processes. However, such
cross-industry collaborations also enable firms to accelerate innovation and
create more competitive market positions. Therefore, there are two kinds of
challenges associated with the digitalization and the creation of new cooper-
ative partnerships and new business models: first, as in the case of Volvo and
Ericsson around the Connected Vehicle Cloud, there is the challenge of
creating new types of bilateral cooperation. Secondly, such as in the case of
“smart cities”, there are the many challenges of establishing functioning
29
partnerships associated with broader sets of cooperating organisations. This

type of wide form of digitalization challenging innovation connects to Vargo
et al (2015): for example, they propose an ecosystem approach for considering
different types of technological and market innovations as being driven by
new forms of institutionalisation processes.
7. M ANAGING THE NE W DIGITAL ENTREPRENEURS

AND THEIR R APID PROCESSES OF INTER MEDIATING
There is a tendency to draw a great deal of attention to the steps and moves of
the so-called Unicorns and successful digital start-ups when discussing and
analysing digital transformation. In reality, digital transformation processes in
all sectors involve a wide array of different relationships between both young
and established incumbent firms being dependent upon each other and being
part of the same transformation. Digital transformation is a collaborative inno-
vation process in which small and large companies create joint strategic partner-
ships. The small digital start-ups may take advantage of the fact that many of
them are digital from the start and may also possess a scalable business model
from the outset. They are often in need of access to the established incumbents’
advantages, however: in terms of financial resources, established cooperation
networks, experiences, regulatory knowledge, and so on. In some cases, they
can rapidly scale successful digitalization experiments across multiple markets.
An example can be seen in the service sector. New digital actors, such as
Booking.com and Tripadvisor, have created new intermediate positions in a
short period of time between hotels (including big global hotel chains) and
customers: where the former struggles to attract hotel customers back to the
hotels’ own web/booking sites. As a result of digitalization processes, similar
intermediation can be seen in other service industries, thus, challenging the
incumbents regarding who is going to have most of the direct contact with the
customers. As a previous study stated:“… service innovation processes might
require, or stimulate, changes in intermediation – sometimes also the entry of
new actors as intermediaries” (Andersson & Mattsson 2015). One company
expressed the process as going in both directions: that is to say, embracing
both dis-intermediation and re-intermediation processes: “Increased usage of
ICT has led to a complex and dynamic process of ‘disintermediation,’ as
producers are able to generate direct sales and creators can directly distribute
their work online. Telecom and IT players, meanwhile, create a move toward
30
‘re-intermediation,’ allowing smaller companies that may not have large mar-
keting budgets to participate in the market.”3 The challenge of managing the
new digital entrepreneurs and their rapid processes of intermediating has a
mirror effect; the digital start-ups need to embed their business in established
network settings by relating to incumbents. No business is an island.
8. CRE ATING NE W BUSINESSES ACROSS

INDUSTRY BOUNDARIES AND INDUSTRY LOGICS
Digital technologies cause markets to converge in many new and sometimes
unexpected ways. The previous wave of convergence between the sectors for
information, communication, media, and household appliances is being
replaced by much more radical forms of convergence (Hackling, 2013). And,
digital technologies are at the centre of these processes: pulling different and
sometimes distant industries and markets toward each another. Previously
unrelated sectors become dependent upon one another. Connectivity and
interdependence between networks change with digitalization, due to techni-
cal and market convergence: for example, between the ICT sectors, the auto-
motive industry, and in various public spheres (road authorities, and so on)
when launching more integrated connected vehicle concepts. Following these
processes of overlapping between industrial networks, unexpected new pat-
terns of cooperating and competing companies emerge and create new market
situations. Such overlapping between digitalizing industrial networks implies
the need to consider and address new relationships, changes in old ones, and
closer indirect contact. Overlapping confronts actors with uncertainties and
tensions, and sometimes threatens established network positions (Andersson
& Mattsson 2015). Through digitalization, companies that were not previously
considered competitors are pulled closer together; categories of firms that were
once distinct begin to converge, and build new cooperative business models.
Processes of technical and industrial convergence create new digitalized
markets: for example, through the proliferation of technical platforms. In the
emerging smart home ecosystem, a good example is Google’s acquisition of
Nest (thermostat, smoke detector, and alarm system hardware). With these
two companies, players as wide-ranging as telecoms, energy corporations,
3 From official presentations made by the ICT company Ericsson under their label The Networked
Society. These presentations are listed and are made available at: http://www.slideshare.net/Ericsson/
industry-transformation-in-the-networked-society
31
gaming systems, and home appliance manufacturers form new Smart home
constellations of both cooperating and competing firms. This includes Google,
thus, positioning itself in relation to operating system or cloud service provid-
ers for smart products and digital selves. And, beneath it all is digitalization,
which enables new potential contributors to enter these emerging networks. A
general implication for management is to develop a preparedness to act upon
unexpected new actors entering these networks; this becomes the new normal
situation in many industrial areas.
9. RE-ORGANISATION OF DIGITALIZING CUSTOMERS:

M ANAGING ORGANISATIONAL BUYER ALLIANCES
The digitalization of companies and industries also affects the buyer side.
Purchasing power moves to executives outside of regular purchasing and IT
departments and functions. Digitization brings a more permanent change to
the ICT investment and buying processes. In turn, this creates a more com-
plex sales environment for technology providers engaged in their customers’
digitalization processes. Furthermore, large-scale digitalization processes –
such as investments in “Smart cities” – create new and very complex buying
situations. Complex constellations of both public and private organisations
need to engage in the creation of functioning buyer constellations in order for
large-scale digitalization processes – such as smart city projects – to move
from idea to pilot, and then to large-scale implementation. Successful exam-
ples, such as the city of Dubai, can be explained by the fact that constellations
on the buyers’ side have been created and engaged in the process. The grow-
ing influence of joint business and public buyer constellations in purchasing
decisions when digitalizing also constitutes a major challenge for the supply
side. Providers of digital solutions will need to support the creation of func-
tioning buyer constellations, thus, creating new go-to-market models that
meet these organisational challenges on the buyers’ side.
10. M ANAGING THE POLITICAL AND

INSTITUTIONAL CHALLENGES OF DIGITALIZ ATION
Digitalization, in general, and the access to and use of big data, in particular,
pose a number of challenges for both companies and policy makers. As one
manager in health care expressed it: “One of the biggest challenges for us right
now is how to relate to policy makers and policy making – often (the) lack of
32
policy making – when it comes to digitalization, in general, and issues con-

cerning patient data, in particular. Understanding how to handle these new
strategic issues and how to influence politicians and policymakers are two of
our major concerns….” Addressing privacy and security issues will become
paramount as more data increasingly travels across boundaries for various
purposes and as a result of increased digitalization.4 Security issues and intel-
lectual property issues are becoming part of companies’ data strategies, yet it
is perhaps becoming more importantly a privacy and trust issue when it con-
cerns customers and other stakeholders. Meanwhile, one of the strategic
challenges for companies in many digitalizing sectors is how to influence
policy makers, so the choices they make also help individual firms in their
quest to capture value from using big data. For digitalizing companies, it is a
matter of complying with the role that policy makers have of developing poli-
cies that balance the interests of companies that want to create value from data
and citizens who wish to protect their privacy and security. This new situation
for companies is often accentuated by the fact that one of the most important
enablers of value creation from big data combines data from multiple sources.
This is a new situation in many sectors and digital data policies are still often
lacking or need to be adapted. Still, this has become one of the major strategic
challenges for many companies: the way in which to relate to – and sometimes
influence – new policy-making. The ownership, access, collection, storage, use,
and dissemination of information require rules and policies; companies and
public organisation tend to become highly involved in these processes.
Digitalization Challenges are Connected

The focus of the ten challenges of digital transformation is different: some are
more closely connected; some are not. Those that are concern technology: for
example, platforms and big data management and user orientation – such as
big data analyses, user-centric and user-network driven actions. Several man-
agerial challenges are business-model related: creating new cooperate busi-
ness models and shifting to service-based business models; some of these also
connect to broader ecosystem-related challenges for digitalizing companies;
4 For example, both public and private organisations need to interpret and relate to various new
principles: for example, OECD’s eight Privacy Principles, which concern: Collection Limitation,
Data Quality, Purpose Specification, Use Limitation, Security Safeguards, Openness, Individual
Participation, and Accountability.
33
thus, creating new business across industry boundaries and managing new
forms of buyer/user constellations.
A digital transformation challenge for an incumbent firm seldom comes
alone. The medical technology company to which we referred in the intro-
duction began one part of its digital transformation with a delimited interest
in how to manage the big data generated from the use of its intensive care
machines. Successively, major business model issues emerged: cooperation or
competition with other machine suppliers and users: doctors, hospital admin-
istration units, and other related issues. Seeing how firms in various sectors
take different initiatives into digital transformation, we can see that digital
transformation steps often lead to new digitalization challenges.
Managers need to cope with the fact that digitalization challenges do not
come alone. We can assume by building upon our first insights into this issue
when going into the fifteen business areas of digital transformation, that
companies take different paths in their digital transformation processes.
There are differences with regard to initial drivers and managerial problems,
as well as when and the way in which these connected challenges are han-
dled. Hence, the emphasis shifts over time. Technical platform issues in this
digital transformation might dominate a certain period; more attention could
be given to user-centric issues and the processes of business modelling or
creating functioning cooperation between involved stakeholders in the emerg-
ing ecosystems might dominate the other stages. Incumbents in many indus-
tries have begun their digitalization journeys; the starting point creates differ-
ent digital transformation paths in different industries and in different
organisational contexts.
The management of digital transformation should not be seen as purely an
intra-organisational or operational issue. Instead, it is a strategic and societal
issue, which is often a challenge of highest priority. As seen in Table 1, the
studies and cases of digital transformation, however, all reveal that the ten
strategic challenges are connected to big internal challenges and tensions
within the organisations.
Implications: Pressures on Internal

Organisational Structures, Processes, and Resources
Digitalization is becoming a dominant driver of internal change: both in pri-
vate and public organisations. For incumbent companies as well as public
34
organisations, digitalization often requires administrations to explore new

opportunities while still operating with mature technologies in mature mar-
kets. Many incumbent organisations experience profound changes in organ-
isational structures, daily operations, and in other modes of doing business.
Thus, dual forces affect internal operations and organisations in profound
ways. Keeping the digitalized business or operations separate or integrating
them with current traditional ones is but one of the organisational challenges
that needs to be managed.
We conclude this chapter by discussing the demands upon internal organ-
isational structures, processes, and resources when entering a digital transfor-
mation. Interviewees have forwarded these issues that were taken from
reports, which represent all fifteen digitalization arenas listed in Table 1 and
have also been compared with other studies: for example, Soule et al, 2016.
Many of the issues described are structural and concern new principles for
organising digitalized operations, as well as internal resources and capabili-
ties. Many of the mentioned internal challenges are change related. The focus
of internal organisational challenges includes managing and leading digital
innovation and transformation processes. Comments from managers centred
around seven general areas: 1) leadership challenges; 2) new skills, resources,
and internal capabilities; 3) customer orientation and customer oriented work
practices; 4) internal organisational structures and responsibilities; 5) internal
processes for continuous experimentation and user orientation; 6) internal
cultural challenges; and 7) change management challenges.
A LE ADERSHIP CHALLENGE
A common view among managers in all sectors is that digitalization can only
be successful if top management support is ensured. The presence of a dedica-
ted CEO and a central team to propel the new digital development is central in
achieving successful transformation. This is not a back office or an IT depart-
ment problem. Some firms and organisations hire a new chief digital officer to
spearhead the changes in a digital transformation. Few, however, believe that
installing a new chief digital officer or one in a similar position is a guarantee
for success. One can argue that digital transformation is a big challenge, and
that support for the process needs to be driven from top management, thus,
fostering a (new) corporate culture. That is to say, the responsibility of creating
a corporate culture that effectively drives digital transformation ultimately
35
rests with the CEO.5 There is a strong common belief that top management is
responsible for setting the digital vision and strategy of the company or organ-
isation. Some also argue that a new type of leadership is needed, thus, moving
away from hierarchical autocratic top-down approaches and looking instead to
create more open collaborative environments, powered through digital collab-
oration tools.6
NE W SKILLS, RESOURCES, AND INTERNAL CAPABILITIES

Many argue that, apart from the many technological resources that are
required to ensure a digital transformation, there is also a need for new human
skills. Interviewed managers often come back to various workforce-related
issues. There is a need for people with skills and experiences with different
digital technologies (cf. Soule et al 2016). When digitalizing, desired employ-
ment skills will include technological experience with social and mobile tech-
nology, artificial intelligence, big data analytics, internet of things, and more.
The competition for such skills and capabilities is expected to increase. The
ways in which to apply these technology skills operationally, tactically, and
strategically will also be part of the skills that will be needed when going dig-
ital. Furthermore, the increasing importance of networked and connected
resources and capabilities means that platforms and skills for such networking
will have an increasingly important role in digitalizing organisations.
CUSTOMER ORIENTATION AND CUSTOMER-ORIENTED WORK PR ACTICES

A central recurrent idea among interviewed managers about the new internal
requirements for digitalizing organisations is the fact that organisations with
digitalization are coming closer to end users: that is to say, customers,
patients, clients, and so on. Digitalizing organisations are becoming more and
more user centric. Actively participating and closer connected users require
internal skills and capabilities – as well as processes and organisational struc-
tures – that support new user empowerment. The challenge for existing
marketing and customer operations includes understanding, following, anal-
ysing, and trying to influence customers’ “digital and analogue journeys”
5 WEF Report: World Economic Forum White Paper: Digital Transformation of Industries: Digital Enterprise,
January 2016
January 2016
36
(Edelman & Singer 2015) across various channels, platforms, and physical
settings.
INTERNAL ORGANISATIONAL STRUCTURES AND RESPONSIBILITIES

During the transition, organising the new digitalized operations in relation to
the old established operations sometimes create “ambidexterity problems” for
incumbents (O’Reilly & Tushman 2004; Birkinshaw et al 2016). That is to say,
the new digitalized operations need to be organised and handled in relation
to existing traditional business operations. How this should be managed in
practice is often challenging and requires organisational skills. In addition,
new organisational processes based upon changed internal and external
connectivity often require new digital platforms in order to enable new forms
of cooperation. Some argue that digitalization and platforms for cooperation
might lead to a shift toward flatter hierarchies. One of the new internal chal-
lenges is creating structures that enable organisations to draw upon networks
both within and beyond traditional organisational boundaries.
INTERNAL PROCESSES FOR CONTINUOUS

E XPERIMENTATION AND USER ORIENTATION
Leadership and organisations in general need to adapt to two central shifts in
operations; the first concerns the need for less focus on long-term planning,
thus, moving toward more continuous development. Traditional internal
roles and responsibilities need to be adapted, including leadership and CEO
roles. Some reports7 argue that, as a consequence of internal digitalization,
leadership needs to embrace the role of moving to shorter data-driven predic-
tive analytical planning cycles with greater focus on experimenting. Digital-
ization leaders need to embrace the role as drivers of new forms of develop-
ment, thus, becoming promoters of continuous innovation processes. The
second challenge concerns the new role of end users; internal organisational
processes and operations need to be changed and coordinated in new ways to
adapt to this new orientation. Understanding and organising in relation to
end users’ digitalization processes is a challenge that requires new leadership
orientation, new internal processes, structures, interactions, and capabilities.
January 2016
37
INTERNAL CULTUR AL CHALLENGES

Creating an effective digital strategy is critical, as is fostering a corporate
culture that is open to innovation and will be supportive of the new strategy.
Digital transformation processes equate cultural challenges, which are often
a part of the new demands facing CEOs and leaders of the transformation.
CHANGE M ANAGEMENT CHALLENGES

Many incumbent organisations are not prepared for digital transformation. A
big challenge is the actual change process. One report8 describes three ways in
which organisations seem to approach the change management challenge
when digitalizing. The first is a direct approach: that is to say, transforming
existing business processes. Initiatives can come from, and within, various
parts of an organisation: marketing, supply chain, production, and operations.
They can also be connected to a number of partnerships of the organisation.
One of the many challenges with the direct approach is how to coordinate
different initiatives and how to engage the entire organisation. The second
approach starts by creating an autonomous “digital unit” free from corporate
legacy and standardised business processes. When it works, the separate unit
(“skunk works” unit) functions as a start-up with no learning barriers or aver-
sions to risk. The second approach has a niche focus upon building new digital
models. The third change approach builds a parallel digital business, which is
a parallel business aimed at repositioning the existing enterprise as a start-up
geared towards digitalization. This becomes a sort of incubator aimed at deliv-
ering new business models and insights that can be leveraged by the tradi-
tional organisation. Overall, there are multiple ways to begin removing barri-
ers to digitalization and to learn from the change processes within organisa-
tions. Irrespective of the change management principle, one of the major
challenges is how to digitally transform the entire organisation and how to
coordinate different change initiatives. Traditional mindsets, practices, and
resources can be difficult to adapt. Developing new (digital) mindsets, prac-
tices, and resources on every level, and within every function, is often per-
ceived as being a difficult and long-term change management issue.
8 From official presentations made by the ICT company Ericsson under its label The Networked Society.
These presentations are listed and are made available at: http://www.slideshare.net/Ericsson/
industry-transformation-in-the-networked-society
38
WHAT CHAR ACTERISES A DIGITAL ORGANISATION?

Our interview findings can be compared to those of Soule et al (2015), who
investigated organisations’ modes of becoming digital and state that the char-
acteristics of a digital organisation could be summed up in four critical areas
concerning mindset, practices, workforce, and resources. First, Soule et al
state a distinguishing feature of a digital organisation is a (digital) mindset: an
attitude that reflects a broad tendency to initially seek digital solutions, use
technology as a tool for advantage, and approach enterprise data in a system-
atic fashion. Secondly, as concerns practices, they argue that four fundamen-
tal behavioural norms are present in a digitalized organisation: first) digitized
operations based upon extensive use of data and information exchange; sec-
ond) a readiness for boundary-crossing collaborative learning, problem solv-
ing, and the discovery of new insight, and data-driven decisions; third, digital
organisations view their workforce in broad terms, thus, acknowledging the
contributions of all when it comes to technology experience, digital skills, and
high engagement in digitalization issues; and fourth, digital tools and data are
critical inputs to the functioning of the digital organisation, and the main
resources are often real-time customer data, accessibility to integrated data
about internal operations, and digital collaborative tools to support commu-
nication, collaboration, and rapid feedback within the organisation.
Conclusion: Connecting External Strategic and

Internal Operational and Organisational Challenges
As aforementioned, digitalization challenges do not come alone. In this chap-
ter, we have elaborated on one broad question: What are the general, strate-
gic, and managerial challenges associated with digitalization? We described
different general paths of digital transformation building upon our insight
from the fifteen areas of digital transformation, arguing that there are strong
overlaps between them: in terms of the accompanying managerial issues that
are involved and managed. For example, big data problems and issues appear
in many of the studied business contexts. There are also differences with
regard to initial drivers and managerial problems, as well as when and the
way in which external strategic challenges and internal operational chal-
lenges are handled. In these terms, the digital transformation paths are
always unique, while the overall external and internal digital transformation
challenges are very similar across industries and organisations.
39
References
Andersson, P., Markendahl, J. and Mattsson, L.G. (2011). Technical development and the
formation of new business ventures – the case of new mobile payment and ticketing
services. The IMP Journal, Vol. 5 No. 1: 23—41.
Andersson, P. and L-G. Mattsson, (2015). Service innovations enabled by the ‘internet of
things’. IMP Journal, Vol. 9 Iss 1: 85—106.
Berman, S.J., (2012). Digital transformation: opportunities to create new business models.
Strategy & Leadership, Vol. 40 Iss 2: 16—24.
Birkinshaw et al (2016). How Do Firms Adapt to Discontinuous Change? Bridging The
Dynamic Capabilities And Ambidexterity Perspectives. University Of California, Berkeley
Vol. 58, No. 4 Summer 2016
Edelman, B. and Singer (2015). Competing on customer journeys: you have to create new
value at every step. Harvard Business Review, Vol. 93 Issue 11: 88—98.
Ehret, M. and J. Wirtz (2017). Unlocking value from machines: business models and the
industrial internet of things. Journal of Marketing Management, 33:1-2: 111—130.
Frenken, K. and Schor, J. (2017). Putting the sharing economy into perspective. Environmental
Innovation and Societal Transitions, 23: 3—10.
Gawer, A., and Cusumano, M. A. (2014). Industry platforms and ecosystem innovation.
Journal of Product Innovation Management, 31(3): 417—433.
Hackling (2013). Strategic Choices in Converging Industries. Fall MIT Sloan Management
Review, Fall: 64—73.
O’Reilly, C.A. and Tushman, M.L. (2004). The ambidextrous organisation. Harvard Business
Review, April: 74—83.
Soule, D., Puram, A., Bonnet, D. and Westerman, G. (2016). Becoming a Digital Organisa-
tion: The Journey to Digital Dexterity. MIT Center for Digital Business, Working Paper #301,
January 5 1016.
Vargo, S.L. and Lusch, R.F., (2004). Evolving to a New Dominant Logic for Marketing.
Journal of Marketing, Vol. 68 ( January): 1—17.
Vargo, S.L., Wieland, H., and Akaka, M.A. 2015. Innovation through institutionalization:
A service ecosystems perspective. Industrial Marketing Management, 44: 63—72.
Westerman, G and Bonnet, D. (2015). Revamping Your Business Through Digital Transfor-
mation. MIT Sloan Management Review, Spring 2015: 2—5.
Yoo Y., Henfridsson O. and Lyytinen, K. (2010). The new organizing logic of digital
innovation: An agenda for information systems research. Information Systems Research, 21(4):
724–735.
40
CHAP TER 2
Reaping Value From Digitalization

in Swedish Manufacturing Firms:
Untapped Opportunities?
M AGNUS M ÄHRING, K ARL WENNBERG, AND ROBERT DEMIR 1
Introduction
Digitalization has emerged as one of the hottest management buzzwords of the
past few years. Media and industry experts forcefully argue that broad-ranging
digitalization is a competitive must and that speed is of the essence (Kiron et al.
2016). This prompts several key questions for companies: for example, are we
“at peak trend”; that is to say, at the summit of inflated expectations? And, will
the next developments move firms through the “trough of disillusionment”
(Burton and Barnes 2017)? Or are we, in fact, not in a trend cycle at all, but in a
massive adoption phase instead: where transformation of companies and indus-
tries will continue and even accelerate (Brynjolfsson and McAfee 2014)? If so,
what are firms actually doing? What influences them? And, is there a gap
between talk and action when it comes to digitalization?
In a cross-national survey conducted in 2015, we asked executives in 400
large firms in Scandinavia, Europe, North America, and the Asia-Pacific
region about their perceptions and respective strategies regarding digitaliza-
tion (Andersen et al. 2015). We found that Scandinavian firms appeared less
concerned with and less active in, pursuing digitalization than did their North
American and Asian counterparts. Top management devoted relatively less
time to digitalization in their strategic dialogue and companies devoted less
attention to acquiring and deploying potentially disruptive technologies.
1 The authors are grateful for financial support from the Peter Wallenberg foundation. Jonas Yakhlef
and Kristina Karlsson provided excellent research assistance. All conclusions and interpretations are
our own.
41
M AGN US M Ä H R I NG, K A R L W EN N BERG, A N D ROBERT DEMI R
Correspondingly, chief information officers (CIOs) predominantly had a

traditional view of competition: for example, not being very active in technol-
ogy monitoring and focusing their activities on providing IT services rather
than supporting or driving business innovation.
A lot has changed since then. As digitalization became a buzzword in the
Swedish business press in 2016, many large corporations expressed their
intentions to become “digital leaders” in their respective industries. Experi-
ence from digitalization and service transformation is scarce and highly
sought-after in the recruitment of executives and directors (Carlsson 2016;
Karlsson 2016). In many firms, developing strategies that incorporate the
challenges and benefits of digitalization is seen as a pressing need.
In this chapter, we take a fresh look at what is actually happening in the area
of digitalization, with a particular focus on the Swedish manufacturing sector.
We sent out surveys by email during the winter of 2016–2017 to 1250 CIOs
and Chief Technology Officers (CTOs) at companies in Sweden that have
been active in a wide range of manufacturing sectors: such as road construc-
tion equipment, gear motor production, agricultural production systems,
precision components, water monitoring and management instruments, and
wood processing equipment. We received 206 completed surveys from the
firms, of which the average grosses approximately 135 million SEK in revenue.
The firms in our dataset are distributed across revenues in the following way:
45 per cent with 500+ million SEK, 40 per cent with 50–500 million SEK, 10
per cent with 10–49 million SEK, and 5 per cent with 1–9 million SEK. In order
to shed further light upon the patterns we see in the survey data, we also draw
upon interviews and focus groups conducted with 18 key decision makers
involved in digitalization projects and initiatives in manufacturing firms.
Our data provides a recent snapshot of digitalization activity levels, prac-
tices, and strategic readiness in Swedish manufacturing companies. We partic-
ularly focus upon patterns in the ways in which they seek to develop innova-
tions and explore new business models from their activities related to product
sensors and wireless data, cloud-based data warehouses, computer-aided
manufacturing and 3D printing, big data technologies, and application pro-
gramming interfaces (APIs). Our findings suggest that while many Swedish
industrial firms have developed a strong edge through a combination of
high-quality products, international presence, and decentralization, the latter
42
R EA PING VA LU E F ROM DIGI TA LI ZAT ION IN SW EDISH M A N U FACT U R ING F IR MS
in particular poses challenges when it comes to digital transformation. Digital-

ization may necessitate large investments across business segments, standard-
isation, and knowledge sharing regarding both customers and digital solutions
in order to create new customer offerings. Points for reflection are then dis-
cussed, along with recommendations for scholars that are seeking to develop
new and relevant knowledge by studying the transformation of Swedish
industry, as well as for managers seeking to benchmark their digitalization
activities to others.
How do Firms Assess the Strategic Importance of

Digital Opportunities in Their Processes and Offerings?
The future is already here — it’s just not very evenly distributed (William Gibson)
We begin our exploration of digitalization activities and strategies among

Swedish manufacturing firms by providing an overview of the overall strate-
gic importance of digital technologies. We posed questions regarding the
opportunities related to both digital processes and digital offerings. Digital
process-related opportunities capture how firms run their operations, includ-
ing logistics and manufacturing processes. These opportunities include con-
nected and digitally (or remotely) controlled production equipment, data
transfer within and across factories to optimise process flows, and the use of
data to analyse and improve processes. Digital offering-related opportunities
capture digital capabilities in products and associated services. Such improve-
ments include the embedding of sensors and software in “smart” products,
enabling the capturing of customer use patterns and product performance,
thus, new kinds of services, new pricing models and, ultimately, new business
models (Porter and Heppelmann 2014). We asked how respondents perceived
their firms to be active in both sets of opportunities during the past two years,
and what they saw as their main foci during the next two years.
43
Strategic importance of digital process-related opportunities

40
35
30
25
Share (%)
20
15
10
0
Not at all important Slightly important Moderately important Very important Extremely important
Small firms (bottom 50% of sample) Large firms (top 50% of sample)
Figure 2.1: Strategic importance of digital process-related opportunities.
Figure 2.1 depicts the strategic importance of process-related opportunities, as

per the 206 responding companies. The black bars display the answers from
the smaller firms: those that gross less than the average 135 million SEK in
revenue; the grey bars represent the answers from the larger firms: those that
gross more than the average 135 million SEK in revenue. Two important
insights can be gleaned from this figure: first, it is apparent that larger firms
are more active in digitalization; secondly, the distribution of responses high-
lights that digitalization is of high strategic importance to many of the Swed-
ish manufacturing firms, and that more than half of all respondents rate
digital process-related opportunities as being “not at all”, “slightly” or “mod-
erately” important. For half of the industrial firms surveyed, the digitalization
of processes is simply not the top priority. Apparently, many firms either do
not perceive the risk of digital disruption in their industry, or they do not
think that developing digital opportunities drives new business opportunities.
Of course, we do not know whether or not this perception is accurate. What
we do know, however, is that over the past few years progressive industrial
firms in Sweden have addressed the challenges of digitalization in their strat-
egies. For example, one of the firms in our sample stated: “The ability to be
connected to the product has been a strategy since 2010”.2 Conversely, a
2 Director, Strategy & Business Development, large industrial firm.
44
breakdown of the respondents’ answers to our question regarding the strate-

gic importance of offering-related opportunities is depicted in Figure 2.2:
Strategic importance of digital offering-related opportunities
50
45
40
35
30
25
20
15
10
5
0
Not at all important Slightly important Moderately important Very important Extremely important
Small firms(bottom 50% of sample) Large firms(top 50% of sample)
Figure 2.2: Strategic importance of digital offering-related opportunities.
A few insights can be gleaned from Figures 2.1 and 2.2 and illustrated with
our qualitative data. Digitalization affects not only the processes; it also
affects the products and services of a majority of firms in our survey. The
challenges are not only – or even primarily – about technology itself; they are
more about how to generate continuous revenues from digital solutions
(Autio et al. 2017). As one of the interviewed executives noted: “It’s not the
technology in itself that’s interesting; it’s the ecosystem and business model
that you can create around the digital content that’s interesting”3.
A common key to creating new business models is to exploit digital offer-
ing-related opportunities: that is to say, those that relate to the firms’ potential
to generate new solutions for their customers. Therefore, we also asked our
respondents about the strategic importance of digital offering-related opportu-
nities (Figure 2.2) in their firms. As aforementioned, the black bars display the
answers from the smaller firms: those that gross less than the average 135
million SEK in revenue; the grey bars represent the answers from the larger
firms: those that gross more than the average 135 million SEK in revenue. On
3 CEO, large Scandinavian industrial firm.
45
average, the firms in our sample reported digital offering-related opportuni-

ties to be somewhat less important compared to process-related opportunities
(cf. Figure 2.1). Consequently, this indicates that Swedish manufacturing
firms perceive current opportunities from digital technologies to lie, to a
greater extent, in process innovation rather than in the potential to produce
new types of products and services to their customers.
In addition to firm size, several other factors contribute to the firm’s per-
ceived importance of digital process-related and offering-related opportunities.
Two plausible reasons are “whether managers have the willingness and com-
mitment to spend the time and effort on pursuing those opportunities”, and
“whether the firm’s IT department (where a great deal of digital competency
tends to reside) is involved in the digitalization efforts”. We posed both of these
questions in our survey, finding a weak yet statistically significant correlation
between the strategic importance of digital process-related opportunities and
the willingness and commitment to spend time and effort on pursuing those
opportunities (r=0.28 and p<0.05). We found no correlation, however, between
the strategic importance of digitalization and the extent to which the IT
department is involved in exploring digital opportunities (r=0.01). This raises
an important question concerning whether internal expertise on digital tech-
nologies—specifically in the IT department—is not being sufficiently leveraged
in developing digital process-related opportunities and, correspondingly,
whether some IT departments do not possess the proper capabilities for digital
transformation efforts. The quality of existing resources, technology, and data
assets may also be poor, forcing firms to explore new and different avenues in
their digitalization initiatives. As one executive told us:
“The data quality of the ERP systems is so substandard that you could not
build anything reliable on that data without first sanitizing the information.
But, then you have destroyed the whole idea of digitalization.”4
From Where do the Influences to

Pursue Digital Opportunities Come?
Without a strong willingness and commitment to spend the time and effort
on exploring digital technology’s business implications, there is a risk that
technologies are being deployed in ways that do not create strategic value.
Our focus group discussion with decision makers in six companies suggests
46
that a key aspect of committing to digitalization in manufacturing firms

might stem from the ways in which external and internal parties influence the
shaping of a firm’s strategic agenda. Therefore, we now explore what kind of
internal and external actors influence Swedish manufacturing firms when it
comes to developing process-related and offering-related opportunities
enabled by digital technologies. We asked our respondents about the relative
importance of various influences on their efforts to develop digital processes
and offerings (see Tables 2.1 and 2.2).
Table 2.1: Important Influences on Manufacturing Firms’ Efforts to Develop Digital Processes
Not
Slightly Moderately Very Extremely
Important
Important Important Important Important
At All
Owners and Board Directors 5,6% 12,8% 23,2% 40,0% 18,4%
B2B Customers 11,2% 9,6% 29,6% 38,4% 11,2%
Non-Management
4,0% 12,8% 36,8% 36,0% 10,4%
Employees
Middle Management 4,8% 10,4% 38,4% 39,2% 7,2%
Subcontractors / Suppliers 6,4% 24,8% 32,0% 32,0% 4,8%
B2C Customers 34,4% 16,8% 22,4% 21,6% 4,8%
Existing Competitors 17,6% 23,2% 35,2% 20,0% 4,0%
Start-ups / New Entrants
25,6% 32,0% 32,8% 6,4% 3,2%
From Other Industries
Consultants 16,0% 32,8% 25,6% 21,6% 4,0%
Media and Public Debate 30,4% 40,8% 16,8% 10,4% 1,6%
Government Institutions 37,9% 29,8% 19,4% 11,3% 1,6%
Table 2.1 displays that the most important influences on industrial firms’
efforts to digitally transform processes come from owners and the board of
directors, followed by corporate (B2B) customers. The emphasis upon owner
influence highlights the importance of corporate governance in setting not
only the strategic agenda in general, but also in articulating the role of digital-
ization in the strategic development of the company (Benaroch and Cher-
nobai 2017; Mähring 2006). The patterns for development of digital offerings
(new, digitally enabled products and services) are quite similar (Table 2.2);
47
however, respondents were less likely to rate three of the major influences on
firms’ efforts to develop digital offerings (B2B customers, non-management
employee, and middle management) as being “very” or “extremely” impor-
tant, compared to the same three influences on their efforts to develop digital
processes.
Table 2.2: Important Influences on Manufacturing Firms’ Efforts to Develop Digital Offerings
Not
Slightly Moderately Very Extremely
Important
Important Important Important Important
At All
B2B Customers 18,6% 16,1% 25,4% 24,6% 15,3%
Owners and Board

12,7% 11,0% 28,0% 38,1% 10,2%
Directors
Non-Management
13,7% 18,8% 29,9% 30,8% 6,8%
Employees
Middle Management 11,9% 17,8% 31,4% 34,7% 4,2%
Subcontractors / Suppliers 17,9% 28,2% 21,4% 24,8% 7,7%
B2C Customers 35,9% 22,2% 17,1% 20,5% 4,3%
Existing Competitors 20,3% 22,0% 33,1% 19,5% 5,1%
Start-ups / New Entrants
29,7% 33,9% 22,9% 11,0% 2,5%
From Other Industries
Consultants 23,9% 30,8% 27,4% 13,7% 4,3%
Media and the Public
33,1% 41,5% 16,1% 8,5% 0,8%
Debate
Government Institutions 46,4% 28,8% 12,7% 9,3% 2,5%
Clearly, the influence from the top in many firms also includes digital leader-
ship of the CEO:
“The godfather of everything regarding digitalization has been our Group
CEO, who has challenged the organization by saying: ‘This thing with digi-
talization, you need to dig it; whether you like it or not, but that’s how it’s
going to be’”.5
48
In many companies, digitalization initiatives are started, guided, and sup-

ported by senior executives (Gregory et al. 2015). Other external and internal
stakeholders are also important:
“If we disregard the customer as being naturally the largest stakeholder in
all of this, it is the service organization and those responsible for service orga-
nizations. Then we have those who actually perform services in different
ways, such as field technicians, sales people, all who are out there… those
who manufacture the goods… those who develop new services”6
Digitalization initiatives are challenging due to their often wide-ranging
and partly emergent impact upon organisation structures, work processes and
work content, as well as due to their challenging project dynamics. For exam-
ple, extensive knowledge sharing and knowledge recombination is often
required, as are repeated changes in goals and shifts in the initiative priorities.
Boundary spanning activities are needed that may or may not result in sus-
tainable working groups supported and resourced by unit and division level
managers. As one executive said: digitalization “requires extensive collabora-
tion and formation of new cross-functional groups… to take the full benefit of
creating an organization that supports big data and IoT services”.7
This suggests that “upper echelons” are important for the activity level in
digitalization; the interests of owners, directors, and executives, which can be
partly driven by trend sensitivity and “copying” behaviours, also need to be
matched by in-depth knowledge amongst lower-level employees and middle
management involved in digitalization efforts. This knowledge needs to
encompass both areas ripe for internal innovation and opportunities that can
be captured in collaboration with customers, suppliers, and entirely new
“third-party” collaborators (Autio et al. 2017).
Both demand side and supply side influences are important. On the
demand side, business-to-business customers are particularly important
sources of influence for manufacturing firms’ digitalization efforts. On the
supply side, empowering employees seems to carry extensive potential for
bottom-up innovation initiatives, though companies also need to be prepared
for some of these initiatives to take the form of “bootlegging” projects hidden
from managers and executives (Criscuolo et al. 2013; Globocnik and Salomo
2015). As one firm reports:
6 Program Manager for Connectivity, large industrial firm.
7 Senior IT executive, large water technology firm.
49
“It all started just over a year-and-a-half ago. I was working as a product
development manager and ran a few smaller pilot projects revolving around
the electronics of our control systems. Initially, I was doing this on my own
but, after a while, I tried to prompt others’ interest around [digitalization]:
acting as an ambassador, trying to get people’s [management’s] attention”.8
A notable difference between processes and offerings is that B2B customers
are seen as having a less important influence on digital offerings than do sub-
contractors. A widely known fact is that innovation and knowledge exchange
activities may take place across organisational boundaries, thus, including
buyer-supplier interactions and even strategic alliances with competing firms.
For example, Toyota developed an organisational unit to better exchange
knowledge within its wide network of suppliers; Nestlé collaborated with Coca-
Cola to develop a distribution model for its hot canned drinks using Coca-Cola’s
expertise in distribution and vending machine network (Dyer and Singh 1998).
One possibility might then be that firms are dependent upon the digital capabil-
ities of suppliers in developing digital offerings, as well as on the extent to which
components supplied by upstream partners can be digitally enabled.
Alternatively, companies might consider locating digitalization activities in
business hubs. For example, one company we studied gathered their analytics
and digitalization activities in a specific location where most of this expertise
was both internally and externally located:
“Here are our consultants and partners... We could not have attracted
employees if we had not been here…” 9
This quotation also leads us to consider the influence patterns that are not
seen as being very important. Firms in our sample pay considerably less atten-
tion to outside influences, and even less to influences outside of their existing
business relationships. Start-ups and new entrants from other industries have
little influence, and even existing competitors are considerably less important
than B2B customers and internal stakeholders. This can be understood as a
natural consequence of organisational structure and interaction patterns; it
also suggests a vulnerability to disruptive forces and a lack of intelligence
activities extending into other sectors (Dyer et al. 2011). Yet, we know that
Uber did not come from the taxi industry, and that Tesla was not an incum-
bent automotive company. This means that incumbent firms need to keep an
8 Senior IT executive, large water technology firm.
9 Program Manager for Connectivity, large industrial firm.
50
eye on new entrants who may take advantage of new digital technologies that
help them overcome the entry barriers that incumbents have set. In other
words, digitalization might lower the barrier to entry for new entrants,
thereby, increasing the threat to incumbents should they ignore the fact that
actors outside of the traditional industry or strategic group boundaries may
use new technologies more advantageously.
A specific domain of internal influence relates to the IT department that tra-
ditionally has been responsible for digital technologies. Our 2015 cross-national
survey revealed that Scandinavian CIOs have a rather “traditional” view on
competition: for example, they choose to focus their activities upon “IT services”
rather than on facilitating business development (Andersen et al. 2015). Thus, we
posed a question to the managers in the surveyed Swedish manufacturing firms
regarding the extent to which their IT department is involved in the formulation
of the firm’s strategy. The result for this question is illustrated in Figure 2.3.
Our IT department is closely involved in the formulation of the organizational strategy
80
70
60
50
40
Share (%)
30
20
10
0
1
Disagree or Indifferent Agree
Figure 2.3: Our IT department is closely involved in formulation of organisational strategy
Figure 2.3 shows that 32 per cent of all IT departments in our survey are
closely involved in the formulation of the firm’s strategy. Compared to the
rather negative results in our previous study two years ago (Andersen et al.
2015), this suggests that Swedish manufacturing firms do value the role of IT
departments. In fact, it is remarkable that a comparatively large proportion of
IT departments take an active role in the formulation of strategy, which is
traditionally the domain of TMTs and income generating units. A plausible
interpretation is that IT departments are shifting in focus from internal
51
efficiency (business process improvement) to supporting external/competi-

tiveness aspects (strategy, digital opportunities). This would suggest a shift
from IT as service delivery toward business process improvement. Such a
shift is likely required to establish credibility for IT to be involved in innova-
tion and strategy work (Mark and Monnoyer 2004).
However, while some IT departments seem to be actively engaged in strat-
egy formulation, it begs the question to what extent are they involved in
exploiting digital opportunities? As previously discussed, there is a common
lack of capability to effectively generate knowledge from information: that is
to say, to store, manage, and mine all of the data generated by products and
production processes in Swedish manufacturing firms.
Which Technologies Underlie

Digitalization Initiatives at Industrial Firms in Sweden?
The specific type of technologies a company uses is a key source of opportu-
nity for enhancement in both the processes and customer offerings using dig-
ital technologies. Thus, we asked respondents to report what kinds of technol-
ogies are currently being used in their respective firms. We asked for a large
set of different technologies, and categorised the answers in the four groups: 1)
Products (sensors in products, wireless data transfer in products, and 3D print-
ing), 2) Manufacturing (computer-aided manufacturing, computer-integrated
manufacturing process, and APIs), 3) Analytics (statistical analysis tools, cloud
base data warehouse, and big data) and 4) Other (computer-aided design and
publication technologies). The results from these questions are reported in
Table 2.3.
Table 2.3: Types of Digital Technologies Currently Used by Swedish Manufacturing Firms
Which of the following digital technologies

No Not Sure Yes
or practices is your firm currently using?
Products (sensors in products, wireless data transfer in products,
59,8% 4,2% 35,9%
and 3D printing)
Manufacturing (computer-aided manufacturing, computer-
45,1% 19,1% 35,0%
integrated manufacturing process, and APIs)
Analytics (statistical analysis tools, cloud base data warehouse,
51,2% 14,6% 35,1%
and big data)
Other (computer-aided design and publication technologies) 46,7% 42,2% 11,1%
52
What do these scores tell us about the type of digital technologies currently
used by Swedish industrial firms? As indicated by the relatively high answers
to “No” and “Yes” across companies, there is clearly a significant heterogene-
ity between firms. Furthermore, as indicated by the relatively high answers
on “Not Sure” for these two categories, the CIOs and CTOs surveyed do not
possess complete knowledge regarding the use of digital technologies in man-
ufacturing processes or for data analytics purposes in their firms. When it
comes to the specific sub-questions, the most common categories within
“Products” are wireless data transfer and sensors in products. A surprisingly
high number of companies also report they are using 3D printing. Many of
the companies surveyed are obviously very advanced in what they do. How-
ever, as a group, they appear to be lagging on big data as part of their analyt-
ics. Current Swedish industrial firms are perhaps struggling in the analytics
domain since the resources and capabilities needed to store, manage, and
mine all the data they generate are often lacking.
These types of technologies are important since they are intimately related
to specific firms’ digitalization strategies and their potential to develop inno-
vations and explore new business models from their activities. However,
corresponding customer-facing idea generation activities are also essential to
create a match between technology adoption and related knowledge acquisi-
tion, and customer engagement in the innovation process. For example, in
order to reap the benefits of digitalization, one company has initiated and
engaged in “research projects around the world… and we have jams and
hackathons and have lots of such fun stuff”10. This suggests that digital inno-
vation is inherently emerging in interaction with the customer. In order to
make this happen, the company has sought to rapidly increase digital service
functionality vis-à-vis customers, as well as initiating collaborations with
leading smartphone handset manufacturers. Their hope is that internal and
external developers will be better equipped to develop new services and apps
related to the product itself. Customers will also benefit by being able to per-
sonalise connected services to their needs, thereby, generating user informa-
tion that feeds back to the company. Through the expertise of external actors
for developing digital solutions that speak to customers’ needs, the company
seeks to create a recursive flow of proprietary data that continuously helps
improving, adapting, and innovating services (Svahn et al. 2017).
10 Senior IT executive, large industrial firm.
53
Future Digital Ambitions of Swedish Manufacturing Firms

We now turn to the firms’ assessments of their respective ambitions and plans
concerning digitalization during the coming two years. Here, it is first notable
that the projection of future digitalization activity is quite ambitious: as much
as approximately 75 per cent of companies expect changes in customer inter-
action patterns and relationships to be somewhat or extremely likely (Figure
2.4). This suggests a rather widespread belief that digital innovation in the
industry will trigger changes in the relationships between manufacturing
firms and their customers. For the individual company, this might be either
caused by own first-mover initiatives or by the need to catch up with competi-
tors’ initiatives. In either case, the distribution of responses suggests that many
firms view the competitive impacts of digitalization as potentially disruptive
for their industry and clearly beyond incremental changes and adjustments.
Next two years: Introduction of new ways to interact with customers and manage relationships
55
50
45
40
35
Share (%)
30
25
20
15
10
5
0
Extremely unlikely Somewhat unlikely Neither likely nor unlikely Somewhat likely Extremely likely
Figure 2.4: How likely is it that you will introduce new ways to interact with customers and manage
relationships in order to meet the changes in the market caused by digitalization?
We next show in Figure 2.5 respondents’ assessment of the likelihood that

their firms will use digitalization to reduce costs from their internal processes
and operations. As is apparent, further cost reductions from the digitalization
of processes are seen to be even more likely to occur than does new ways to
interact with customers and manage relationships. This reflects aspects of
digitalization that are less disruptive, essentially focusing upon cost savings
from process efficiency improvements. This most likely reflects both a wide-
54
spread expectation that regular productivity gains are necessary to remain

competitive and profitable. As such, the responses to this question are not
surprising since productivity improvements to a considerable extent are
driven by IT investments (Brynjolfsson and Hitt 1998). What is perhaps more
interesting is that the percentage of respondents who believe in disruptive
strategic changes involving key external actors is about as large as the per-
centage that expects “business-as-usual” improvement work.
Next two years: Reducing costs from your internal processes and operations
50
40
30
Share (%)
20
10
0
Extremely unlikely Somewhat unlikely Neither likely Somewhat likely Extremely likely
nor unlikely
Figure 2.5: How likely is it that you will reduce costs from your internal processes and operations in
order to meet the changes in the market caused by digitalization?
Interestingly, the correlation between the responses to the two questions in

Figures 2.4 and 2.5 above is quite low (r = 0.28). This suggests that companies
either plan to focus upon changing customer interaction and relationships,
which suggests introduction of new services: for example, through the intro-
duction of digital capabilities in products that allow automated monitoring
and servicing of the installed base at customer sites, or they see themselves as
focusing upon increasing process efficiency. In turn, this suggests that compa-
nies are at different stages of maturity, with the assumption that process
improvement often comes before more advanced rethinking of how the com-
pany delivers offerings to customers (Mark and Monnoyer 2004). As afore-
mentioned, the future might have arrived – but not for everyone.
Managerial Attention to Digitalization:

are Firms Willing and Able to Walk the Talk?
Knowing that expressed ambitions are high in the firms and that owner and
board involvement is important for digitalization efforts, it becomes impor-
tant to ask to what extent management is seen as willing and able to engage
55
in digitalization. We measure this in several ways: we assess the time spent by

respondents (CIOs/CTOs) scouting for input on new digital opportunities
(see Figure 2.6); we look at their commitment of resources to cross-functional
teams (see Figure 2.7); and we focus upon the CEO’s change of leadership,
particularly as it pertains to questioning the status quo and rethinking how
things are doneManagers at our 2.8).
(see Figure firm will devote a large share of their time
to searching for information about digital opportunities
35
30
25
Share (%)
20
15
10
5
0
Strongly disagree Somewhat disagree Neither agree nor disagree Somewhat agree Strongly agree
Figure 2.6: Managers at our firm will devote a large share of their time to searching for information
about digital opportunities.
Figure 2.6 shows that, for a majority of firms in our sample, managers are
unlikely to actively devote a major portion of their time to scout for digitaliza-
tion opportunities (the responses “somewhat agree” and “strongly agree”
summarise to about 25 per cent). However, managers’ willingness to devote
resources to cross-functional teams and to provide leadership for digitaliza-
tion initiatives by challenging the status quo represent approximately 45–50
per cent of firms. This coincides very closely with the percentage of firms that
are more active in their digitalization efforts (cf. Figures 2.1 and 2.2). One way
to interpret these figures is that executives in the surveyed firms seek struc-
tural organisational solutions to digitalization (cross-functional teams and
widespread activity increase) rather than addressing digitalization through a
radical reprioritisation of their personal agendas. A more critical interpreta-
tion of Figure 2.6 would be that it puts into question whether managers are
willing to put their time where their mouth is: that is to say, whether they are
willing to “walk the talk”. Since top-down influences are important, it might
well be that a shortage of managerial attention can hamper the level of digi-
talization activity and/or lead to insufficient managerial guidance of the
efforts that are undertaken. Two of our respondents in large industrial firms
note:
56
“There are some people in top management who appreciate the future
importance of digitization, because they are on the boards of other more
digitized companies. One must consider that these are people who mostly
meet other engineers whose focus is on the functionality of the product itself,
and not how do the company work: business people, organisational structure,
communication, and all these other issues where digitization has a very, very
big impact. For these people, digitalization is limited to automation issues.” 11
“The CEO is extremely central. Replacing a major part of the top manage-
ment team might be necessary to push through the huge cultural change that
When searching for information about digital opportunities, our managers will make
is required.” 12
looking for new information a top priority for how they will spend their time
50
45
40
35
30
Share (%)
25
20
15
10
5
0
Strongly disagree Somewhat disagree Neither agree Somewhat agree Strongly agree
nor disagree
Figure 2.7: When searching for information about digital opportunities, our managers will make
looking for new information a top priority for how they will spend their time.
The answers shown in Figure 2.7 call into question whether top managers are
willing to put their time where their mouth is, so to speak. These findings
relate to our interviews that indicate decision makers might be cognitively
and habitually bound to their existing practices and practice domains, thus,
limiting them from exploring opportunities beyond their core competencies
and areas of personal interest. They also reflect the hectic, issue-packed, and
often response-driven nature of managerial work (e.g., Stewart 1982): where
new domains might be difficult to incorporate into an already full agenda,
particularly when there are knowledge gaps that raise the threshold for, and
cost of, initial engagement (Loch et al. 2017).
11 Program Manager Connectivity, large industrial firm.
12 Director, Strategy & Business Development, large industrial firm.
57
On the other hand, one could argue that the key to successful digitalization
efforts might not be strongly correlated with time spent by top management;
rather, it is by the priority they assign to the task compared to other tasks, as
well as to how they allocate resources to address complex challenges. In par-
ticular, we know that complex and novel challenges often need to be tackled
by teams that break the boundaries of functional silos, thus, potentially
enabling problem solving and solution development to be more creative and
innovative (Love and Roper 2009). We, therefore, asked managers in the sur-
veyed firms about the extent to which they allocate resources to forming and
deploying cross-functional teams (see Figure 2.8). Here, we see a level of activ-
ity, which is much higher than it is for direct top management time allocation.
Our managers formally allocate resources to the use of cross-functional teams
45
40
35
30
Share (%)
25
20
15
10
5
0
Strongly disagree Somewhat disagree Neither agree Somewhat agree Strongly agree
nor disagree
Figure 2.8: Our managers formally allocate resources to the use of cross-functional teams.
Lastly, as a way of exploring whether technical and managerial staff are

actively involved in the implementation of digital solutions, we also asked
respondents to gauge the percentage to which time in executive meetings is
devoted to digitalization, as well as the percentage to which time in research
and development (R&D) is designated to the implementation of digital solu-
tions (see Figure 2.9). Looking at the light grey line, we see that the majority
of Swedish industrial firms in our study – about 87 per cent – spend approxi-
mately 20 per cent or less of their time in management meetings discussing
issues related to digitalization. The dark grey line shows us that 80 per cent of
58
firms use approximately 30 per cent of their R&D time on digital solution
development and implementation.
To summarise, relatively limited time in executive meetings is typically
devoted to digitalization; somewhat more time in R&D is devoted to imple-
menting digital solutions. Our data does not suggest that top executive atten-
tion is undergoing a major shift towards digitalization. Rather, the agenda of
top executives is likely to remain broad and diverse, while also allowing to
focus upon new trends and strategic shifts. The challenge for corporate boards
and executive teams will be to assess how much attention to digitalization is
appropriate, given the current level of, and future potential for, transformation
and disruption in the specific industry within which the company is active.
0,9
0,8
0,7
Cumulative frequency
0,6
0,5
0,4
0,3
0,2
0,1
0
0 10 20 30 40 50 60 70 80 90 100
In the next two years, approximately what percentage of your unit’s or the corporation’s research
and development time will be devoted to developing and/or implementing new digital solutions?
In the next two years, approximately what percentage of Top Executive meeting time
(including off sites) will be devoted to digitalization matters?
Figure 2.9: Percentage of R&D time devoted to the implementation of digital solutions (dark grey) and
executive meetings time devoted to digitalization matters (light grey)
Concluding Discussion
In this chapter, we have sought to provide a snapshot of digitalization in
Swedish manufacturing firms and to display how these firms seek to develop
59
and explore business opportunities from digital technologies. Our survey of

CIOs and CTOs in Swedish manufacturing firms – with 206 firms respond-
ing – suggests that, while digitalization is of strategic importance to many
Swedish manufacturing firms, more than half of the firms in our sample do
not rate opportunities for value creation from new digital processes or digital
offerings as being very important. The finding that digitalization is simply
not a top priority for many firms poses some interesting questions for future
studies and also some challenging questions for corporate decision makers.
Our analyses suggest that the Swedish manufacturing sector currently
experiences a somewhat “digital divide”13 among manufacturing firms. As
digital transformation continues to be driven by the increased attention and
activity of many progressive firms, the digital laggards in the manufacturing
firms risk being left behind; knowledge acquisition and capability develop-
ment may be compromised, and competitiveness could simply slip through
their fingers. As aforementioned, it appears that management in these firms
does not perceive a large risk of digital disruption in their industry and/or
they do not see new business opportunities as being driven primarily by
developing digital opportunities. The future has clearly arrived, yet it is not
evenly distributed. Furthermore, most CIOs and CTOs responding to our
survey perceived current opportunities from digital technologies to lie pri-
marily in their companies’ processes, rather than in the potential to produce
new types of offerings for their customers. This goes against the current
international discourse to a certain extent: where, along with the servicifica-
tion of current products, digitalization is heralded as facilitating the develop-
ments of new revenue-generating products (Brynjolfsson and McAfee 2014;
Porter and Heppelmann 2014). This raises the question whether the firms in
our sample underestimate the risks for disruptive forces as well as the oppor-
tunities of digital offerings. Moreover, those who responded to our survey
may have been more focused upon internal issues rather than market offer-
ings, compared to those that did not respond.
Our analyses also illustrate that, while digitalization is quite high on the
corporate agenda, the actual time spent discussing digitalization in top exec-
utive meetings is limited. Correspondingly, prioritised efforts over the next
two years appear to include using digital technologies to lower costs and
13 Originally, the “digital divide” denoted the risk that underprivileged individuals would be left behind
in the digital era, due to lack of access to IT resources and knowledge (see e.g. Kvasny & Keil, 2006).
60
streamline production and process efficiency, rather than to develop new

customer offerings.
Our survey also shows that managers perceive their current portfolio of
digital technologies to be primarily within manufacturing and in technologies
embedded in products sold to customers. When it comes to technologies
related to analytics (big data, statistical tools, and so on) there is still a relative
dearth.
Summarising, while digitalization seems to have changed from a buzzword
to something of real importance for a large portion of Swedish manufacturing
companies, our study suggests there might still be a discrepancy in terms of
“talking the talk” and “walking the walk” regarding the actual time and effort
spent on developing business opportunities from digital technologies. We
interpret this from the perspective that many Swedish industrial firms have,
thus far, managed to develop a strong edge by focusing upon high-quality
products, an international outlook and footprint, and decentralised decision-
making that concentrates upon serving customer needs wherever they are.
However, the key focus on physical products and decentralised decision-
making also poses challenges when it comes to digital transformation, which
may necessitate large coordinated investments across business segments,
standardisation, and knowledge regarding both customers and tools for man-
aging and using digital information to drive business processes and create
new customer offerings.
For managers seeking to develop and benchmark their digitalization activ-
ities to others, we believe that the global nature of digitalization makes it
necessary for Swedish manufacturing firms to also keep a keen eye on studies
of digitalization processes and advancements in other regions, such as North
America and Asia. Finally, generating value from digitalization often
demands that firms are able to work across departments and avoid thinking
in silos. Knowledge and resources to collect and manage digital data may
reside in IT departments, while customer contacts and ideas for new products
and services are often developed elsewhere in the organisation. Efforts to
improve the customer journey typically require lateral and integrative
approaches. We believe that firms best able to successfully reap the value
from digital technologies will be those that actively avoid departmentalising
questions and, instead, see digitalization as an overarching strategy for their
business processes and business development.
61
References
Andersen, E. S., Demir, R., Mähring, M., Sannes, R., Wennberg, K., & Woerner, S. L. (2015).
Organizing for digitalization – A cross-country study of CIO attention to digital
technology. Working paper: Ratio Institute.
Autio, E., Nambisan, S., Thomas, L. D., & Wright, M. (2017). Digital affordances, spatial
affordances, and the genesis of entrepreneurial ecosystems. Strategic Entrepreneurship
Journal, Forthcoming.
Benaroch, M., & Chernobai, A. (2017). Operational IT failures, IT value-destruction, and
board-level IT governance changes. MIS Quarterly, 41(3) 729–762.
Brynjolfsson, E., & Hitt, L. M. (1998). Beyond the productivity paradox. Communications of the
ACM, 41(8), 49–55.
Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and prosperity in a
time of brilliant technologies: WW Norton & Company.
Burton, B., & Barnes, H. (2017). Hype Cycles Highlight Enterprise and Ecosystem Digital Disruptions:
A Gartner Trend Insight. Stamford, Connecticut: Gartner Group.
Carlsson, S. (2016). Åtta av tio styrelsemedlemmar saknar digital kompetens. DI Digital,
February 21. URL: https://digital.di.se/artikel/atta-av-tio-styrelsemedlemmar-saknar-digital-
kompetens, Accessed: Nov. 12, 2017.
Criscuolo, P., Salter, A., & Ter Wal, A. L. (2013). Going underground: Bootlegging and
individual innovative performance. Organization Science, 25(5), 1287–1305.
Dyer, J., Gregersen, H., & Christensen, C. M. (2011). The Innovator’s DNA: Mastering the Five
Skills of Disruptive Innovators: Harvard Business Press.
Dyer, J., & Singh, H. (1998). The relational view: Cooperative strategy and sources of
interorganizational competitive advantage. Academy of Management Review, 23(4), 660–679.
Globocnik, D., & Salomo, S. (2015). Do formal management practices impact the emergence of
bootlegging behavior? Journal of Product Innovation Management, 32(4), 505–521.
Gregory, R. W., Keil, M., Muntermann, J., & Mähring, M. (2015). Paradoxes and the nature of
ambidexterity in IT transformation programs. Information Systems Research, 26(1), 57–80.
Karlsson, P. (2016). Rekryteringsenkäten 2016: Rekrytering – när teknikutveckling och digitalisering
förändrar jobben: Svenskt Näringsliv, URL: https://www.svensktnaringsliv.se/migration_
catalog/Rapporter_och_opinionsmaterial/Rapporter/rekryteringsenkaten-2016pdf_642897.
html/BINARY/Rekryteringsenkäten%202016.pdf, Accessed: Nov. 12, 2017.
Kiron, D., Kane, G. C., Palmer, D., Phillips, A. N., & Buckley, N. (2016). Aligning the
organization for its digital future. MIT Sloan Management Review, 58(1), 3–26.
Kvasny, L., & Keil, M. (2006). The challenges of redressing the digital divide: A tale of two US
cities. Information Systems Journal, 16(1), 23–53.
Loch, C., Mähring, M., & Sommer, S. (2017). Supervising projects you don’t (fully) under-
stand: Lessons for effective project governance by steering committees. California Manage-
ment Review, 59(2), 45–67.
62
Love, J. H., & Roper, S. (2009). Organizing innovation: complementarities between

cross-functional teams. Technovation, 29(3), 192–203.
Mark, D., & Monnoyer, E. (2004). Next-generation CIOs. McKinsey on IT, 2, 2–8.
Mähring, M. (2006). The role of the board of directors in IT governance: A review and
agenda for research. AMCIS 2006 Proceedings, 3104–3112.
Porter, M. E., & Heppelmann, J. E. (2014). How smart, connected products are transforming
competition. Harvard Business Review, 92(11), 64–88.
Stewart, R. (1982). Choices for the Manager. Englewood Cliffs, N.J.: Prentice-Hall.
Svahn, F., Mathiassen, L., Lindgren, R., & Kane, G. C. (2017). Mastering the digital innova-
tion challenge. MIT Sloan Management Review, 58(3), 14–16.
63
CHAP TER 3
Digital Platforms:
A Critical Review of the Core Concepts
HENRIK GLIMSTEDT
The Inverted Smile

Remember the smile?
Comments on the economics of computer manufacturing in ‘new econ-
omy’ in the 1990s often referred to a U-shaped curve, which illustrated the
uneven distribution of profitability between the different kinds of actors in
the personal computer industry. Both manufacturers of branded personal
computers and the manufacturers of PC clones operated in the shadows of
two specialized component suppliers: Microsoft and Intel. These two busi-
nesses organizations provided most of the value added, whilst they also cap-
tured the lion’s share of the profit pool. Hardware components and software
applications lived somewhere in between those two polar positions, depend-
ing upon degree commoditization.
High Operating Systems (Microsoft)

Processor chipset design (Intel)
Software Applications
Added value and
Value captured
Hardware modules (e.g. printers)

PC System integration
Peripherals (key boards) etc
Subcontracted
Low Manufacturing
Sub-sectors of the PC industry
Figure 3.1. Wintel: The Smile Shaped Curve
65
H EN R I K GLI MST EDT
Recent developments in the contemporary digital economy have turned the

concept of the smile into a less useful template. In particular, system integrating
device manufacturers have risen from rags to riches; successful device integra-
tor manufacturers, such as Apple, sees profit margins far superior to those of
even the more component suppliers. The economic value of operating systems
– once the stellar performers of the digital economy – seems to have eroded
definitively, especially since Google released its free of charge open source OS
for mobile devices1. How about software applications? What about software?
The growth in terms of software output has been staggering since the smart-
phone revolution, especially in the app markets that Apple and Google/Android
have organized. Apple likes to talk about its AppStore, in terms of a booming
business. A little more than 0.7 billion customers/users can now choose from 2.1
million different apps, accounting for 100 billion accumulated downloads (Reis-
inger, 2017). While growth, in terms of output of software, is striking. The
independent software developer’s revenue figures tell us a different story. Game
apps, with market leading Clash of Clans, absorb 85% of the profit pool. For
most developers, the App Store resembles a lottery: for every hit like Candy
Crush, hundreds or even thousands of apps languish in obscurity. In 2016,
Apple paid approximately $50 billion to app developers (Perez, 2016), indicating
that app developers offer a substantial part of all apps for free or at very low
cost (e.g. $1). Forbes already flagged in 2013 that no less than 55% of all for-
profit apps failed to even fetch $1,000 in revenue; only a fraction of the more
successful app development companies reporting revenues above $5000 per
month, thus, concluding that “a hard-working developer on iOS will eventually
be able to get a new car, while Android and Microsoft developers will be forced
into the used car market, if they plan to take those earnings on the road.“(Louis,
2013) As for hardware, the sector suffers from exposure to concentrated
demand: for example, Apple and Samsung. Steep investments in the latest sub-
22 nanometre technologies must also be made at prohibiting costs to all who
remain viable as suppliers. Leading chip vendors, such as Qualcomm and
Broadcom, respond in the same fashion as low margin contract manufacturers
have done for decades: mergers in search of control of even greater production
1 According to numbers released officially by Redmond, Windows mobile operating system dropped
from a 1.2 per cent market share at the end of 2015 to a new low of 0.3 per cent by the end of Microsoft’s
third financial quarter in 2016. Those numbers show that the business of selling operating systems as a
standalone product for profit is dead.
66
DIGI TA L PLAT FOR MS: A CR I T ICA L R EV I EW OF T H E COR E CONCEP TS
volumes to amortize sky-rocking manufacturing costs for the latest generations

of chip technologies. According to Handel Jones, a semi-conductor industry
analyst, only five companies are making sufficient investments to support
leading edge manufacturing capabilities today: down from nearly 20 a decade
ago. Lesser chip vendors either make their exit, or stick with older generations
of processor node technologies: that is to say, >32nm. Whereas software and
hardware companies find themselves entangled in intensive competition and
“price taking”, the major platform companies capture the largest chunk of the
profit pool of the digital economy. Tech strategist Wos Ahmed2 (2016) writes:
“Today, the consumer captures a lot of end use value. The functions in my
iPhone are a testament to this. Nearly all the economic value goes to Apple:
from the application processor through to retail, apps, and services. Foxconn
earns low single-digit operating profit margins and the rest of the value system
– vendors of IP, semiconductors and display panels, etc. – fights over the rest.”
In 2016, the top 15 public platform-based companies represented no less
than US$2.6 trillion of the world market capitalization. Some platforms are
household names: such as Microsoft, Amazon, Apple, Google, and Alibaba.
Others have emerged more recently or hail from parts of the world that get
less attention: such as ARM (Great Britain), Rakuten ( Japan), Delivery Hero
(Germany), Naspers (South Africa), Flipkart (India), or Javago (Nigeria).
Thus, a new general orthodoxy has emerged as a “strategy of last resort” for
tech companies. Writing for Accenture, the global advisory, Lacy, Hagen-
mueller, and Ising (2016) offered the following view to prospective customers:
“Players across industry clusters are entering existing platforms or collabo-
rating to build new products, services, and customer experiences on enabling
platforms. Or businesses are expanding into other industries by using existing
platforms—or creating their own. Previously ‘independent’ products and ser-
vice suppliers are now part of one large competitive set. This leads to a new
landscape where former competitors are now working closely together, and
former collaborators become competitors. And while this expanded competi-
tive circle may seem a threat, it is also an opportunity.”
Observations of value creation and value capture in platform economies
bring business strategies to the forefront, as well as governance of networks
and industry architecture. The billion-dollar question concerns whether the
2 https://www.linkedin.com/pulse/business-semiconductor-part-one-what-happened-woz-ahmed/
67
actions in which a firm takes may, indeed, shape the industry architecture:
with the intent to skew to their advantage capacity for capturing value from
innovations along the value chain.
Executives, consultants, and academics are armed with showcase examples
ranging from personal computers in the 1990s to the contemporary case of
Uber. They push hard to support the idea that modular platform will change
industrial architecture, bringing massive productivity gains, and even con-
tribute to the collapse of old established incumbents. In the late 1990s and
early 2000, a widely believed notion was that modular platforms would
change the architecture of the global automotive industry: shifting the capac-
ity for innovation from incumbents’ OEMs to “first tier mega-suppliers”, thus,
servicing the OEMs with modules the way that Microsoft and Intel innovated
on behalf of HP and other manufacturers of personal computers3. More
recently, loud voices including those at PwC, Accenture, McKinsey & Co,
and KPMG (the list is long) all advocate that “open platform banking” collec-
tively organized by ecosystems of innovative FinTech companies, or some
version thereof, will disrupt the giant incumbents. And, as the argument
runs, that will be the end of banking as we know it (e.g. deJong, Little and
Gagliardi, 2016). The numbers are certainly suggestive. Europe anticipates
banking regulations that require incumbent banks to share proprietary data
through open “application interfaces” 4. Uncertain if they have the right ideas
for open platform banking and financial innovation, incumbent banks and
investors congregate around the new generation of FinTech start-ups. Global
venture investment in FinTech grew by 11%: up to $17.4 billion in 2016; it is
the first time China – with its $7.7 billion of investment in FinTech – outpaced
the US with its $6.2 billion. Ant Financial, formerly Alipay and a subsidiary
of Alibaba, led 2016 with a whopping $4.3 billion venture round: the largest in
FinTech’s venture history (Wintermeyer, 2017). Yet, the jury for open plat-
form banking is still out, leaving us to question whether investments in open
3 According to a Bain & Company report by Donovan (1999), “The new giant suppliers will quickly
move to designing vehicle systems that can be ‘standardized’ within and across OEMs—in other words,
used in multiple models of an OEM and eventually by multiple OEMs.” According to some academics,
autos would mirror IT: “Chrysler has played the role of the Compaq of the automotive industry.
Chrysler’s strategy allows suppliers—even Ford’s and GM’s internal suppliers—to strengthen their
capability to develop whole automotive subsystems, thereby, pushing the entire structure of the
industry from vertical toward horizontal (Fine, 1998, p. 62).”
4 Regulation PSD2 in the European Union, and Open Banking Standard in the United Kingdom.
68
platform banking will be as futile as the billion dollars invested in automotive

modular mega-suppliers around the new Millennium.
Emergence: The Basic Enabling Technologies and Definitions

In a sense, today’s digital platforms are just the more complicated cousins of
two 19th-century innovations: the self-playing piano and Joseph Marie
Jacquard’s famous silk loom from 1801. Both innovations separated pre-
programmed instructions (punctuated paper music rolls and silk patterns
stored in sets of punch cards) and the machinery execution that, in turn,
enabled the formation of rudimentary ecosystems of composers and silk pat-
tern designers5. Modern platforms, of course, support more elaborate forms of
integrations. Generally speaking, we define this by comprising three elements:
a core technology that serves as a foundation, additional modular technologies
that integrate or connect with this core, and the interfaces in-between
(Baldwin and Woodard, 2008; Tiwana, Konsynski and Bush, 2010). Accord-
ing to the loosest possible depictions, the term “digital platform” simply points
to a set of online digital arrangements whose algorithms serve to organize and
structure economic and social activity (Kenney and Zysman, 2016). The core
technology is typically formed around a specific standard (for example, GSM,
VHS, and Ethernet) or the arrangement of standards compiled into an oper-
ating system: such as Microsoft Windows (David, 1985; Farrell and Saloner,
1985; Cusumano, Mylonadis and Rosenbloom, 1992; Besen and Farrell, 1994;
Von Burg, 2001) Especially in business, these arrangements also point to a set
of digital frameworks for social and marketplace interactions. Therefore,
platforms tend to build upon the formation of digital ecosystems. In sharp
contrast to the sequential and linear notion of value creation of Porter’s infa-
mous value chain model (Porter, 1985), the general thrust in later conceptual-
izations of value creation has revolved around the idea of horizontal linkages
and concurrent co-specialization between independent value network partici-
pants or “value constellations” and “value co-creation” as famously proposed
5 Joseph Marie Jacquard’s loom was indeed the first binary information processor. At any given point,
the thread in a woven fabric can be in one of two states or positions: on the face of the fabric or on the
back. Pattern cards were punched or cut according to the required fabric design. A hole in the card
signified that the thread would appear on the face of the fabric, while a blank meant that the end would
be left down and appear on the back of the fabric. The Jacquard head was used on the weaving loom or
machine for raising and lowering the warp threads to form desired patterns based upon the lifting plan
or program embedded in the cards. Thus, the Jacquard mechanism set the stage for modern day binary
information processing.
69
by ( Normann and Ramirez, 1993; Vargo and Lusch, 2004). Indeed, a substan-
tive subset of the literature proposes platforms as the coordinating artefact that
a hub firm uses or the services, tools, and technologies that other members of
the ecosystem can use to enhance their own performance (Gawer and Cusu-
mano, 2002; Iansiti and Levien, 2002, 2004b, 2004a; Li, 2009)6, 7.
Product Platforms: Internal and Supply Chain Platforms

Product development researchers generally see modular product platforms as
the final answer to the question of how to develop and offer a greater variety
of products to different market segments at a reduced cost. Their thinking
starts with Herbert A. Simon’s theory about hierarchies of interdependent
problem having dynamics that are approximately independent to those of
other subsystems (Simon, 1965). Herbert Simon famously said: “Every prob-
lem-solving effort must begin with creating a representation for the problem”
and “solving a problem simply means representing it so as to make the solu-
tion transparent” (Simon, 1996: pp. 108 and 132). In other words, good repre-
sentations can help to illuminate important dimensions of a problem. Simon’s
basic inclination was to differentiate between two approaches to solving
problems in complex systems:
ÜIntegral systems where the sub-system are tied together through a large
number of technical interconnections with un-specified ‘interface rules’
for how different sub-systems work together, and
ÜModularized systems where sub-systems are decomposed and tied
together through a reduced number of technical interconnections with
clearly defined interfaces between the different sub-systems.
6 As Jansen and Cusumano (2012) point out, the field of digital ecosystems is evolving. Originally, the
concept of ecosystem was applied to study how traditional monolithic software service-oriented
software architectures evolved into collaborative architectures: processes in which innovation by
autonomous agents, self-organization, and sustainability were the main topics, More recently, this
previous application of the concepts has faded into the background, giving way to a more strategic
definition. Increasingly, the term digital ecosystem is being used as strategic behavior in digital
business ecosystems.
7 The comparison to biological and natural ecosystem is easily made, but analogies only stretch so far.
The main difference between digital and natural ecosystems is that biological ecosystems are mainly
studied to observe influences from external factors, whereas software ecosystem dynamics are mainly
analysed with the aim of growth and success. Software ecosystems are also made up of participants
harboring intentionality, whereas the beings in a biological ecosystem have no means to consciously be
part of the ecosystem. The largest difference between participants in software ecosystems and those in
natural ecosystems, however, is that participants can consciously decide to exit the ecosystem or even
destroy it in software ecosystems.
70
Simon suggested that problem solving in integral systems will be always

cumbersome and time consuming because modifications to one part of the
system may result in deleterious side effects elsewhere in the system. By con-
trast, Simon’s research suggested that modularization of complex systems into
nearly independent modules allow engineers to modify and improve sub-
systems independently, with limited unintended side-effects in other parts of
the system. Hence, the idea about modular problem solving and the re-use of
modules in ever-increasing ranges of product configurations to meet diversi-
fied demand is simple and powerful. Wheelwright and Clark (1992) famously
introduced the concept of product platforms to describe a framework “for new
products that meet the core customer requirement, but are designed for easy
modification into derivative products through addition, substitution or removal
of features” (p 73). Apart from the reduction in complexity (Simon, 1965;
Parnas, 1972), the advantages of modular product platforms involve economies
of substitution (Garud, Jain and Kumaraswamy, 2002), enhanced customer flexibil-
ity (Baldwin and Clark, 2000), and organizational agility in responding to chang-
ing environmental conditions (Galunic and Eisenhardt, 2001). Moreover,
many have argued that modularity increases innovation (Baldwin and Clark,
2000). Because many firms are involved in the design and production of a
modular system, there are more opportunities for innovation as there are
potential innovators. Further, several firms are involved in the design and
production of a product; its modules increase competition which, in turn, also
spurs experimentation and innovation. Not only are there more potential
experimenters who face increased competition; the costs of experimentation
are also lower, given the fact they are split among multiple firms. Thus, mod-
ularity results in an elevated rate of trial-and-error experimentation, and in
increased competition and innovation on the module-level (Langlois and
Robertson, 1992).
Within the discussion on internal product platforms – particularly in
research on the development of product platforms in machine tools, consumer
electronics, and the automotive industry – product platforms have been
defined by its degree of modularity. Most definitions of product platform
focus upon re-using and sharing common elements – or use-cases – across
complex products. The focus has been placed, to a large extent, on four
dimensions: architecture, platform, modules and design rules (Ulrich 1994;
Baldwin & Clark, 2000).
71
ÜPlatform: The collection of bundled technological [physical] assets that

are shared by a set of related products.
ÜArchitecture: The [abstract] scheme by which the functions of a product
is allocated to physical components, also defined as modules.
ÜModule: A unit whose internal structural elements are powerfully and
integrally connected among themselves and relatively weakly connected
elements in other modules.
ÜDesign Rules [or interfaces]: The principles that govern the relationship
between modules.
The general established criteria are that a platform embodies certain con-
straints or design rules, in terms of interfaces (or crossing points) between
components; these govern the relationship between components. At these
points, the interdependencies between components are defined by the inter-
faces, whilst other forms of interdependencies are ruled out by the design
rules Therefore, internal product platforms refer the to modularization of
complex system in which the platform itself remains stable, while modules are
encouraged to vary in a cross-section or over time. The most stable element
in a platform is the interfaces that control the mediation and point of interac-
tions between modules. In turn, this defines the degree of modularity.
This product-oriented definition emphasises commonality of the systematic
re-use of components across different products within a product family, which
allow economies of scope in production to occur. Hence, the systematic cre-
ation and harnessing of economies of scope and mass-customization in innova-
tion can be seen as one fundamental principle of platform-based new product
development. Led by these inclinations, empirical studies within the product
engineering studies have identified that these kinds of economies of scope can
occur in a variety of industrial contexts (such as semi-conductors, machine
tools, commercial aircraft manufacturing, automotive manufacturing, aero-
space engine manufacturing, and consumer electronics). Expanding the focus
from internal product platforms, such as Black & Decker’s successful operation
of its much-discussed internal platform for consumer electrical hand tools,
empirical research has also documented how manufacturing platforms were
increasingly being shared across firms within supply chains. (Helper and Sako,
1995; Helper, MacDuffie and Sabel, 2000; Brusoni, Prencipe and Pavitt, 2001;
72
Sturgeon, 2002; Zirpoli and Caputo, 2002; Becker and Zirpoli, 2003; Doran,
2003; Berger, 2005; Brusoni, 2005; Huang, Zhang and Liang, 2005; Park et al.,
2009; Sako, 2009; MacDuffie, 2013; Jacobides, MacDuffie and Tae, 2016).
Industry Ecosystem Platforms

(Product Innovating and Transactional)
Parallel to the evolution of platform thinking in product engineering, various
scholars and industry observers began to discuss the networking of personal
computing technologies’ “industry-wide platforms” for information technol-
ogy; these include Michael Bourrus & John Zysman (1997); William Lazonick
(Lazonick, 2005, 2009); and, Michael Cusumano & Anna Gawer (2002).
Zysman and Bourrus, coined the phrase Wintelism, to describe the rise of a new
industry platform, competing against the vertically integrated computer man-
ufacturers. The aforementioned research originally drew upon insights from
business history on how the policies and processes (e.g. anti-trust) led to the
raise of independent software companies a new generation of merchant chip
manufacturers, epitomized as the Fairchildren. Eventually, this new generation
American tech firms experienced the rise of Japan’s successful semi-conductor
manufacturers, which fiercely competed upon both price and quality in the
late 1970s and early 1980s. This cut-throat context in semiconductor technolo-
gies, such as memory chips, lead tech firms in Silicon Valley-based pioneers to
more advanced technologies in personal computing, which they identified as
way of diversifying into new and less competitive segments. Hardware special-
ists such as Intel, however, lacked the competence to develop and market
complex consumer electronics products. While the initial attempts to diversify
into computers and other consumer products failed (for example, calculators),
Intel and other Silicon Valley tech firms purified their specialisation strategies.
They particularly embraced the idea of platforms linked together with sophis-
ticated and less advanced components, thus, according principles of “open-but-
owned” systems of standards8. The making of a new Wintelist era enabled a
dramatic shift in the character of electronics production, moving away from
8 Key product standards under Wintelism, especially the interface specifications that permit inter-oper-
ability with the operating system or system hardware, are owned as intellectual property, yet are made
available to others who produce complementary or competing components, systems or software
products. Hence, the systems are “open-but-owned”. The relevant technical standards are licensed
rather than published, with either the universe of licensees, the degree of documentation of the
technical specifications, or the permissible uses restricted in some fashion.
73
the dominance of vertically-integrated organizations that were built upon a

closed-proprietary standard over to a decentralized value chain that both col-
laborated and competed within platforms, such as the PC.
More recently, platforms have been found to operate within even larger
networks of firms that are not necessarily linked through buyer–supplier
relationships. This is also known as “innovation ecosystems” (Adner and
Kapoor, 2010; Nambisan and Sawhney, 2011) or “ecologies of complex inno-
vation” (Dougherty and Dunne, 2011). Such industry platforms are then defined
as ‘ …a building block, providing an essential function to a technological
system, which acts as a foundation upon which other firms, loosely organized
in an innovation ecosystem, can develop complementary products, technolo-
gies, or services (Tee and Gawer, 2009).
At the end of 2016, four of the top five public firms by market capitalization
used platform business models. An open platform business model offers dis-
tinct economic advantages since it allows a firm to harness external innova-
tion as a complement to internal innovation (Chesbrough, 2003a). While
prevalent in information intensive industries such as search (Google), operat-
ing systems (Microsoft), and video games (Sony), open platforms have
emerged in aerospace (Lockheed Martin), food spices (McCormick), T-shirts
(Threadless), 3-D printing (MakerBot), and shoes (Nike). Thus, industry
platforms are mainly viewed by a research hub or a central point of control
within a technology-based business system (Gawer and Cusumano, 2002,
2008; Huang et al., 2013). Cusumano (2010) has argued that that an industry
platform differs from product platforms in two ways:
ÜWhile similar to an internal product platform in that it provides a com-

mon foundation or technological system that a firm can reuse in differ-
ent product variations, the industry platform defines a more or less
“open” technological system whose components are likely to come from
different companies (or maybe different departments of the same firm),
which we call “complementors.”
ÜThe industry platform has relatively little value to users without these
complementary products or services.
As an example, Cusumano maintains that the Wintel PC or a smartphone
are just “boxes with relatively little or no value without software development
tools and applications or wireless telephony and Internet services” (ibid, p.
74
33). The company that makes the platform is unlikely to have the resources
or capabilities to provide all the useful applications and services that make
platforms such as the PC or the smartphone so compelling for users.
Hence, in order to allow their technology to become an industry wide
platform, companies generally must have a strategy to open their technology
to complementors and create economic incentives (such as free or low licens-
ing fees, or financial subsidies) for other firms to join the same “ecosystem”
and adopt the platform technology as their own. A third key point is that, as
various authors have noted, the critical distinguishing feature of an industry
platform and ecosystem is the creation of network effects. These are the power-
ful feedback loops, which also are referred to as demand-side economies of scale
(Katz and Shapiro, 1986), that can grow at geometrically increasing rates as
adoption of the platform and the complements rise.
Central to industry platforms, by the way of summary, appear to be the
combined logics of platform leverage and architectural openness. As its most
basic definition, platform leverage refers to a process of generating value and
market impact that is disproportionally larger than the input required in
other types of value chains: for example, integral (non-modular) architec-
tures. In the area of strategic management, platform leverage is directly linked
to the organization’s sustainable competitive advantage. Following Thomas,
Autio and Gann (2014):
ÜProduction leverage is based upon the (re) use of a collection of assets and
the interfaces and standards that enable sharing these to drive econo-
mies of both scale and scope. In the case of product families, the reuse of
production assets and product components helps to realize both scale
and scope economies through reduced manufacturing costs and
improved design quality, such as better product architecture.
ÜInnovation leverage is similarly based upon the (re) use of a collection of
assets and the interfaces and standards that enable sharing. However,
instead of sharing to achieve economies of scale and scope, the goal is to
drive economies of innovation and complementarity and, hence, facili-
tate the creation of new goods and services. When the product family is
extended to supply chains and the platform system is decoupled from the
focal firm, potential innovation benefits also emerge in the form of com-
75
ponent innovation, enhanced by the distribution of self-interested deci-

sion making across the ecosystem among competing complements.
ÜTransaction leverage, in contrast, is based upon the manipulation of the
market pricing mechanism and market access, which drives transaction
efficiency and reduces search costs in the exchange of goods and ser-
vices. In the same manner as a conventional market intermediary, a
platform ecosystem extracts the surplus value generated by leveraging
its position as a value hub linking multiple sides of the market. In this
sense, the platform ecosystem leverages its position within industry
architecture to benefit from the economies of transactions and search.
Given the critical importance of complements and network effects, the key
ways of defining industry platforms revolve around concepts of degree of
platform openness and the governance of complements. Before discussing the theme
of governance, we will briefly touch upon the economics of platforms, com-
bining demand-sided economies of scale (or network externalities) with
forceful lock-in effects.
Value Capture in Digital Platform Ecosystem

Demand-side economics of scale is widely held as constituting the driving
force of digital platforms. Researchers have developed an explicit platform
theory to explain how, despite interdependence in technologies and comple-
mentary assets, some technology firms can control an industry’s value chain
and capture a disproportionate share of the total value. This draws primarily
upon the literature on technological standards (David, 1985; Katz and Sha-
piro, 1985; Besen and Farrell, 1994), network economics (Katz and Shapiro,
1985; David and Bunn, 1988) (David Bunn, 1988; Katz & Shapiro, 1985 and
1994) and multi-sided markets (Evans, 2003b, 2003a; Rochet and Tirole, 2003;
Evans and Schmalensee, 2010). As many have observed, a first key point is
the critical distinguishing feature of an industry platform and ecosystem is
the creation of critical mass and network effects: for example, (e.g. Molina,
Bremer and Eversheim, 2001; Evans and Schmalensee, 2010; Prasarnphanich
and Wagner, 2011). These are positive feedback loops that can grow at geo-
metrically increasing rates as adoption of the platform and the complements
rise. Thus, the network effects can be very powerful.
76
T WO-SIDED-M ARKETS
Digital platforms serve as integrators and bottlenecks in two-sided markets
(Roson, 2005; Hagiu and Hałaburda, 2014). They can take many guises and
provide infrastructure and rules that facilitate the two groups’ transactions.
In some cases, platforms rely upon physical products, as with consumers’
credit cards and merchants’ authorization terminals. In other cases, they are
places that provide services, such as shopping malls or websites: Monster,
eBay, and so on. Two-sided networks differ from other offerings in a funda-
mental way. Value moves from left to right in the traditional value chain,: cost
is to the left of the company; revenue is on the right. Since the platform in
two-sided networks has a distinct group of users on either side, cost and reve-
nue are both found on the left and on the right: for example, as is the case of
Google Search, which is subsidized by revenues from the advertising busi-
ness: Google Ad.). In this case, the perspective shift goes from supply-side
economics to demand-side of economics.
In two-sided markets, the number of agents on the other side determines
the value that an agent derives from joining a platform: that is to say, the
cross-group network effects. Examples include payment systems such as Pay-
Pal or Visa, videogame systems such as PlayStation 3 and Xbox 360, smart-
phone platforms similar to Apple’s iPhone or Google’s Android, and so on.
According to Parker and van Alstyne (2005), two-sided markets require the
interaction of three groups of actors: a group of technology buyers, a group of
sellers, and an intermediation “platform”, which creates tools or mechanisms
for helping both parties strike a deal9. It works like this: a company quickly
enters a new market and attracts customers, and those customers attract more
customers, and so on. In turn, the first mover experiences explosive growth
and assumes a dominant market position while earning wonderful profits.
The most important aspects of the network effect are that the more external
adopters in the ecosystem that create or use complementary innovations, the
more valuable the platform (and the complements) become. This dynamic,
driven by direct or indirect network effects or both, encourages more users to
9 More precisely, according to Rochet and Tirole (2006, pp. 664-665)“a market is two-sided if the platform can
affect the volume of the transactions by charging more to one side of the market and reducing the price paid by the other
side… The market is one-sided if end-users negotiate away the actual allocation of the burden; it is also one-sided in
the presence of asymmetric information between the buyer and the seller, if the transaction between buyer and seller
involves a price determined through bargaining or monopoly”.
77
adopt the platform, more complementors to enter the ecosystem, more users
to adopt the platform and the complements, almost ad infinitum.
NET WORK EFFECTS: DIRECT AND INDIRECT

Economists developed the theory of network effects in the 1980s, and bur-
nished it in the 1990s. Business gurus, entrepreneurs, and the tech media
cherished it as one of the guiding lights of the new economy. Two concepts
are central here: critical mass and “indirect” network effects; it is widely rec-
ognized that sufficient value from the use of products such as the telephone,
fax machine, or other networked services is closely associated with ‘critical
mass’ (Rohlfs, 1974). Without such a critical mass (defined as “a minimum
network size that can be sustained in equilibrium”) users will not receive
sufficient value, and growth will not continue (Oren and Smith, 1981; Econo-
mides and Himmelberg, 1995).
Strong direct network effects builds upon the number of users. A telephone
becomes more valuable to an individual as the total number of telephone
users increases. Following the received wisdom of Metcalf ’s Law, the value of
a platform is attributable to the size of the network: that is to say, the number
of nodes (Gilder, 1993; Metcalfe, 2013)10. Companies or platforms compete
through creating “bandwagon” among the users that make the outsiders to
those bandwagons experience a loss of value, and even an extra cost of
remaining outsiders. Therefore, network industries often involve ‘tipping’ at
a certain point at which the joint existence of two incompatible products may
be unstable, with the possible consequence that a single product and standard
will dominate. Given the idea that the numbers of nodes – that is to say,
complements and users – determine the value of a network, economists see
markets with strong network effects as being prone to a “winner-takes-it-all-
outcome” (Katz and Shapiro, 1985; Boschma and Lambooy, 1999).
Strong indirect network effects arise when critical mass of complementary
products – for example, hardware and software – enable users to receive suffi-
cient value from the use of the networked technology. As the variety of avail-
10 Research in industrial economics introduced the concept of network externalities (popularized as

Metcalf’s Law) to describe a situation in which “the utility that a user derives from consumption of the
good increases with the number of other agents consuming the good” (Katz & Shapiro, 1985: 424). The
“number of other agents consuming the good,” often referred to as total network size, is defined in a
straightforward way: “The network size is simply the total number of consumers owning units of
hardware that are compatible with the individual’s unit” (Katz & Shapiro, 1992: 59)
78
able DVDs increases, a DVD player becomes more valuable through indirect
network effects, and this variety increases as the total number of DVD users
increases. A major stream in the literature on indirect network effects demon-
strates how the value of ownership of core products – for example, phones,
VHS and DVD players, game consoles, and other networking technologies –
increases with the number of complement products. Standardisation, there-
fore, is a likely outcome (Gandal, Kende and Rob, 2000; Dranove and Gandal,
2003; Gandal, Salant and Waverman, 2003; Rohlfs, 2003; Clements and
Ohashi, 2005). Theory also suggests that such effects should drive faster mar-
ket growth due to the bandwagon effects (Shapiro and Varian, 1998; Rohlfs,
2003). Shapiro and Varian (1999) first attributed network externalities to posi-
tive feedback and then suggested that “if a technology is on a roll…positive
feedback translates into rapid growth: Success feeds on itself” (p 176).
Research also point to the opposite effect of slowing growth in what is
sometimes labelled “excess inertia” (Srinivasan, Lilien and Rangaswamy,
2004; Goldenberg, Libai and Muller, 2010; Peres, Muller and Mahajan, 2010).
Early in the product life cycle, most consumers see little utility in the product,
as there are few adopters; therefore, they may take a “wait-and-see” approach
until there are more adopters. Hence, diffusion early on may be very slow and
occur among the few consumers that see enough utility in the product even
without adoption on the part of other consumers. Therefore, the process may
be characterized by a combination of excess inertia and excess momentum:
that is to say, slow growth followed by a surge (Rogers, 2003).
Managing the Degrees of Openness

Governance of platforms and strategizing within platforms gradually
becomes relevant for companies that are trying to establish themselves in the
digital economy. Decentralization of value chains is not without its ambigui-
ties. First, the building of decentralized value chains for a more or less perma-
nent innovation economy invites free riding and opportunism, thus, raising
questions about governance of management of networks. (Piore and Sabel,
1984; Jarillo, 1988; Dyer and Singh, 1998). The central dilemma of growth lies
in reconciling the demands of learning with the demands of monitoring
(Sabel, 1994). In this view, there is a contradiction between openness to attract
and coordinate learning for innovative product or services with the control
over value capture and the distribution of the gains from the collaboration
79
within the ecosystem. Here, it is easier to write contracts between partners

within the ecosystem that cover contingencies associated with transactions
between them when the market conditions are stable and economies of scale
are, thus, predictable. Innovations undermine stability because they disrupt
the regularity of markets. Hence, the dilemma is that learning and innovation
within the ecosystem undermines the stability that is normally required for
value capture and monitoring insofar as each transacting party in the collab-
oration fear possible hold-ups: that investments will not be matched and the
terms of value capture from the investments remain uncertain.
With concepts such as system integrators (Hobday, Davies and Prencipe,
2005), architectural capabilities (Henderson and Clark, 1990; Galunic and Eisen-
hardt, 2001; Roy and McEvily, 2004; Jacobides and Winter, 2005; Jacobides,
Knudsen and Augier, 2006; Baldwin, 2015; Jacobides, MacDuffie and Tae,
2016), architectural knowledge control (Henderson and Clark, 1990) and platform
leadership (Gawer and Cusumano, 2002, 2014), iterative pragmatic collaboration
(Helper, MacDuffie and Sabel, 2000; Gilson, Sabel and Scott, 2009) research
tries to describe that decisive capacity to coordinate collaborative knowledge
creation and simultaneously manage value capture.
Gawer and Cusumano (2002, 2008) argued that the main problem of plat-
form leaders11 can be identified in two key features of contemporary plat-
forms: the increasing interdependency of products and services and the
increasing ability to innovate by more actors, especially in the high-tech sec-
tors. The combined effect of these two elements determines that the evolution
and improvement of one element in the product/service/organisation of the
platform is complementary and interdependent to the development of all
other elements. Furthermore, they focus upon how firms can drive industry
innovation and “architect” or influence competition through four particular
“platform levers”:
ÜFirm scope: The choice of which activities to perform in-house versus

what to leave to other firms. In particular, this decision is about whether
11 Iansiti and Levien (2004) also differentiate between two types of platform leaders: “keystone” and
“dominator” leaders. In particular, the keystone leader has developed capabilities from which to benefit
and, at the same time, generate significant externalities within the platform in order to sustain the
collective performance. Keystone leaders strike a productive balance value appropriation and value
sharing between platform’s partners. By sharp contrast, the “dominator” leaders integrate vertically and
horizontally this in a predatory way, seeking to appropriate most of the value produced by the network.
80
the platform leader should make at least some of its own complements
in-house.
ÜTechnology design and intellectual property: This refers to what functionality
or features to include in the platform, whether the platform should be
modular, and to what degree and at what price the platform interfaces
should be open to outside complementors.
ÜExternal relationships with complementors: This is the process by which the
platform leader manages complementors, and encourages them to con-
tribute to a vibrant ecosystem.
ÜInternal organization: This regards the way and the extent to which plat-
form leaders should use their organizational structure and internal pro-
cesses to give assurances to external complementors that they are genu-
inely working for the overall good of the ecosystem. This last lever often
requires the platform leader to create a neutral group inside the com-
pany, with no direct profit-and-loss responsibility, as well as a Chinese
Wall between the platform developers and other groups that are poten-
tially competing with their own complementary products or services.
Gawer’s and Cusomano’s highly influential book has created a bandwagon of
related research, converging on a general theory on platform leadership.
Essentially, there is a general agreement on the critical role of establishing an
“optimal degree of openness”, which ensures wide ecosystem participation
and positive network externalities, while still leaving the control of the core
element of the platform firmly in the hand of the platform leader in order to
ensure the disproportionate distribution of value captured. First, the literature
indicates that platform owners face a key challenge in designing the structure
of their platform, such that they maintain ownership and control over the
critical elements that deliver value. For example, platform owners must deter-
mine the optimal “openness” of the platform in terms of interoperability,
disclosure of IP, and collaboration with complementors that will spur innova-
tion and network effects (Chesbrough, 2003a; West and Gallagher, 2006;
Parker and Alstyne, 2008). Secondly, the firm must balance these require-
ments with the need to maintain control of the platform in a way that allows
it to capture value in a sustainable fashion (Boudreau, 2010; Eaton et al., 2015).
This tension is present in strategies for day-to-day governance, which include
81
determining the boundaries of innovation and value capture by the platform

owner and complementors (Eaton, 2012; Tilson, Sorensen and Lyytinen,
2012; Ghazawneh and Henfridsson, 2013), as well as pricing and other
revenue generation strategies for each side of the market (Rochet and Tirole,
2003).
Research on “industry architecture” takes a more structured approach to the
analysis of platform governance. Jacobides, Knudsen and Augier (2006) say
firms attempt to strategically develop architectural competencies to determine
the firm’s vertical and horizontal specialization, as well as influence the institu-
tions that shape markets. Thus, researchers explore how firms shape and
redefine the strategies and templates that determine “who does what” in a sec-
tor, because they appreciate that this will affect “who takes what” ( Jacobides,
Knudsen and Augier, 2006). Departing from a structural view on agency, the
industry architecture literature suggests that a platform firm can intentionally
construct the value network in such a way as to create barriers of entry for its
own position, while increasing competition in other nodes around its network
location, thereby positioning itself as the “bottleneck” or “control points”,
which is defined as the location in the platform that extracts most value while
locking-in customers most forcefully (For a revew, see: Ballon et al., 2008).
In the PC industry and other related sectors, the OS and application layer
have famously been the locus for value capture (for example, Microsoft)12;
with the advent of the Internet, however, the recent opportunities for control
and profitable growth have migrated upward, away from the operating sys-
tem layer and into the software application layer, which is higher in the IT
stack (for example, Google’s search function and other online applications13).
12 Though most persons attribute Microsoft’s dominance to its control of the Window’s OS, equally, or
perhaps more important, is the Microsoft Office productivity suite, which is the consumers’ connection to
Microsoft and is likely more important for the mindshare lock-in than the desirability of Windows. More
precisely, each new improved generation of Microsoft Office package also involves up-graded file formats
(e.g. docx for Word). Once users start upgrading –they might be universities, large companies, parts of the
government, or public institution—all users with older versions will find that sharing and opening of the
new files will be more complicated (for example, they might save in older formats with a loss of functions
and formats) unless they also upgrade to the latest version of MS Office. In that sense, Microsoft’s position
in the computing depends upon the proprietary file format rather than the operating system.
13 As Jonathan Murray, Microsoft Worldwide CTO, revealed in a private conversation with the present
author, Microsoft’s concern of course is that large public sector customers start pressing for support of
widely recognized non-proprietary formats, such as XML for Excel, which would reduce the pressure
on up-grading and increase the capability between MS products and free software offered by, say,
Google as cloud services.
82
Research on Android shows that Google’s launch of its open source oper-
ating system diluted the existing OS-based bottlenecks in general – particu-
larly, Microsoft’s position – whilst also generating a shift in the locus of con-
trol points. Pon, Seppälä and Kenney (2014) demonstrate, based upon case
studies of evolution of control points and gate keeper roles in Google’s
Android, Amazon and Xiaomi, that Amazon and Xiaomi have built new
bottlenecks by designing complementary services on top of Google free OS
to create new bottlenecks and find ways to lock-in customers. Google’s open
source OS allowed Amazon and Xiaomi to tap into Google’s massive installed
base and offer a significant short-cut, meaning they could forego the massive
investments in attracting users into a two-sided market and, instead, allow
them to focus their resources upon providing value-adding services for two-
sided markets. Amazon builds its own versions of Android application inter-
faces to offer unique services that other Android-powered platforms cannot
match. In this fashion, Amazon extends, for example, its popular AWS cloud
services into its line of Kindle and Fire-tablets, thus, creating strong incentives
for developers to focus upon Kindle and Fire-tablet applications while locking
in customers at the other end of this two-sided market. Pon, Seppälä, and
Kenney (2014) follow this same logic by also revealing that Google responded
by raising the bar for mobile OEMs that seek to implement Android without
adding Google Mobile Service apps: for example, Maps, Gmail, Google
Drive, Calendar, and Search. Thus, Google quietly adds the highest value-ad-
ding innovations to a proprietary version known as “certified” Android.
Mobile phone OEMs that aim to offer its own version of added-value services
on top of the open source Android will find the “bare” open source OS
becomes less and less competent compared to the certified version, thus,
increasingly demanding more and more resources to turning the open source
version into a competitive offering.
General Caveats
In this chapter, I attempt to provide a first impression of some of the more
central ideas that guide our understanding of the evolution of the digital
economy. These generalizations must, however, be taken at face value – at
least partially. Some of the cornerstones – particularly, advantages from mod-
ularity in design and production, and network effects, may be weaker than
often agreed. As for management implications, the consequences might be
83
significant. If the digital platforms strongholds of monopoly power are

weaker than presumed, the chances of dethroning digital platforms in various
sectors of the economy are not as bleak as we may think.
NET WORK EFFECTS: HOW STRONG IS E XPL ANATORY POWER?

First, the existing conceptualisations of network effects do not fully explain
observable market outcomes, especially when it comes to technology adop-
tion. If consumers would have based their decision on network size, it could
be hard to explain market outcomes in wireless standards. Take cellular
standards as an example: by the early 1990s, the AMPS systems ranked, by
far, as the most successful standard in the world. By 1991, wireless operators
in 21 countries had adopted AMPS, accounting for approximately 75% of the
world subscribers (Garrard, 1998). Even if AMPS originally outnumbered
European standards in terms of connected phones, the European GSM stan-
dard gained traction in the 1990s and the first decade into the millennium.
Two decades later, the GSM family of standards (GSM, UMTS, LTE)
became the dominant global standard.
In a similar way, we must also ask how well does network theory really
explain why Symbian and Blackberry did not win against iOS and Android
in the race for dominance in smartphone operating. By 2009 – that is to say,
two years after Apple’s introduction of iPhone and Google’s decision to offer
Android for free – Symbian’s global market share was at 60%, whereas the
second network in terms of size (RIM/Blackberry) had captured 20% of the
global market through its dominating position in the US market. At 60%
dominance, the market should have tipped in the favour of Symbian. Yet,
Android and iOS stormed in. If the number of nodes determines the value of
a network, then Google’s Android arm, or Samsung’s mobile phone business,
would be valued higher by investors than Apple’s. What is noteworthy is that
Apple’s 75% market share of the global MP3-player fell apart the moment
entrants such as Nokia, Samsung, and Sony-Ericsson entered the music plat-
forms with music phones, even though their initial platforms lacked the
strong network externalities, such as Apple’s iTunes services, which con-
nected the music industry to Apple’s platform product. Similarly, in video-
games, several competing platforms fight for dominance; however, smaller
players still hold strong in profitable pockets of the market. The deviations
from predicted outcome are significant in all cases. What anomalies such as
84
these suggest is that other factors in network externalities theories besides

network effects are at play. Network effects might, indeed, explain market
outcomes cetris paribus. But then again, how often is everything equal in the
innovative platform economies?
THE LIMITATIONS OF MODUL ARIT Y

Closely related to the above, we need to also treat the concept of modularity
with analytical care, which explains the seamless integration of positive net-
work externalities (or complements) into the platform environment. Following
Simon’s notion of nearly decomposable systems, modularity has been a key
concept in system design as well as in the discussion on the rise of highly spe-
cialized supplier networks. Modularity allows firms to apparently respond
more quickly and flexibly to shifts in product markets since modular architec-
ture reduces the cost of providing a greater variety of product and services
because the standardisation of interfaces drastically reduces the volume of
information required for inter-firm coordination. Since the modules them-
selves can serve many purposes, they can be produced in high volume and
combined to yield a variety of customized goods matched to differentiated
consumer demand (Langlois, 2004). In modularity literature, the implicit
assumption is that supply-chain can be ‘virtualised’. Langlois (2003) intro-
duced the concept of the “vanishing hand” to illustrate how platform technol-
ogies, such personal computers and stereo equipment systems, emerged as a
consequence of perfectly modularised product architectures under the liberal-
isation slash globalisation of trade and manufacturing in the 1980s and 1990s.
The stabilisation of technical interface standards, however, constitutes a
two-edged sword (Glimstedt, 2001; Chesbrough, 2003b; Sabel and Zeitlin,
2004). Some standardisation is obviously necessary to allow specialists to
focus upon the complex subsystems in which they have distinctive capabili-
ties. Too much standardisation, however, can just as obviously become a
barrier to systematic innovation, thus, locking component manufacturers into
a potentially obsolete product architecture. Excessive commitment to a par-
ticular product architecture and accompanying interface standards can, thus,
lead to a modularity trap, with the following two associated risks: loss of inno-
vative capacity and loss of product distinction.
Henry Chesbrough writes: “Within the firm, the focus on developing
products to compete within the standard eventually erodes the amount of
85
system-level knowledge. While focused firms are effective in linking to the

established architecture, they lack the knowledge to envision how to connect
to a new architecture. Within the industry, the collection of focused compet-
itors that modularity enthusiasts celebrate . . . now lack the collective knowl-
edge of how to evolve the system. They may also lack the ability to take
collective action, necessary to coordinate a shift from one system of highly
interconnected parts to a new system of connections” (Chesbrough, 2003b, p
181)
The far-reaching delegation of modules’ R&D diminishes organisation’s
absorptive capacities (Cohen and Levinthal, 1990), with firms losing their abil-
ity to evolve their product and to innovate (Hobday, Davies and Prencipe, 2005).
If architects of a product no longer control inter-module interactions and no
longer understand the technological and functional opportunities offered by
module-related innovations, they will be unable to design radically new archi-
tectures. The product will then freeze in its current state, which will weaken
the manufacturer who has taken the outsourcing decision, thus, affecting its
ability to make radical innovations: especially in comparison to more inte-
grated competitors. For these very reasons, firms in most industries seek to
avoid risky and irreversible commitments to a single product architecture and
technical specifications.
While having embraced the concept of modular mega suppliers in the 1990s,
major automotive OEMs backed away from modularisation as a key path for
the creation and capture of value. Automotive executives feared that modu-
larisation and outsourcing to 1st tier (mega) suppliers would lead to the hollow-
ing out of OEMs, that is shifting the capacity to innovate from OEMs to the
suppliers (MacDuffie, 2013; Jacobides, MacDuffie and Tae, 2016). Takeishi
(2002)has also shown that the quality of suppliers’ developments depends
strongly upon the degree of carmakers’ technological prowess, notably when
it regards architectural knowledge. Carmakers that have substantially
reduced the scope of their ancillary (non-core) competencies are less success-
ful in innovation terms than carmakers that have maintained and continue
to achieve significant learning regarding detailed module architecture. Man-
agers and technology strategists in the automotive sectors understand the
problem: “It is naive to believe you can integrate a system without having
in-depth and detailed knowledge of the components that are going to affect
the performance of the whole car. Managing each system performance does
86
not, in fact, automatically result in effective system integration. The perfor-

mance is the ultimate objective, not systems... We realised you cannot inte-
grate the performance of components you know very little about... if you have
never designed a component or a system it will be very difficult to understand
the subtle interactions with the rest of vehicle” (FIAT Director of Vehicle
Concept and Integration: 2006, quoted in (Becker and Zirpoli, 2011)14.
While reporting on modular networks in the electronics industry, Timothy
Sturgeon (2002) acknowledges that “as contractors seek new sources of reve-
nue by providing additional inputs to lead firm design and business processes,
and new circuit-board assembly technologies appear on the scene, such as
those for boards with optical components, the hand-off of design specifica-
tions is becoming more complex and less standardized, making it harder for
lead firms to switch and share suppliers,” while requiring “closer collabora-
tion in the realm of product design”. Thus, Sturgeon also concedes contract
manufacturing of modular design accounted in 2000 for just 13 per cent of the
global market for circuit-board and product level electronics (Sturgeon, 2002).
Concerning loss product distinction through far-reaching modularity, Joakim
Ingers, a smart phone veteran and Apple’s expert witness in Samsung versus
Apple over patent rights, noted:
“Any entrepreneur can hire a team of 10 engineers to create a new smart
phone based on standard modules, such as Qualcomm Snapdragon [hardware
platform] and Android [operating system] and a slew of other standards com-
ponents. In a year so, it would actually work. But it would be a very mediocre
device without features that make it stand out in the competition. As such, it
would compete in Asia’s rock-bottom division of mediocre white-box phones,
catching minimal attention and minimal revenues per sold phone.”
14 Statements of this kind echo the insight from research on system integration in complex systems: that
is to say, “firms need to know more than they make” (Brusoni, 2001). Task and knowledge boundaries
will not always coincide (Takeishi, 2001). Firms that have historically integrated the components of a
complex product risk a competency trap if, from outsourcing, they lose their systems integration
capability (Zirpoli and Becker, 2011). Thus, firms that no longer produce certain components may still
need to retain the knowledge of how to make them; as Brusoni et al. (2001) had it, such firms need to
“know more than they make”. Indeed, given risks of imitation from modularity (Pil and Cohen, 2006),
firms may benefit from preserving the interdependencies of a near decomposable product design— even
when more decomposition is possible— to maintain the tacit knowledge associated with managing
those interdependencies (Ethiraj and Levinthal, 2004).
87
Conclusion
There is widespread consensus among management consultants on digital
platforms with a new pillar of profitable growth or even a fourth industrial
revolution. These ideas are trickling down into business schools and into
buzzword-driven academic business research. While citing more or less fan-
ciful examples of digital platforms to illustrate the potential for innovative
value creation concerning value capture, the literature tends to be less dis-
tinct, or even conceptually misleading.
88
References
Adner, R., & Kapoor, R. (2010). Value Creation in Innovation Ecosystems: How the Structure
of Technological Interdependence Affects Firm Performance in New Technology
Generations. Strategic Management Journal, 31(3), 306–333.
Baldwin, C. Y. (2015). Bottlenecks, Modules and Dynamic Architectural Capabilities Harvard
Business School Working Paper, No. 15-028, October 2014. (Revised May 2015.).
Baldwin, C. Y., & Clark, K. B. (2000). Design Pules: The Power of Modularity (Vol. 1). Cambridge,
MA: MIT Press.
Baldwin, C. Y., & Woodard, C. J. (2008). The Architecture of Platforms: A Unified View. Working
Paper (03-030), Harvard Business School: http://www.hbs.edu/faculty/Publication%20
Files/09-034_149607b7-2b95-4316-b4b6-1df66dd34e83.pdf
Ballon, P., Walravens, N., Spedalieri, A., & Venezia, C. (2008). Towards Platform Business
Models for Mobile Network Operators. Paper presented at the 19th European Regional ITS
Conference, Rome, Italy.
Becker, M. C., & Zirpoli, F. (2003). Organizing New Product Development: Knowledge
Hollowing-out and Knowledge Integration–the Fiat Auto Case. International Journal of
Operations & Production Management, 23(9), 1033–1061.
Becker, M. C., & Zirpoli, F. (2011). Beyond Product Architecture: Division of Labour and Competence
Accumulation in Complex Product Development. SSRN Working paper, available at http://papers.
ssrn.com/sol3/papers.cfm?abstract_id=1031931:
Berger, S. (2005). How We Compete: What Companies around the World Are Doing to Make It in
Today’s Global Economy. Danvers, MA: Crown Business.
Besen, S. M., & Farrell, J. (1994). Choosing How to Compete: Strategies and Tactics in
Standardization. The Journal of Economic Perspectives, 8(2), 117–131.
Borrus, M., & Zysman, J. (1997). Globalization with Borders: The Rise of Wintelism as the
Future of Global Competition. Industry and innovation, 4(2), 141–166.
Boschma, R. A., & Lambooy, J. G. (1999). Evolutionary Economics and Economic Geography.
Journal of Evolutionary Economics, 9(4), 411–429.
Boudreau, K. (2010). Open Platform Strategies and Innovation: Granting Access Vs.
Devolving Control. Management Science, 56(10), 1849–1872.
Brusoni, S. (2005). The Limits to Specialization: Problem Solving and Coordination in
‘Modular Networks’. Organization Studies, 26(12), 1885–1907.
Brusoni, S., Prencipe, A., & Pavitt, K. (2001). Knowledge Specialization, Organizational
Coupling, and the Boundaries of the Firm: Why Do Firms Know More Than They Make?
Administrative Science Quarterly, 46(4), 597–621.
Chesbrough, H. (2003a). The Logic of Open Innovation: Managing Intellectual Property.
California Management Review, 45(3), 33–58.
Chesbrough, H. (2003b). Towards a Dynamics of Modularity: A Cyclical Model of Technical
Advance. In A. Prencipe, A. Davies, & M. Hobday (Eds.), The Business of Systems Integration.
Oxford: Oxford University Press.
89
Clements, M. T., & Ohashi, H. (2005). Indirect Network Effects and the Product Cycle: Video
Games in the Us, 1994–2002. The Journal of Industrial Economics, 53(4), 515–542.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive Capacity: A New Perspective on
Learning and Innovation. Administrative Science Quarterly, 35(1), 128–152.
Cusumano, M. (2010). Technology Strategy and Management the Evolution of Platform
Thinking. Communications of the ACM, 53(1), 32–34.
Cusumano, M. A., Mylonadis, Y., & Rosenbloom, R. S. (1992). Strategic Maneuvering and
Mass-Market Dynamics: The Triumph of Vhs over Beta. Business History Review, 66(1), 51–94.
David, P. A. (1985). Clio and the Economics of Qwerty. The American Economic Review, 75(2),
332–337.
David, P. A., & Bunn, J. A. (1988). The Economics of Gateway Technologies and Network
Evolution: Lessons from Electricity Supply History. Information Economics and Policy, 3(2), 165–202.
deJong, B., Little, M., & Gagliardi, L. (2016). Open for Business. Accenture Strategy. Retrieved
from https://www.accenture.com/t20170629T215524Z__w__/us-en/_acnmedia/PDF-56/
Accenture-Strategy-Digital-Open-Banking-POV.pdf
Doran, D. (2003). Supply Chain Implications of Modularization. International Journal of
Operations & Production Management, 23(3), 316–326.
Dougherty, D., & Dunne, D. D. (2011). Organizing Ecologies of Complex Innovation.
Organization Science, 22(5), 1214–1223.
Dranove, D., & Gandal, N. (2003). The Dvd-Vs.-Divx Standard War: Empirical Evidence of
Network Effects and Preannouncement Effects. Journal of Economics & Management Strategy,
12(3), 363–386.
Dyer, J. H., & Singh, H. (1998). The Relational View: Cooperative Strategy and Sources of
Interorganizational Competitive Advantage. Academy of Management Review, 23(4), 660–679.
Eaton, B., Elaluf-Calderwood, S., Sorensen, C., & Yoo, Y. (2015). Distributed Tuning of
Boundary Resources: The Case of Apple’s Ios Service System. Management Information
Systems Quarterly, 39(1), 217–243.
Eaton, B. D. (2012). The Dynamics of Digital Platform Innovation: Unfolding the Paradox of Control
and Generativity in Apple’s Ios. (PhD Thesis), The London School of Economics and Political
Science, London.
Economides, N., & Himmelberg, C. P. (1995). Critical Mass and Network Size with Application to the
Us Fax Market. Working paper, NYU Stern School of Business EC-95-11. Available at SSRN:
https://ssrn.com/abstract=6858 or http://dx.doi.org/10.2139/ssrn.6858:
Evans, D. S. (2003a). The Antitrust Economics of Multi-Sided Platform Markets. Yale Journal
on Regulation, 20, 325–342.
Evans, D. S. (2003b). Some Empirical Aspects of Multi-Sided Platform Industries. Review of
Network Economics, 2(3).
Evans, D. S., & Schmalensee, R. (2010). Failure to Launch: Critical Mass in Platform
Businesses. Review of Network Economics, 9(4).
Farrell, J., & Saloner, G. (1985). Economic Issues in Standardization.
Galunic, D. C., & Eisenhardt, K. M. (2001). Architectural Innovation and Modular Corporate
Forms. Academy of Management Journal, 44(6), 1229–1249.
90
Gandal, N., Kende, M., & Rob, R. (2000). The Dynamics of Technological Adoption in
Hardware/Software Systems: The Case of Compact Disc Players. The RAND Journal of
Economics, 43–61.
Gandal, N., Salant, D., & Waverman, L. (2003). Standards in Wireless Telephone Networks.
Telecommunications Policy, 27(5), 325–332.
Garrard, G. A. (1998). Cellular Communications : Worldwide Market Development. Boston, MA:
Artech House.
Garud, R., Jain, S., & Kumaraswamy, A. (2002). Institutional Entrepreneurship in the
Sponsorship of Common Technological Standards: The Case of Sun Microsystems and
Java. Academy of Management Journal, 45(1), 196–214.
Gawer, A., & Cusumano, M. A. (2002). Platform Leadership: How Intel, Microsoft, and Cisco Drive
Industry Innovation. Boston: Harvard Business School Press.
Gawer, A., & Cusumano, M. A. (2008). How Companies Become Platform Leaders. MIT Sloan
Management Review, 49(2), 28.
Gawer, A., & Cusumano, M. A. (2014). Industry Platforms and Ecosystem Innovation. Journal
of Product Innovation Management, 31(3), 417–433.
Ghazawneh, A., & Henfridsson, O. (2013). Balancing Platform Control and External
Contribution in Third-Party Development: The Boundary Resources Model. Information
Systems Journal, 23(2), 173–192.
Gilder, G. (1993). Metcalfe’s Law and Legacy. Discovery Institute, Technology and Democracy
Project. Availible at: http://www.discovery.org/a/41.
Gilson, R. J., Sabel, C. F., & Scott, R. E. (2009). Contracting for Innovation: Vertical
Disintegration and Interfirm Collaboration. Columbia Law Review, 109(3), 431–502.
Glimstedt, H. (2001). Competitive Dynamics of Technological Standardization: The Case of
Third Generation Cellular Communications. Industry and innovation, 8(1), 49–78.
Goldenberg, J., Libai, B., & Muller, E. (2010). The Chilling Effects of Network Externalities.
International Journal of Research in Marketing, 27(1), 4–15.
Hagiu, A., & Hałaburda, H. (2014). Information and Two-Sided Platform Profits. International
Journal of Industrial Organization, 34(May 2014), 25–35.
Helper, S., MacDuffie, J. P., & Sabel, C. (2000). Pragmatic Collaborations: Advancing
Knowledge While Controlling Opportunism. Industrial and Corporate Change, 9(3), 443–488.
Helper, S. R., & Sako, M. (1995). Supplier Relations in Japan and the United States: Are They
Converging? Sloan Management Review, 36(3), 77.
Henderson, R. M., & Clark, K. B. (1990). Architectural Innovation: The Reconfiguration of
Existing Product Technologies and the Failure of Established Firms. Administrative Science
Quarterly, 35(1), 9–30.
Hobday, M., Davies, A., & Prencipe, A. (2005). Systems Integration: A Core Capability of the
Modern Corporation. Industrial and Corporate Change, 14(6), 1109–1143.
Huang, G. Q., Zhang, X., & Liang, L. (2005). Towards Integrated Optimal Configuration of
Platform Products, Manufacturing Processes, and Supply Chains. Journal of Operations
Management, 23(3), 267–290.
91
Huang, P., Ceccagnoli, M., Forman, C., & Wu, D. (2013). Appropriability Mechanisms and the
Platform Partnership Decision: Evidence from Enterprise Software. Management Science,
59(1), 102–121.
Iansiti, M., & Levien, R. (2002). The New Operational Dynamics of Business Ecosystems:
Implications for Policy, Operations and Technology Strategy Working Paper 03-030:
Division of Research, Harvard Business School.
Iansiti, M., & Levien, R. (2004a). The Keystone Advantage: What the New Dynamics of Business
Ecosystems Mean for Strategy, Innovation, and Sustainability. Boston, MA: Harvard Business Press.
Iansiti, M., & Levien, R. (2004b). Strategy as Ecology. Harvard Business Review, 82(3), 68–81.
Jacobides, M. G., Knudsen, T., & Augier, M. (2006). Benefiting from Innovation: Value Creation,
Value Appropriation and the Role of Industry Architectures. Research Policy, 35(8), 1200–1221.
Jacobides, M. G., MacDuffie, J. P., & Tae, C. J. (2016). Agency, Structure, and the Dominance
of Oems: Change and Stability in the Automotive Sector. Strategic Management Journal,
37(9), 1942–1967.
Jacobides, M. G., & Winter, S. G. (2005). The Co-Evolution of Capabilities and Transaction
Costs: Explaining the Institutional Structure of Production. Strategic Management Journal,
26(5), 395–413.
Jarillo, J. C. (1988). On Strategic Networks. Strategic Management Journal, 9(1), 31–41.
Katz, M. L., & Shapiro, C. (1985). Network Externalities, Competition, and Compatibility.
The American Eeconomic Review, 75(3), 424–440.
Kenney, M., & Zysman, J. (2016). The Rise of the Platform Economy. Issues in Science and
Technology, 32(3), 61.
Lacy, P., Hagenmueller, M., & Ising, J. (2016). Platform Strategies: How the Rules of
Competitiveness Have Changed in the Era of Ecosystems. Accenture Strategy, availible at
https://www.accenture.com/t20160916T010708Z__w__/us-en/_acnmedia/PDF-27/
Accenture-Strategy-Platform-Strategies.pdf:
Langlois, R. N., & Robertson, P. L. (1992). Networks and Innovation in a Modular System:
Lessons from the Microcomputer and Stereo Component Industries. Research Policy, 21(4),
297–313.
Lazonick, W. (2005). Evolution of the New Economy Business Model. Business History and
Economic History OnLine(2005:3), 1–60.
Lazonick, W. (2009). Sustainable Prosperity in the New Economy? Kalamazoo, Michigan: WE
Upjohn Institute for Employment Research
Li, Y.-R. (2009). The Technological Roadmap of Cisco’s Business Ecosystem. Technovation,
29(5), 379–386.
Louis, T. (2013, Aug 10). How Much Do Average Apps Make? Forbes.
MacDuffie, J. P. (2013). Modularity-as-Property, Modularization-as-Process, and ‘Modularity-
’-as-Frame: Lessons from Product Architecture Initiatives in the Global Automotive
Industry. Global Strategy Journal, 3(1), 8–40.
Metcalfe, B. (2013). Metcalfe’s Law after 40 Years of Ethernet. Computer, 46(12), 26–31.
92
Molina, A., Bremer, C. F., & Eversheim, W. (2001). Achieving Critical Mass: A Global
Research Network in Systems Engineering. Foresight, 3(1), 59–65.
Nambisan, S., & Sawhney, M. (2011). Orchestration Processes in Network-Centric Innovation:
Evidence from the Field. The Academy of Management Perspectives, 25(3), 40–57.
Normann, R., & Ramirez, R. (1993). From Value Chain to Value Constellation: Designing
Interactive Strategy. Harvard Business Review, 71(4), 65–77.
Oren, S. S., & Smith, S. A. (1981). Critical Mass and Tariff Structure in Electronic
Communications Markets. The Bell Journal of Economics, 467–487.
Park, Y., Ogawa, K., Tatsumoto, H., & Hong, P. (2009). The Impact of Product Architecture
on Supply Chain Integration: A Case Study of Nokia and Texas Instruments. International
Journal of Services and Operations Management, 5(6), 787–798.
Parker, G., & Alstyne, M. V. (2008). Managing Platform Ecosystems. ICIS 2008 Proceedings, 53.
Parnas, D. L. (1972). On the Criteria to Be Used in Decomposing Systems into Modules.
Communications of the ACM, 15(12), 1053–1058.
Peres, R., Muller, E., & Mahajan, V. (2010). Innovation Diffusion and New Product Growth
Models: A Critical Review and Research Directions. International Journal of Research in
Marketing, 27(2), 91–106.
Perez, S. (2016). Apple’s App Store Hits 2m Apps, 130b Downloads, $50b Paid to Developers.
TechCrunch ( June 23, 2016). Available at: https://techcrunch.com/2016/06/13/apples-app-
store-hits-2m-apps-130b-downloads-50b-paid-to-developers/ https://techcrunch.
com/2016/06/13/apples-app-store-hits-2m-apps-130b-downloads-50b-paid-to-developers/
Piore, M., & Sabel, C. (1984). The Second Industrial Divide: Possibilities for Prosperity. NYC, New
York: Basic Books.
Pon, B., Seppälä, T., & Kenney, M. (2014). Android and the Demise of Operating
System-Based Power: Firm Strategy and Platform Control in the Post-Pc World.
Telecommunications Policy, 38(11), 979–991.
Porter, M. E. (1985). Competetive Advantage: Creating and Sustaining Superior Performance.
New York: Free Press.
Prasarnphanich, P., & Wagner, C. (2011). Explaining the Sustainability of Digital Ecosystems
Based on the Wiki Model through Critical-Mass Theory. IEEE Transactions on Industrial
Electronics, 58(6), 2065–2072.
Reisinger, D. (2017). Here’s How Many Iphones Are Currently Being Used Worldwide. Fortune.
Rochet, J. C., & Tirole, J. (2003). Platform Competition in Two-Sided Markets. Journal of the
European Economic Association, 1(4), 990–1029.
Rochet, J. C., & Tirole, J. (2006). Two-Sided Markets: A Progress Report. The RAND Journal
of Economics, 37(3), 645–667.
Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). New York: Free Press.
Rohlfs, J. (1974). A Theory of Interdependent Demand for a Communications Service.
The Bell Journal of Economics and Management Science, 16–37.
Rohlfs, J. H. (2003). Bandwagon Effects in High-Technology Industries. Cambridge, MA: MIT Press.
Roson, R. (2005). Two-Sided Markets: A Tentative Survey. Review of Network Economics, 4(2).
93
Roy, R., & McEvily, S. K. (2004). Incumbent Survival During Market Fusion in Matured
Industries: The Influence of Component and Architectural Capabilities on the Survival of
Us Machine Tool Manufacturers During 1975–1995. In J. A. C. Baum & A. M. McGahan
(Eds.), Business Strategy over the Industry Lifecycle (Advances in Strategic Management, Volume 21)
(pp. 199–224): Emerald Group Publishing.
Sabel, C. F. (1994). Learning by Monitoring. In N. J. Smelser & R. Swedberg (Eds.),
The Handbook of Economic Sociology. Princeton, N J Princeton University Press.
Sabel, C. F., & Zeitlin, J. (2004). Neither Modularity nor Relational Contracting: Inter-Firm
Collaboration in the New Economy. Enterprise & Society, 5(3), 388–403.
Sako, M. (2009). Technology Strategy and Management Globalization of Knowledge-
Intensive Professional Services. Communications of the ACM, 52(7), 31–33.
Shapiro, C., & Varian, H. R. (1998). Information Ules: A Strategic Guide to the Network Economy.
Boston, MA: Harvard Business Publishing.
Simon, H. A. (1965). The Architecture of Complexity. General systems, 10(1965), 63–76.
Srinivasan, R., Lilien, G. L., & Rangaswamy, A. (2004). First in, First Out? The Effects of
Network Externalities on Pioneer Survival. Journal of marketing, 68(1), 41–58.
Sturgeon, T. J. (2002). Modular Production Networks: A New American Model of Industrial
Organization. Industrial and corporate change, 11(3), 451–496.
Takeishi, A. (2002). Knowledge Partitioning in the Interfirm Division of Labor: The Case of
Automotive Product Development. Organization Science, 13(3), 321–338.
Tee, R., & Gawer, A. (2009). Industry Architecture as a Determinant of Successful Platform
Strategies: A Case Study of the I-Mode Mobile Internet Service. European Management
Review, 6(4), 217–232.
Thomas, L. D., Autio, E., & Gann, D. M. (2014). Architectural Leverage: Putting Platforms
in Context. The Academy of Management Perspectives, 28(2), 198–219.
Tilson, D., Sorensen, C., & Lyytinen, K. (2012). Change and Control Paradoxes in Mobile
Infrastructure Innovation: The Android and Ios Mobile Operating Systems Cases. Paper presented at
the 2012 45th Hawaii International Conference on System Science.
Tiwana, A., Konsynski, B., & Bush, A. A. (2010). Research Commentary—Platform Evolution:
Coevolution of Platform Architecture, Governance, and Environmental Dynamics.
Information Systems Research, 21(4), 675–687.
Vargo, S. L., & Lusch, R. F. (2004). Evolving to a New Dominant Logic for Marketing.
Journal of marketing, 68(1), 1–17.
West, J., & Gallagher, S. (2006). Challenges of Open Innovation: The Paradox of Firm
Investment in Open-Source Software. R&d Management, 36(3), 319–331.
Wheelwright, S. C., & Clark, K. B. (1992). Revolutionizing Product Development: Quantum Leaps
in Speed, Efficiency, and Quality. NYC, New York: Simon and Schuster.
Wintermeyer, L. (2017, Feb 17). Global Fintech Vc Investment Soars in 2016. Forbes Magazine.
Von Burg, U. (2001). The Triumph of Ethernet: Technological Communities and the Battle for the Lan
Standard. Palo Alto, CA: Stanford University Press.
Zirpoli, F., & Caputo, M. (2002). The Nature of Buyer-Supplier Relationships in Co-Design
Activities: The Italian Auto Industry Case. International Journal of Operations & Production
Management, 22(12), 1389–1410.
94
The Digital Customer
CHAP TER 4
Catering to the Digital Consumer:

From Multichannel to
Omnichannel Retailing
SAR A ROSENGREN, FREDRIK L ANGE, MIK AEL HERNANT, AND ANGELICA BLOM
Introduction
Digitalization has a long history in retailing. In fact, is has been a driving
force since the 1960s with developments being made, for example, in terms of
electronic cash systems, barcodes, point-of-sale data, and electronic data inter-
change (EDI) with suppliers (e.g., Hagberg et al, 2017). Since the early 2000s,
however, digitalization driven by operations and supply-side concerns has
been complemented by digitalization following changing consumer beha-
viours and demands. The way consumers search and shop for products has
changed dramatically during this period and, consequently, retail offers have
changed with it. Although this change has affected retailing in general, it has
been especially disruptive in service retailing (such as travel and banking)
and for retailers focusing upon products that have been digitized: such as, for
instance, music and books (Verhoef et al, 2015). The intangibility of these
offers has made them highly suitable for online transactions that, in turn,
make physical stores more or less obsolete.
Still, the shift in consumer behaviour is increasingly being felt in other
retail sectors. A clear indication is the many recent store closings and bank-
ruptcies of large retailers in US. As is the fact that online retailing comprised
all growth in Swedish non-grocery retailing in the first half of 2017. With
increasing use of mobile, tablets, social media and technological advances in
analytics, the integration of channels in online and offline retailing continues
to develop; as a result, the retail landscape continues to change. As a conse-
quence, it has been argued that digitization of consumer behaviours will
97
SA R A ROSENGR EN, F R EDR I K LA NGE , M I K A EL H ER NA N T, A N D A NGELICA BLOM
fundamentally change retail as we know it (Grewal et al, 2017; Verhoef et al,

2015).
This chapter discusses contemporary research on digital shopper beha-
viour and its implications for retailers to better understand this development;
it does so by reporting on two ongoing research projects at the Center for
Retailing at Stockholm School of Economics: more detailed descriptions can
be found in Hernant and Rosengren (2017) and Blom et al (2017). While pre-
vious research has primarily focused upon the growth of online retailing
(Hagberg et al 2017) we focus upon the integration between online and offline
retailing, which is often referred to as omnichannel retailing (Verhoef et al
2015). The question raised in this chapter is not whether an omnichannel
strategy should be implemented; rather, what should be the omnichannel
strategy. We particularly focus upon the importance of understanding, man-
aging, and evaluating consumer behaviour across channels and touchpoints
when crafting an omnichannel retail strategy.
In our first project, we use customer-level data of actual customer purchases
offline and online to track changes in consumer behaviour over time (Hernant
and Rosengren, 2017). The data comes from three Swedish retailers that operate
in three different retail industries: pharmacy, grocery, and beauty. This project
contributes to omnichannel retailing by offering a way to understand changes
in customer behaviours over time. In the second project (Blom et al. 2017), we
use experimental methods to test the effectiveness of efforts to integrate consum-
ers’ digital traces (for example, online browsing history) with promotional offers
from a retailer. This project contributes to omnichannel research by examining
digitalization related to individual shopping trips and shopper reactions to
omnichannel-based promotions that are created for short-term sales effects. In
both projects, we adopt the perspective of an established retailer adapting their
offer to increasingly digital shopper behaviours. Still, it should be noted that the
challenges in combining the online and offline retail offer is equally pressing for
pure online players trying to grow, as indicated, for instance, by more and more
online retailers opening physical stores (for example, Amazon buying Whole
Foods or Zalando adding outlets and pop-up stores to their online offers).
From Multichannel Retailing to Omnichannel Retailing

Although interest in digital consumers dates back to the 1990s, it is in the past
decade or so that digital consumer behaviour has become mainstream.
98
CAT ER I NG TO T H E DIGI TA L CONSUMER
According to eMarketer (n.d.), a leading source of statistics on digital retail,

over 2.14 billion people worldwide are expected to buy goods and services
online in 2021, which is up from 1.66 billion global digital buyers in 2016. In
2016, 58.3 percent of global internet users had purchased products online. By
2019, this figure is expected to grow to 63 percent. This development clearly
shows that retailers need to cater to consumers who are constantly online via
smart phones and tables and for whom social and digital media are an ever-
present part of everyday life and, thus, the buying process.
Verhoef et al (2015) discuss in their seminal article how digital consumer
behaviour has led retailers to move from multichannel retailing: that is to say,
offering their products in several different online and/or offline channels, to
omnichannel retailing: offering customer experiences that are integrated
across different channels and touchpoints. Whereas multichannel retail man-
agement focuses upon “the design, deployment, coordination, and evaluation
of channels to enhance customer value through effective customer acquisi-
tion, retention, and development”, they argue that omnichannel retail man-
agement should be concerned with “the synergetic management of the numer-
ous available channels and customer touchpoints, in such a way that the
customer experience across channels and the performance over channels is
optimised” (p. 175). The main differences in perspectives lie in the focus upon
synergies rather than coordination, on holistic customer experiences rather
than conversion (customer acquisition, retention, and development), and on
combining sales (channel) and brand (touchpoint) perspectives. Table 4.1
provides an overview of the different perspectives.
Table 4.1: Differences Between Multichannel and Omnichannel Management in Retailing

(Adapted from Verhoef et al, 2015)
Multichannel Omnichannel
Focal channels Interactive (transactional) Interactive (transactional) and
mass communication channels
Channel integration Channels separate Channel synergies
Focal interactions Customer, store Customer, store, brand
Management Per channel Cross-channel
Performance measures (KPI) Sales Sales and brand equity
99
The literature on multichannel retailing is vast; however, Verhoef et al

(2015) points to the need for more research on omnichannel issues Such
research has, indeed, emerged over the past couple of years as illustrated, for
instance, in special issues coming out in several of the leading academic retail
journals dealing with omnichannel issues (for example, Grewal et al, 2017;
Hagberg et al, 2017).
From a retailer perspective, the move to omnichannel retailing seems inev-
itable. Consumers are digital shoppers and, thus, retailers need to offer expe-
riences that cater to their demands. Still, there is plenty of room for variations,
in terms of what the omnihchannel strategy should be. We will now discuss
three key issues to consider in developing this strategy: namely understand-
ing, managing, and evaluating movements between channels and touchpoints.
UNDERSTANDING MOVEMENTS BET WEEN CHANNELS AND TOUCHPOINTS

Previously, researchers discussed channel migration, as consumers were
thought to switch from offline stores to online stores. Nowadays, there is a
general understanding that most consumers move back and forth between
channels and use them interchangeably. There is a vast literature document-
ing different tools for understanding consumer behaviours and decision
journeys in order to create strong customer experiences (for a review: see
Lemon and Verhoef, 2016). Customer experience is a multi-dimensional con-
struct focusing upon cognitive, emotional, behavioural, sensorial, and social
responses to a firm’s offerings during the entire purchase journey (Lemon
and Verhoef, 2016: p. 71). This means the overall customer experience is cre-
ated in interactions before, during, and after a specific purchase has been
made. The interactions are not all controlled by the retailer; they also com-
prise touchpoints created by brands and partners as well as other consumers,
influencers, and media (see Figure 4.1 below).
Pre-purchase customer experiences are created through interactions taking place
as the consumer recognises a need and begins to search for ways to fulfil it.
In order to serve these effectively, retailers should consider different types of
shopping needs. For example, it is often fruitful to distinguish between utili-
tarian (also refered to as task oriented or problem solving), and hedonic (also
refered to as recreational, or experiential) shopping needs, as they lead to
different decision journeys. From a retailer perspective, it is also vital to
understand where the need occurs and whether it will lead consumers to
100
search for information. In this regard, a distinction can be made between

planned (the need occurs prior to purchase and typically leads to information
search) or unplanned (the need occurs at the point of purchase without much
previous information search).
Considering the interplay between online and offline channels in the pre-
purchase stage in an omnichannel context is especially important. Key
behaviours to consider are the extent to which consumers use webrooming
(that is to say, searching online and then moving in to physical stores to make
a purchase) and showrooming (searching offline and then moving to digital
stores to make a purchase).
Purchase customer experiences cover all interactions with the brand and its envi-
ronment during the actual purchase. From a retailer perspective, key beha-
viours to consider at this stage are how choice, ordering, and payment are
facilitated. Shopping experiences can also be created during the purchase stage.
Traditionally, the focus in retailing has been on the offline store as the
arena for purchase (sales), whereas online platforms have been used as touch-
points to convey messages and provide information (brand). However, this
logic can easily be turned around today: as indicated by novel showrooming
concepts, such as the Samsung concept store in NYC (where consumers are
not able to leave the physical store with a product) and American department
store retailer Nordstrom’s new concept: small neighbourhood stores without
inventory. In addition, all retailers need to be sensitive to the fact that shop-
pers may look for the same product in competing online stores while shopping
in a specific online or offline store.
Post-purchase consumer experiences comprise interactions that consumers have
with the brand and its environment following the actual purchase. These
interactions are usually related to usage and consumption; important
behaviours from a retailer perspective deal with service and loyalty. Thus, the
focus in this step has been on issues related to the brand rather than to sales.
In omnichannel retailing, however, this stage of the decision journey is
more complex. Whereas the shopping experience in traditional physical
retailing typically stopped with the purchase, the purchase and delivery of
products is typically extended in time when an online purchase is made. This
means that omnichannel retailers need to be more concerned about the expe-
rience offered in delivering the product to customers. This has led to an
increased focus upon different delivery options: such as click-and-collect
101
(where products purchased online are picked up in offline stores) and last-mile
delivery as part of the customer decision journey.
Understanding how consumers make decisions, mapping the use of differ-
ent touchpoints in different stages, and identifying trigger points that lead
customers to continue or discontinue in their purchase journey is crucial for
understanding how to manage an omnichannel retail offer. Shoppers today
may use a retailer’s digital channels and touchpoints on the path to purchase
and, thereby, disclose information to the retailer about, for instance, the types
of products in which they are interested. This behaviour produces digital
trace (for example, from website browsing with the mobile phone) about shop-
ping goals that may be a valuable source of information for retailers that they
can use in order to analyse movements between channels and touchpoints,
which they are able to manage them over time.
Process Model for Customer Journey and Experience
Current Customer Experience (t)

Feedback
Previous Experience Future Experience

(t-n) (t+n)
Postpurchase Stage
Prepurchase Stage
Purchase Stage
Prepurchase Stage
Postpurchase Stage
Prepurchase Stage
Postpurchase Stage
Purchase Stage
Purchase Stage
Behaviors: Behaviors: Behaviors:

need, recognition, choice, ordering, consumption, usage,
consideration, payment engagement, service
search requests
Customer Journey
Figure 4.1. Process model for customer journeys (adapted from Lemon and Verhoef, 2016)
M ANAGING MOVEMENTS BET WEEN CHANNELS AND TOUCHPOINTS

Based upon thorough analyses of the consumer decision journey, retailers are
better equipped to come up with strategies on how to manage movements
102
between channels and touchpoints. A review of existing literature on omni-

channel retailing highlights that integration between channels and touchpoints
needs to be managed differently depending upon the type of product being
offered and the type of decision journey used by consumers (Blom et al 2017).
We investigated in a recent research project how consumers respond when
retailers offer a promotion in the physical store that is based upon pre-purchase
digital data. We compared such tailored promotions (that is to say, promotions
that are adapted based upon the customer’s pre-purchase behaviour online
before entering the store) with a generic promotion (see Blom et al, 2017 for a
more detailed description). The results reveal that shoppers are more respon-
sive to tailored promotions compared to generic traditional promotions. We
also found the advantages of using tailored promotions vary depending upon
the type of product (utilitarian versus hedonic motivation) and decision pro-
cess (planned versus unplanned) of consumers.
In terms of product categories, our study found that, although tailored
promotions have a positive effect on both sales and brand, this effect is atten-
uated for hedonic compared to utilitarian product categories. Since a tailored
promotion evokes benefits that are more valued by those shopping for utili-
tarian rather than hedonic products, consumers who are shopping for a utili-
tarian product category redeem the promotion to a greater extent and have a
more positive attitude towards the retailer after being offered such promo-
tions. The tailored promotion facilitates the shopping trip, thus, enabling the
consumers to satisfy their shopping goal in an effortless manner. Conversely,
consumers seek exploration, variation, and enjoyment when shopping for
hedonic product categories. In order to be relevant in promotions targeting
such consumers, promotions that satisfy their shopping goal in an experien-
tial fashion is required, as the shopping experience is part of the shopping
goal. The tailored promotion is, therefore, of less relative importance since
consumers shopping for hedonic products also perceive the shopping trip
experience as such as part of their shopping goal.
In terms of consumer decision processes, our study found that, although
tailored promotions have a positive effect on both sales and brand, the effect is
attenuated when consumers make an unplanned versus a planned purchase.
Consumers that have made a planned purchase have previously identified a
problem. This typically means they engage in a more rational shopping pro-
cess with clearly defined shopping goals. In contrast, unplanned purchases
103
correspond to more irrational, emotional, affective, hedonic, experiential, and

variety seeking aspects of shopping. An unplanned purchase occurs without
much thought or consideration and is less rational, thus, meaning that it is
made with relatively little cognitive control and is usually linked to abstract
shopping goals. A tailored promotion is more helpful for consumers that have
planned their purchase. Since consumers that have made a planned purchase
have engaged in cognitive effort, thought and involvement, and have formed
a concrete shopping goal, a promotion that is congruent with this goal is more
valuable. As a consequence, consumers who make a planned purchase are
more likely to redeem the promotion and become more positive toward the
retailer.
Overall, the results from the Blom et al. (2017) study indicate that examin-
ing different types of shopping situations is important to better understand
omnichannel promotional initiatives and how to manage movements between
channels and touchpoints in an omnichannel setting. We also found that
integration is less important in hedonic (versus utilitarian) product categories
and for unplanned (versus planned) purchases.
We also explored the benefits to consumers that drive these effects. The
use of digital traces in promotions may produce both utilitarian and hedonic
benefits and value for shoppers; these benefits and values are linked in differ-
ent ways to purchase behaviour and brand perceptions. The results reveal
that integration is more important for omnichannel performance (sales) than
for omnichannel experience (brand). If retailers are interested in short-term
effects on sales (promotion redemption), they should strive for designing
omnichannel promotions that provide benefits and shopping value, which are
primarily utilitarian. However, a retailer should offer both utilitarian and
hedonic benefits and experiential values if the main purpose of omnichannel
promotions is to create long-term effects on perceived omnichannel experi-
ence (that is to say, brand building).
The results from our study suggest that omnichannel promotions are more
effective if they are designed with in-depth knowledge about the shopping
goals of specific shopping trips. Based upon an analysis of the customer jour-
ney for different customers it is, thus, possible to come up with more
tailor-made strategies to manage customer movements between channels and
touchpoints. This implication is not restricted to promotions; rather, it could
104
be used to guide other decisions related to managing consumer movements

between channels and touchpoints. For example, the other study discussed in
this chapter (Hernant and Rosengren, 2017) found an asymmetric effect of
converting customers from online to offline compared to converting them
from offline to online, in the context of pharmacy retailing (which can be
characterised as driven by utilitarian products and planned consumer
behaviours). Whereas online customers buy more as they convert offline,
existing offline customers mainly switch channels. This implies that a retailer
should target customers recruited online with offers that will help converge
them offline; while it should market its online channel to current customers
more as an information channel or a complement to the existing store. As
illustrated thus far, there is a lot to gain by analysing and managing move-
ments between channels and touchpoints. These activities also need to be
matched with a clear understanding of how to evaluate these movements.
E VALUATING MOVEMENTS BET WEEN CHANNELS AND TOUCHPOINTS

Retailers must also come up with a system for evaluating movements between
channels and touchpoints in order to fully understand omnichannel retailing.
Such measures play a key role in making insights actionable for the firm (for
example, Lemon and Verhoef, 2016). The omnichannel literature points to
the importance of integrating customer experiences, sales, and brand perspec-
tives; thus, these three aspects must be included when evaluating movements
between channels and touchpoints (Verhoef et al, 2015).
In practice, it is difficult to find a single set of measures that adequately
captures customer experience across different retail sectors and different
channels. This means that firms typically rely upon a set of different mea-
sures (such as satisfaction, net promoter score, or similar) for different parts
of the omnichannel offer. There ideally should be proven measurement
approaches for the overall customer experience at each stage in the customer
journey (pre-purchase, purchase, and post-purchase) and for all touchpoints.
This measure should then be complemented with measures of sales and
brand perceptions. However, there is not yet agreement on robust measure-
ment approaches to evaluate all aspects of customer experience across the
customer journey. Lemon and Verhoef (2016, p. 81–82) conclude the following
in their review of customer experience research:
105
ÜLong-tested approaches, such as SERVQUAL (reliability, assurance,

tangibles, empathy, and responsiveness) may offer a good starting point.
ÜCustomer satisfaction and NPS (net promoter score) perform equally
well in predicting firm performance and customer behaviour, although
the predictive performance differs between specific contexts.
ÜTransformations of metrics are useful to account for potential non-
linear effects due to notions such as customer delight.
ÜCustomer feedback metrics focusing upon a specific domain of the
customer are not strong in predicting future performance.
ÜMultiple customer feedback metrics predict customer behaviour better
than a single metric.
Lemon and Verhoef (2016) recommend crafting dashboards that combine

different measures to evaluate customer experience across channels and
touchpoints, based upon this review. Such dashboards also need to consider
retailer performance, in terms of both sales and brand, to provide a compre-
hensive evaluation of omnichannel performance. For example, as a way to
facilitate integration between online and offline sales performance, Ikea and
Åhléns attribute additional sales in the online channel to the offline stores in
geographical areas where the product is delivered. This way, store employees
can focus upon customer experience and satisfaction rather than direct con-
version in stores. Similarly, it is important to examine simultaneously the
effects on both purchase behaviour (sales) and brand image (see Blom et al.,
2017) when evaluating omnichannel promotions.
From a customer journey perspective, it is also important to track how
multiple touchpoints encountered throughout the journey affect purchase (or
other behavioral) outcomes. This is often done through attribution or path-to-
purchase models. Although such analyses can be valuable, a general problem
with modelling such behaviours is that touchpoints tend to be used in differ-
ent phases of the decision journey and, as a consequence, touchpoint effects
can be endogenous, thus, leading to erroneous conclusions and resource
allocation (Lemon and Verhoef, 2016). What is more, tools are available in an
online context for tracking shopper behaviour in real time. Offline, however,
these tools typically comprise satisfaction, and sales and are not measured as
frequently as are online measures. As a consequence, there is a risk that too
106
much weight is given to aspects of customer behaviours that are easily mea-
sured in relation to what should be measured and what really drives customer
experience, sales, and brand effects.
A Framework for Analysing

Buying Behaviour Across Channels
Although the overall customer experience is multi-dimensional, retailers
typically have quite elaborate and readily available data in terms of
behaviours in their CRM (customer relationship management) systems or
loyalty programs. This data can be valuable in developing an understanding
of consumer behaviour in different channels, thus, providing a point of depar-
ture for analysing, managing, and evaluating behaviours across channels. In
our second recent project discussed here, we use this type of data to develop
a framework that provides a comprehensive understanding of the implica-
tions of retail digitalization for sales in physical stores and of how to manage
customers across channels (for details, see Hernant and Rosengren, 2017).
Our framework illuminates how customer behaviours of individual custom-
ers change when a new online channel is added to an existing network of
physical stores. Thus, our focus is upon the purchase step of the consumer
decision journey and the extent to which buying behaviours of individual
customers change when they also start using an online channel.
To analyse movements between channels for individual customers, our
framework disentangles overall buying behaviors of each customers into
amount, frequency, and regularity.
Table 4.2: Definitions of Variables (from Hernant and Rosengren, 2017)
Purchase amount Total amount of all purchases the customer has made.
Purchase volume Total number of all products the customer has purchased.
Purchase visits Number of occasions the customer has made purchases.
Purchase months Number of different months for the customer’s purchase visits.
Customer months Total number of months the customer has been a customer.
Monthly amount Purchase amount per customer month, (i.e. the total amount of purchases
divided by the number of months since the first recorded visit in the database).
Monthly volume Purchase volume per customer month, (i.e. the total volume of purchased products
divided by the number of months since the first recorded visit in the database).
Amount per visit Purchase amount per purchase visit (i.e. the total amount of purchases
divided by the number of purchase visits).
107
The variables used in our study are summarised in Table 4.2. All measures
refer to the aggregate behaviour during the time period for which the cus-
tomer has been a customer, and were used to describe overall behaviour as
well as behaviour for the online and offline channel, respectively (for further
details regarding the variables: please see Hernant and Rosengren, 2017).
The variables facilitate a comparison of customers purchase behaviour (in
totalm, as well as online and offline) after they have become online custom-
ers. The monthly amount (volume) can be calculated by multiplying the
amount (volume) per transaction, the purchase frequency, and the purchase
regularity. Mathematically, the monthly amount (volume) can be expressed
by multiplying the amount (volume) per transaction, the purchase frequency,
and the purchase regularity (see Figure 4.2).
Purchase amount Purchase amount Purchase visits Purchase months

=
Customer months Purchase visits Purchase months Customer months
(Monthly amount) (Amount per visit) (Purchase frequency) (Purchase regularity)
Purchase volume Purchase volume Purchase visits Purchase months

=
Customer months Purchase visits Purchase months Customer months
(Monthly volume) (Volume per visit) (Purchase frequency) (Purchase regularity)
Figure 4.2: Calculations of monthly revenues and sales volumes using the variables
(from Hernant and Rosengren, 2017)
UNDERSTANDING MOVEMENTS IN PHAR M ACY RETAILING

Our framework was developed in the context of pharmacy retailing, using the
CRM system of a Swedish pharmacy retailer, and capturing data for a total
of 12,139 online customers. Each sales receipt shows detailed information on
an actual purchase: that is to say, the (anonymised) customer’s identity, the
products purchased, and the date and time of purchase. Furthermore, all
purchases in the retailer’s physical stores in a time period from one year
before the online store opened were identified for all customers who made at
least one purchase in the online store. In all, this dataset enabled us to analyse
customer purchase behaviour for a full year prior to, and 15 months post-
introduction of the online store at the specific retailer (for a more detailed
description: see Hernant and Rosengren, 2017).
108
The analysis using the framework showed that going online facilitated
customer acquisition. A total of 2,954 (24%) of the 12,193 customers who
patronised the online store made their first purchase online. Most of these
(1,798 or 61%) made no purchases offline. The remaining 1,156 (39%) online
customers also became offline customers. The latter customer group provided
twice as much sales every month as online-only customers; this was primarily
due to their higher purchasing regularity. From the retailer’s perceptive, this
difference clearly indicates that it is crucial to convert customers acquired
online to also become customers of the physical store.
We also found that existing customers buying online (9,239 or 76%) were
primarily shifting their purchases from one channel to the other. These cus-
tomers showed only minor increases in total purchasing after their first online
purchase; identical number of products purchased per month, and monthly
amount increased by 2%. However, the average transaction size among these
customers increased (most notably in amount spent per visit, which rose by
13%) after their first online purchase. The purchase frequency of this customer
group increased by 2.4%; purchase regularity decreased by 5%. The net effect,
therefore, was the larger average transaction translated into only a small
increase in monthly amount. Overall, the online buying by existing (offline)
customers represented a switch from offline to online. After their first online
purchase, existing customers changed their offline behaviour, as illustrated by
lower average transactions, lower frequency, and lower regularity.
Based on these analyses we found an asymmetry in channel migration.
Converting online customers into customers at a physical store provided
more opportunities for additional sales than the other way around. The
results also highlighted the importance of looking beyond average transaction
size (see also Pauwels and Neslin, 2015). By using the average monthly sales
from individual consumers as the target variable, our study found that,
despite the larger online transaction size, the interaction between average
transaction, purchase frequency, and regularity for existing customers is a
zero-sum game. The effects on the physical stores for the retailer were nega-
tive for all variables.
This highlingts how retailers can use our framework to design address var-
ious aspects of consumer behaviour (average transaction, frequency, and regu-
larity), as well as describe the offers and their underlying strategies to decision
makers and store employees. The framework (Hernant and Rosengren, 2017)
109
allows for a more insightful discussion of the sales impact of adding an online
channel. It can also serve as a basis for targeting different customer groups with
tailored offers designed to change the facet of customer behaviour (average
transaction size, frequency, or regularity) considered most likely to impact
overall sales among that specific group of customers.
While access to this type of customer-level data is rare in academic
research, it is typically readily available for retailers, which makes the frame-
work easy to implement. The framework can also be used for continuous
tracking and aggregation of individual customers’ purchase behaviour
between channels, and assist retailers in identifying sales-related problems as
well as potential opportunities. For the retailer in our study, the online chan-
nel is still small in relation to the total business. Nevertheless, our research
provides important insights. In particular, managers should carefully evalu-
ate the effects of going online, based upon the transaction size, as well as the
routine or pattern according to which transactions occur. For instance, an
increase in average transaction size for offline customers going online can be
offset by lower frequency and regularity. Since online channels also cannibal-
ize on physical stores (see further Hernant and Rosengren, 2017), priority
should be placed upon marketing activities that encourage customers’ cross-
channel purchasing: both to benefit more from customers acquired online
and to reduce cannibalisation. For example, promotion and reward offers to
customers can be based upon their degree of cross-channel purchasing. There
is an important trade-off between cost functions online and offline from an
operational perspective, as well as between short-term customer responses
and longer-term potential effects on customer behaviour and revenues. For
managers, our study highlights the importance of considering effects on cost,
profit, customer relationships, loyalty, and long-term revenues when consid-
ering multi and omnichannel strategies.
COMPARING MOVEMENTS ACROSS RETAILERS

The framework can also be used to gain a better understanding of move-
ments between online and offline channels in different retail contexts. In a
second step of the project, we added the same type of data from two addi-
tional retailers: one grocery retailer and one beauty retailer. The framework
was adapted to weekly behaviours to better fit the purchase cycle of grocery
retailing, whereas monthly behaviours were used for the other two.
110
As buying motives as well as buying processes differ substantially between

industries, adding an online channel should have differential effects upon
customer behaviour and sales performance between various retail industries.
For example, Kushwaha and Shankar (2013) found that customers that buy
from more than one channel increase their purchases for hedonic product
categories, yet not for utilitarian categories. Their study did not, however,
address the effect upon retailers’ sales performance. Still, it is reasonable to
expect based upon this finding that retailers in industries facing consumers
with hedonic motives for their purchases have more to “win” from also going
online.
In comparing the changes in behaviour for existing customers also starting
to buy online, the three cases reveal the following:
ÜIn all three cases, the average purchases in the physical store are
reduced after customers start buying online (range: -22 to -24%). Custom-
ers who start to also buy online increase their total spending with the
retailer in grocery (+23%) and beauty (+19%). There is no such effect for
the pharmacy.
ÜIn all three cases, the average purchase is larger online than offline.
When customers start buying online, this means the average purchase
in total (considering both online and offline purchases) increases by
13–15%. At the same time, the average purchase offline is lower in all
three cases (Pharmacy: -11%; Grocery: -8%; Beauty: -5%).
ÜIn all three cases, the purchase frequency is lower online than it is off-
line. Customers who start to buy online become less frequent offline
(Pharmacy: -10%; Grocery: -5%; Beauty 6%). The overall effect on fre-
quency is positive for Beauty (+7%), yet it is small for Grocery (-1%) and
Beauty (+2%).
ÜThe purchase regularity in all three cases is lower online than offline.
Purchase regularity increases overall after the first online purchase for
Grocery (+17%), yet it decreases for Pharmacy (-5%) and Beauty (-2%).
The results thus support the notion that omnichannel strategies need to be
adapted based upon product categories and purchase behaviour (Blom et al
2017). Achieving high online purchase regularity among customers seems to
111
be most beneficial for overall sales (online plus offline) in the beauty industry.
Cannibalisation is brought about to a larger extent in the grocery industry,
although overall sales are increasing here too. With its primarily utilitarian
motives and planned purchases, cannibalisation on the offline stores is the
highest in the pharmacy industry.
Conclusion
This chapter set out to provide a better understanding of digital shopper
behaviour and its implications for retailers. Whereas previous research has
primarily focused upon the growth of online retailing, we focused on omni-
channel retailing: the integration between online and offline retailing. More
specifically, we outline the importance of understanding, managing, and
evaluating consumer behaviour across channels and touchpoints when craft-
ing an omnichannel retail strategy. We discuss how it is not possible, based
upon two ongoing projects, to provide a one-size-fits-all key to omnichannel
retailing. Omnichannel strategies must be designed with in-depth knowledge
about the type of products and decision processes used by consumers. There-
fore, it is possible to come up with more tailor-made strategies to manage
customer movements between channels and touchpoints, based upon an
analysis of the customer journey for different customers. The chapter also
points to the importance of combining measures of customer experience,
sales, and brands when evaluating different channels and touchpoints
through the use of dashboard approaches. The frameworks presented in the
chapter can be used as tool in this regard.
112
References
eMarketer. (n.d.). Digital buyer penetration worldwide from 2016 to 2021. In Statista – The
Statistics Portal. Retrieved October 8, 2017, from https://www-statista-com.ez.hhs.se/
statistics/261676/digital-buyer-penetration-worldwide/.
Blom, Angelica, Lange, Fredrik, and Hess, Ron (2017), Omnichannel-based promotions’
effects on purchase behaviour and brand image, Journal of Retailing and Consumer Services,
39, 286–295.
Grewal, Dhruv, Roggeveen, Anne, and Nordfält, Jens (2017), The Future of Retailing, Journal
of Retailing, 93(1), 1–6.
Hagberg, Johan, Jonsson, Anna, and Egels-Zandén, Niklas (2017), Retail Digitalization:
Implications for Physical Stores, Journal of Retailing and Consumer Services, 39, 264–269.
Hernant, Mikael and Rosengren, Sara (2017), Now what? Evaluating the sales effects of
introducing an online store, Journal of Retailing and Consumer Services, 39, 305–313.
Kushwaha, Tarun and Shankar, Venkatesh (2013), Are Multichannel Customers Really More
Valuable? The Moderating Role of Product Category Characteristics, Journal of Marketing,
77 (4), 67–85.
Lemon, Katherine N. and Vehoef, Peter C. (2016), Understanding Customer Experience
Throughout the Customer Journey, Journal of Marketing, 80 (6), 69–96.
Pauwels, Koen and Neslin, Scott A. (2015), Building With Bricks and Mortar: The Revenue
Impact of Opening Physical Stores in a Multichannel Environment, Journal of Retailing, 91
(2), 182–197.
Verhoef, Peter C., Kannan, P.K., and Inman, J. Jeffrey (2015), From Multi-Channel Retailing
to Omni-Channel Retailing – Introduction to the Special Issue on Multi-Channel Retailing,
Journal of Retailing, 91 (2), 174–181.
113
CHAP TER 5
Digital Trace Data:

Which Data Should we Collect and
What Should we do Once we Have it?
CL AIRE INGR A M BOGUSZ
Introduction
Today, we take for granted the fact that websites and other platforms are
optimised to provide different experiences for different people. Individuals’
browsing history, location, and device are all bits of important information
that help web services provide them with a personalised experience. These
personalised experiences are incredibly useful for individuals; someone on a
mobile phone no longer has to navigate a complex web-based version of a
platform, for instance, because data about the individual’s device are received
by the website provider. Similarly, advertisements that individuals see are
tailored for them, based upon what they have looked at — and read — online.
While there is no question that these developments are convenient, some
question whether data collection has gone too far.
Whether or not we are aware of it, these practices form the very backbone
of some of the largest internet firms. Google, for instance, makes much of its
revenue from advertising: in mid-2016, Google’s parent company Alphabet
made 21.5 billion USD in revenue, of which 89 per cent (or 19.1bn USD) came
from advertising ( Johnson 2016). Facebook made 8.81 billion USD profit in
2016, exclusively from advertising — with 84 percent of that coming from
mobile advertising (Constine, 2017).
While there are ways to avoid generating data while online, most of us do
not take the trouble to cover our footsteps. This is despite the fact that every-
thing from how our mouse moves when interacting with a website, to sites
115
CLA I R E I NGR A M BOGUSZ
that we visit, to the data that we enter into online forms — even if we never
submit the form — can be, and often are, collected.
The data that enable personalisation are generated on a diverse number of
websites: from social media to news. And, all of these websites collect these
data, whether it is for their own uses or otherwise. When it comes to their
own uses, many websites pay for their own existence by selling advertising,
and they use our data to match our profiles with the most relevant advertise-
ments. In principle, this should give us the most relevant advertisements
online; however, the process also allows websites to charge advertisers more
for better targeting. Other actors who make use of these data are third party
actors; they bundle data from multiple sources and sell them to other firms:
sometimes in raw form, and sometimes as analytical insights. This direct and
indirect collection and use of online data has come to be called the commodifi-
cation of data, and it has emerged from our desire to have “free” services on the
web, and the fact that digital footprints – or digital traces – are easy to track.
This chapter will define and give boundaries to the collection of what is
called digital trace data: the data that are generated — and collected — as a
by-product of our online activities. This summary then goes on to show how
digital trace data are being used for both business and illicit purposes, and
zooms in on some examples. Lastly, it discusses how to balance the risks for
individual integrity against the opportunities for new businesses, and “free”
services on the web.
What are “Trace Data”?

“Big Data” is the popular term for very large volumes of data; it is often used
to refer to the volumes of data collected about online activities. However, big
data has its origins in scientific inquiry, and internet-generated data is just the
tip of the iceberg: information collected about the weather patterns by satel-
lites, to interaction data from so-called “Internet of Things” devices connected
to one another, to information about online activities are even more common-
place (McAfee et al., 2012). These large volumes of data have led to more
accurate and sophisticated models in areas such as agriculture and weather
patterns (using weather data), allowed connected devices to make predictions
about when to turn the heating on in our homes, or when to refill the refrig-
erator (from Internet of Things data), to complex models around human
behaviour and preferences.
116
DIGI TA L T R ACE DATA
Large volumes of data allow for automated pattern recognition, the testing
of hypotheses at a rate of knots, and automated model development using
machine learning. When it comes to online activities, the more gritty “digital
trace data” is the most relevant for social — and economic — activities. Individ-
uals typically unintentionally leave digital traces as they browse, shop, and
transact online. These data can be used instead of, or as a supplement to, data
already willingly given by individuals in order to build a clearer picture of
their online activities and preferences.
Most of us are aware, at least peripherally, that we generate data while
online. Most websites display, for instance, a “cookie request”: they ask for
permission to store small amounts of data on our computers. These small files
are linked to a particular website; in turn, the files can be accessed both from
the user’s computer and from the website owner’s server. As such, the files
carry information that is used to fine-tune the user’s online experience by
remembering preferences or providing targeted advertising. Often web pages
contain scripts that allow data to be carried from one visit (or page) to the
next: for instance, to optimise advertising.
Cookies are just the tip of the iceberg: not only are we mostly aware of
them; there are limits on what can be shared and are regulated by bodies such
as the European Union (Directive 2002/58/EC). Other traces left online are
not as tightly controlled. However, to understand why (and how) this is the
case, we need to explore what it is that we are talking about when discussing
digital traces.
These different ways in which data are collected, the different kinds of
data, and the extent to which we have control over the data collected allow us
to classify digital traces to some degree. Drawing upon Schneier (2015) and
Ingram Bogusz (2018), this chapter describes a taxonomy of digital trace data,
and discusses the different ways in which these data are being used commer-
cially, both on their own and in combination with other data. This summary
then goes on to discuss the possibilities and pitfalls of data use.
DATA WITH CONTENT, AND DATA ABOUT DATA

Data typically are one of two kinds: data with content or metadata. Data with
content are substantial and personal in nature; they say something about an
individual, and can easily be identified as being linked to that person. Data
with content include not only that which we explicitly share with a firm; they
117
also include other kinds of trace online data, notably content shared on social
media and in forums. When it comes to social media, for instance, photo-
graphs on Facebook are data with content, as are the links to news articles
that we share.
In contrast, metadata are data about data. For instance, metadata around a
Facebook photo might include the size of the photograph, the time it was
shared, and the IP address from which the image was shared. While the
metadata from a single photograph cannot be used to say much about an
individual, the metadata about all of the photographs of you shared on Face-
book can. For instance, if you consistently share large files, from the same IP
address in Stockholm at the same time on a Friday night, algorithms might
determine that you have a high quality camera and therefore are a photogra-
phy enthusiast, who lives in Stockholm, and prefers not to go club-hopping.
Many companies — and countries — treat the collection of metadata (and
other “anonymised” data) as unproblematic. Indeed, metadata are often cen-
tral to, for instance, a telecommunications firm, which ensures their internet
service infrastructure is working as it should. However, depending upon the
patterns searched for in the data, it could reveal more about individuals than
data with content. Therefore, these data could give firms, and anyone else
who can access these data, an unprecedentedly detailed picture of a number
of online habits. For instance, the presence of a mobile phone at an anti-gov-
ernment protest in an autocratic country might reveal the identities of indi-
vidual protestors.
A recent study, for instance, using only the metadata from phone calls and
text messages identified that a small sample of individuals were suffering
from sensitive medical conditions (Mayer et al, 2016). The amount of data
that is currently available about us, combined with advances in data analysis,
have significantly increased the likelihood that an individual can be re-iden-
tified from anonymised data: whether metadata or otherwise. In fact, remov-
ing personal data from digital traces (for instance, by making it illegal to col-
lect personal data) is, therefore, insufficient: identifying an individual
depends upon the number of data traces available, and that to which other
data a dataset can be linked.
118
LE AVING TR ACES IS AL MOST UNAVOIDABLE

Most of us are familiar with giving some of our data to companies and
authorities; it would be impossible to get a bank account, or access healthcare,
without disclosing our names, addresses, and other information. We are fully
aware of these data, however, and what they say about us. This is not always
the case with digital trace data. Although the comparison is imperfect, digital
trace data has been likened to the data left behind at a crime scene.
Take, for instance, when a perpetrator leaves behind a strand of hair or
fingerprints on a doorknob. Criminals are not the only ones that leave these
traces, and neither criminals nor passers-by leave these traces deliberately.
Even so, they can be used to identify individuals—whether at face value by
their appearance or colour, or after further analysis. Thus, these traces may
be said to have “content”.
Avoiding leaving these traces in the physical world, while possible, is tricky.
A perpetrator could wear gloves to avoid leaving fingerprints, or a hat to
avoid leaving hair, but what about footprints or skin cells? It would take
considerable effort, if possible, to avoid leaving any traces at all. Leaving these
traces in the digital realm is similarly difficult to avoid: online activity can be
re-routed through multiple servers, and users can use special web browsers.
Few, however, go to these lengths.
For this very reason, online data are readily accessible. While we give
explicit permission for some services to make use of our data — for instance,
Google and Facebook — data are accessible even to those to whom we do not
give explicit permission. We also often give implicit permission for our data
to be used and stored, simply by using certain websites: such as forums or
news websites.
From Where do They Come?

These two overarching categories are the ones most often discussed in policy
documents, for example. However, individual users have different levels of
control over data, depending upon who shares them and who controls what
is being shared. For this reason, researchers have further characterised data.
While there are many taxonomies, we favour Schneier’s (2015) in this chapter.
Relying upon his framework, we discuss some of the different types of data
out there, their source, and who controls the data in various cases.
119
SERVICE DATA
We are largely familiar with giving our service data to service providers:
one’s name, age, address or credit card number are common service data.
Indeed, these data are willingly given in the offline world to everyone from
banks to state actors. However, despite how widely these data are used, they
are considered to be very sensitive. Ironically, although we often willingly
provide these data, they are the most heavily protected in most countries; for
instance in Sweden, other data may not be used to infer these personal details.
These service data, in the digital world, have the least use: they give only
the most basic of details about an individual, and information contained in
service data can often be inferred from other data. For instance, someone’s
location could as easily be inferred from an IP address as his or her physical
address. Moreover, an IP address can pinpoint where an individual is located
at a given time, and not just what their home address is: useful information if,
for instance, for a targeted advertisement for coffee at 7am.
Considering how service data has been the backbone of the service indus-
try for decades is also illuminating. Today, other sources of data are more
enlightening than this service data. Therefore, it is interesting that individu-
als (and governments) are protective of these data, when other kinds of data
contain the same information: often in real time. Other sources of data often
reveal more about an individual than a name and address can.
Theses service data, however, can be combined with other sources of data,
and others’ service data, to create data that provides more — and deeper —
insight.
DERIVED DATA
Derived Data are data inferred from other data. For instance, combining
service data from thousands or hundreds of thousands of individuals allows
marketers to create segmentations. Offline brokering firms create group
profiles that categorise people according to their shared demographic traits,
while online information brokers tend to use social media networks, device
locations, and online activity.
The creation of these categories is done by individuals or machine learn-
ing, with no input from the individuals whose data are being curated. Thus,
an individual’s membership of a group created based upon either demograph-
ics or online activity is not something that they can influence: even if that
120
categorisation is inaccurate or just plain wrong. Moreover, as third parties

create these categories, the individual user has no influence over how these
group-level data are used by the brokering company or other third parties.
DISCLOSED DATA
Disclosed data, on the other hand, are data that include content that we as
individuals control, according to where we control the platform. This kind of
data includes content such as photos, messages, and/or comments that we post
on a webpage, blog, and/or website that we control, own, and/or host. While
the data are publicly available, we can decide what to share, and for how long.
In principle, this should mean that we could limit access to the underlying
infrastructure, thus, limiting the collecting of digital trace data by third par-
ties. However, the reality is the “public” nature of these data — even though
we control the content — mean that third parties wanting to use it can easily
do so as well.
We often think of data that we put up on social media sites as being dis-
closed and within our control. However, this is not the case. Even data that
we flag as “private” can be used by social media giants for third party ser-
vices, such as advertising targeting. For instance, up until 2016, there was talk
of using Facebook data for credit scoring1.
ENTRUSTED DATA
What we often think of as disclosed is really entrusted data instead. This
includes similar content to disclosed data, yet it is data posted on a platform
we do not control, such as Facebook, LinkedIn, or our employer’s website. As
such, someone else decides what happens to these data, and how easy they are
to use and collect. We can decide the content and whether or not we chose to
post it on these platforms; however, we cannot control what firms subse-
quently do with our trace data.
Entrusted data has been a goldmine for internet giants. For instance, by
making use of entrusted data, Facebook has built some of the world’s most
reliable facial recognition software. By using photographs online and users’
tags of their own friends, Facebook has been able to teach a machine-learning
algorithm how to recognise and classify facial features. This algorithm is now
1 http://fortune.com/2016/02/24/facebook-credit-score/
121
not only better than humans are at facial recognition; its use online has been
called “biometric invasion of privacy” in court proceedings that aim to curb
its use (Brandom, 2016).
Facebook and other social networks are also renowned for their use of data
generated as a result of this entrusted data, namely incidental data.
INCIDENTAL DATA
Incidental Data are data generated as a result of the sharing of entrusted data.
For instance, comments on photographs or on shared links are incidental
data. The tag on a Facebook photo, which identifies an individual, is also
considered incidental data. Incidental data, as with entrusted data, is beyond
the user’s control: both because of its platform and because it is generated by
a third party.
Incidental data in the business world are often used to train machine-learn-
ing algorithms or to generate business insights. The example of Facebook’s
tags is an instance of algorithm training. This data can also be used to gener-
ate business insights when analysts use natural language processing to assess
whether or not a post or online content has been positively or negatively
received — not just whether or not it has been shared.
BEHAVIOUR AL DATA
Lastly, there are behavioural data. These data are created while interacting
with a computer, mobile phone or tablet. Some examples include how long
one spends looking at a particular website or where one clicks. These kinds of
data provide insight into what we do, with whom, how often, and where.
These behavioural data are among some of the most valuable data to collect:
they allow websites to give individuals tailor-made advertisements or special
offers.
Behavioural data have even been used to conduct credit risk assessments.
Wonga, a payday lender in the United Kingdom, claims that its behavioural
data-driven algorithms are so reliable (and quick) that decisions are made
within six minutes, and that money is transferred to user accounts in fifteen
(Deville, 2013). Wonga does this by tracking how a person uses a sliding
credit bar (dragging it straight to the maximum amount is apparently a red
flag). Moreover, Wonga seems to offer individuals higher initial loan amounts
based upon the device from which they access the site.
122
Table 5.1: The Characteristics and Kinds of Digital Trace Data, from Ingram Bogusz (2018)
Deliberately Left Unintentionally Left Left by a Third Party

Data with content Service data Entrusted data
Disclosed data Incidental data
Entrusted data
Metadata Entrusted data Incidental data
Behavioural data Derived data
Derived data
Both behavioural and incidental data are typically unintentionally (or unknow-
ingly) shared. This commonly occurs when we allow one service access to data
contained in other services; for instance, when we allow the Facebook mobile
app to access our phonebook, we ultimately are sharing our friends’ phone
numbers. The fact that data are unintentionally shared, however, does not
affect who has control over when and to whom, data are released.
All of these kinds of data could be either data with content, or metadata:
that is to say, data about data. We unknowingly generate this metadata over
the long term in an organised format.
Having discussed the volumes of data that we generate, and the differences
between them, we now turn to the broader trend of commodifying data,
before discussing how to approach the possibilities implicit in these data with
privacy in mind.
Patterns in Commercial Trace Data Use

Facebook’s average revenue per user in the US and Canada was around $20
in 2016 (Oreskovic, 2017). This revenue is largely a result of the social media
giant’s access to volumes of data that, at scale, it can use to create insights and
new products. These new products include targeted advertisements, news
feeds that contain “recommended” posts and new software, such as the afore-
mentioned facial recognition software.
What is key is that individual data sets are not worth this money: data can
only be used to build new products and train algorithms when an actor con-
trols and maintains vast quantities. While building and maintaining this
infrastructure costs money too, the value of the data is only growing: the
European Commission (2016) estimates that, by 2020, the value of European
123
citizens’ personal data is expected to reach 1 trillion EUR, or 8 per cent of the
Union’s GDP.2
DIGITAL INSIGHTS AND RECOM MENDATIONS

Facebook, Google, and other social media giants are, at their very core, data
brokers. They use entrusted and incidental data to build profiles of individuals
for various purposes. They also access other data to know where we are, such
as our phones’ GPS position. These profiles and physical world indicators
allow them to create some of the following business and service innovations.
Consumer Segments
Segmentation helps retailers online and offline identify potential customers:
for instance, “under 40 without a mortgage” or “young mothers in the Upp-
sala region”. These profiles can then be sold to other companies for advertis-
ing or marketing — whether through the data broker’s platform or otherwise.
The media giants with whom we are familiar, however, are just the tip of the
iceberg. There are even more data brokers of whom we have not heard: US–
based company ID Analytics has information on more than 1.4 billion con-
sumer transactions. The data to which they have access goes far beyond what
Facebook or Google control; instead, they can offer third parties detailed
pictures of consumer browsing and purchasing practises, their interests, hab-
its, hobbies, communities and opinions.3
Consumer Behaviours
Knowing whether a visitor to a site is a “first time visitor” or “everyday browser
(who never buys)” can be vital information for an online retailer. These catego-
ries help online platforms optimise their appearance — and offerings — for dif-
ferent people, depending upon their internet profile and browsing history. This
information, and the resulting personalisation of platforms, is both useful for
individuals and for retailers: it allows the platform to better meet the con-
sumer’s needs that, in turn, increases its own income.
2 http://ec.europa.eu/justice/data-protection/files/data-protection-big-data_factsheet_web_en.pdf
3 Federal Trade Commission (2014) ‘Data Brokers: A Call for Transparency and Accountability.’
Washington DC: Federal Trade Commission. p.iv.
124
NE W PRODUCT DE VELOPMENT
Training Algorithms
Algorithms using machine learning are increasingly common online and
offline. Google and Uber’s self-driving cars, for instance, are steered by algo-
rithms that have been “trained” using environmental feedback. Other kinds
of algorithms are often designed — and trained — using online data: for
instance, ones that automatically calculate credit risk, set prices or recom-
mend products.
These algorithms are given initial instructions, often in the form of exper-
iments, which rely upon one set of data to complete. Based upon initial data
and instructions, a machine-learning algorithm builds a model of some sort.
This model is then tested using either additional data (“supervised model-
ling”) or user feedback (usually “unsupervised modelling”). Therefore, being
able to build these kinds of algorithms requires access to large volumes of
data. And the larger the volumes, the more accurate the algorithm is likely to
be. In a study by US credit assessor FICO, using machine learning was said
to improve the accuracy of a credit assessment by 10-25 per cent, depending
upon the methods used.4 One caveat to this is that only the right kind of data
can generate these results: not only are some data not inherently useful; the
cost of extracting them may be more than the possible benefit they reap.
Making Processes into Products

Building a credit score today is something that is based upon transaction and
financial activities, as well as service data. Individuals build up credit scores
by borrowing and repaying progressively larger amounts of money: by con-
sistently having their salaries paid into a single account, and by paying bills
on time. Financial institutions can lend a consumer money at a given interest
rate, based upon these and other consumer-disclosed or service data: with
higher interest rates correlating to high-risk lending. However, these data
points provide only the broadest frame for assessing an individual’s credit-
worthiness.
Moreover, behavioural and incidental data can provide a clearer picture of
an individual’s disposition to repay a more accurate credit score. Third parties
4 http://www.fico.com/en/blogs/analytics-optimization/how-to-build-credit-risk-models-using-ai-and-
machine-learning/
125
– including data brokers — have, therefore, turned the calculation of these

scores into new products. The ability to screen potential borrowers more accu-
rately and possibly more quickly than competitors is, indeed, a source of
competitive advantage. The provision of these kinds of products can draw new
potential customers into the credit ecosystem. In China, for instance, the use
of digital traces has meant that people who were once ineligible began to get
credit, which served to the benefit of the economy at large (Bateman 2017).
Discrimination and Profiling

While creating consumer profiles for the purposes of providing personalised
services may seem sensible, these methods can sometimes be used to “profile”
individuals for nefarious purposes. As a test, a research team at Stanford
University recently created an algorithm that, by using public images of faces,
could identify the sexual preference of the person in the image. Moreover, the
algorithm was more accurate than the average human.5 The abundance of
public data has meant that, while algorithms can be created to do commer-
cially and socially valuable things — such as track weather patterns and
identify health risks—the data that are out there are use-agnostic. Therefore,
it is possible for data to be used to support immoral and even dangerous
developments.
An algorithm that identifies homosexual men or women is just the tip of
the iceberg. Given how digital traces are increasingly used to train
machine-learning algorithms, even the creators of algorithms lose control of
what it is their creations do with the data — and what kind of heuristics they
create. Algorithms used in hiring decisions have been observed to adopt
human biases because the data upon which they rely contains these biases.6
One price-setting algorithm has been known to use race as a proxy for aca-
demic achievement.7
What is more, these algorithms often self-teach (“unsupervised model-
ling”), making it unlikely that a human would notice and figure out how to
reverse discrimination.
5 https://www.wired.com/story/ai-research-is-in-desperate-need-of-an-ethical-watchdog/
6 https://hbr.org/2016/12/hiring-algorithms-are-not-neutral
7 https://www.propublica.org/article/asians-nearly-twice-as-likely-to-get-higher-price-from-princeton-
review
126
Implications for Businesses

Most businesses today have a digital presence of some sort — whether through
their website, digital advertising, or online customer service (often all of the
above). This means that most businesses have the potential to benefit from
the data that is being generated, which is close to their brands, products, and
services. A few things to think about include the following:
OBTAINING THE DATA M AY COST MORE THAN THEIR WORTH

Data from existing products and services can be hard to access, largely
because online services built in the past did not prioritise data in the same
way. Thus, data are often not collected; they may be siloed or only a very
limited dataset collected. Business owners may be misled into thinking that
all data are gold; however, this is not true. Only data that are accessible, and
relevant to a given question, can improve business outcomes.
BUILD A NE W SERVICE OR PRODUCT WITH DATA IN MIND

Although there are vast troves of data out there, accessing them can be messy
and expensive. For this reason, businesses building new digital services are
advised to construct them a little more slowly, and think about the kind of
data that might be collected — and the way in which they might be used as
part of the service or product innovation process. Collecting and structuring
the data in a useful way right from the beginning makes it easier to work with
later; this maximises its potential.
INFOR M YOUR CUSTOMERS

Not only does informing your customers that you are collecting their data
build trust, it is currently required by a myriad of laws. However, telling them
what you are using it for is not yet required. This will shortly become manda-
tory as a result of the European General Data Protection Regulation (GDPR),
which comes into force in Sweden in 2018. However, even if it were not
required, it makes business sense to be transparent about the way in which
you use your customers’ data: in order to avoid a backlash.
DATA PROTECTION BY DESIGN

The GDPR also encourages what consumers are increasingly demanding any-
way: what is known as “data protection by design”. This principle encourages
127
architects of data services to use techniques such as anonymisation, pseudo-

nymisation, encryption, and other protocols for anonymous communications.
The European Commission encourages the use of these techniques as ex ante
protection from data violations, and has offered to support member states in the
technical implementation of such measures.
Moreover, these laws have also tried to avoid lax data security by making
more severe penalties for data breaches; the GDPR prescribes a fine of up to
20 million EUR or 4 percent of a firm’s global turnover (whichever is higher)
for companies who misuse personal data or fail to take proportionate steps to
prevent data breaches. This protection also helps to deal with some of the
problems previously identified around metadata.
Policy Considerations
Although the abundance of digital trace data allows for the creation and
optimisation of large numbers of new services, using these data runs the risk
of infringing on individual integrity. In examining digital trace data and
legislating around its collection, storage, and use, regulators must find a bal-
ance between the commercial imperative to support new business creation
and the social necessity to support individuals’ data integrity. This section is
devoted to some of the important elements that data legislation should — and
increasingly does — include for the purposes of individuals’ protection.
INFOR MED CONSENT

While consumers are often told that their data are being collected, it is not
always clear for what purpose — or exactly for which data — they are giving
their consent. In fact, the use of machine learning may even mean that corpo-
rate data scientists are not always sure themselves what it is that their algo-
rithms are prioritising.
In 2008, researchers at Stanford University estimated that it would take 7.6
days per year for the average person to fully read all the privacy statements
they encountered in their lives (Symons & Bass, 2017). In practice, the ”take it
or leave it” of most online service terms and conditions means that people
have no choice but to grant access to their data to a large number of compa-
128
nies. That is to say, there is no option to make use of, for instance, Google’s
services without permitting them to analyse and sell the data they collect.8
Moreover, most of us barely acknowledge (and seldom read) notices around
how our data are collected online. In fact, the norm when using a website is
often just to “accept” the terms and conditions of its use, without reading what
they entail. This means that users often do not know that their digital traces
are being collected, and do not know what are the ways in which the data are
being used nor to whom they might be sold.
Users typically make use of services without being able to limit the extent
to which data are collected and used, even if they were aware of it, which
many are not. Moreover, even if individuals were aware that their data was
being used, they would be hard-pressed to understand how its use would
affect the financial (and other) services they receive.
The GDPR requires that individuals give specific and informed consent to
how their data are used and collected. While this is a move in the right direc-
tion, the complexities of algorithms and the fact that individuals seldom read
online terms of service means there is still more to be done.
While it is hard to legislate or avoid over-use of user data when consent is
given, GDPR has also promoted better safety measures to prevent the theft of
data, and the non-consensual identification of individuals.
CRE ATION OF “DIGITAL COM MONS”

While the GDPR goes a long way toward protecting individual data, it has
been suggested that individuals should have control over digital traces about
themselves online: for instance, in the European Court of Justice’s 2014 ruling
on the “Right to be Forgotten”.9 In the UK, one proposal has been to create a
registry of data used by firms (Downey, 2016). In Australia, draft legislation
has proposed a National Data Custodian body to allow individuals to have
greater control over the data collected about them by both public and private
sector actors (Bindi 2016). Germany, known for the importance it places on
privacy, treats data protection as a consumer protection issue, with breach
offenses under the law.
8 Some uses of the data can be limited, but users seldom know that these limitations exist—or how to
make use of them
9 ECLI:EU:C:2014:317
129
Pentland (2013) suggests that our digital trace data should be managed by
data controllers in a way akin to how our banks manage our money. He
highlights the tenets of possession, use and disposal, arguing that these are
the three areas of digital trace data leverage that should be regulated and
overseen. He describes these tenets as follows:
You have the right to possess data about you. Regardless of what entity collects the data, the
data belong to you, and you can access the data at any time. Data collectors, thus, play a role
akin to a bank, managing the data on behalf of their “customers.
You have the right to full control over the use of your data. The terms of use must be opt-in
and clearly explained in plain language. If you are not happy with the way a company uses
your data, you can remove the data—just as you would close your account with a bank that is
not providing satisfactory service.
You have the right to dispose of or distribute your data. You have the option to have data
about you destroyed or redeployed elsewhere. (2013:37)
An experimental project called DECODE (DEcentralised Citizen-owned

Data Ecosystems) is currently underway with partners in Spain and the Neth-
erlands; it aims to develop technology to facilitate this “data commons”. The
intention is to put people in control of their personal data, and give them the
ability to decide how it is shared. The technology will include an architecture
for controlled and, if desired, anonymised data sharing. Crucially, this project
also explores whether there are viable alternative revenue generation models
in an internet economy, which finances itself predominantly through monetis-
ing personal data
Conclusion
As more economic activity has moved online, records of our activities have
improved and are routinely collected: both with and without our consent.
Much of the activity that precedes an economic transaction is also recorded:
individuals’ attitudes to brands, online habits, and even decision-making
processes can be captured through the things they say and do online. This
personalisation of online experiences extends beyond just how and what we
see online, to what we are offered, by whom, and under which terms.
This chapter has explored the kinds of data that have been made available
as a result of improved — and more determined — data collection by internet
giants and data brokers. However, while this abundance of data has made
130
way for new products, services and analytics, it raises concern around indi-
vidual integrity. While the EU’s GDPR goes a long way toward easing these
concerns, a push in the direction of a “data commons” would give consumers
more control over their data. This is important and, thus, should be discussed
in the long term.
131
References
Brandom, R. (2016, May 5). Lawsuit challenging Facebook’s facial recognition system moves
forward. The Verge. Retrieved from https://www.theverge.com/2016/5/5/11605068/
facebook-photo-tagging-lawsuit-biometric-privacy
Constine, J. (2017, February 1). Facebook beats in Q 4 with $8.81B revenue, slower growth to
1.86B users. Techcrunch. Available online at https://techcrunch.com/2017/02/01/facebook-q4-
2016-earnings/
Deville, J. (2012). Regenerating Market Attachments. Journal of Cultural Economy, 5(4), 423–439.
Downey, P. (2016). Registers in a digital ecosystem. Available online at https://data.blog.gov.
uk/2016/09/12/registers-in-a-digital-ecosystem/
Ingram Bogusz, C. (2018). Digital Traces, Ethics and Insight: Data-Driven Services in Fintech
in Teigland,R; Siri, S; Larsson, A; Moreno Puertas, A; and Claire Ingram Bogusz (Eds),
The Rise and Development of Fintech: Accounts of Disruption from Sweden and Beyond (forthcom-
ing). London: Routledge.
Johnson, L. (2016, July 22). Google’s Ad Revenue Hits $19 Billion, Even as Mobile Continues
to Pose Challenges. Adweek. Available online at: http://www.adweek.com/digital/
googles-ad-revenue-hits-19-billion-even-mobile-continues-pose-challenges-172722/
Lorenzetti, L. (2016, February 24). Lenders Are Dropping Plans to Judge You by Your
Facebook Friends. Fortune. Available online at http://fortune.com/2016/02/24/facebook-
credit-score/
McAfee, A., Brynjolfsson, E., Davenport, T.H., Patil, D.J. and Barton, D. (2012). Big data:
The management revolution. Harvard Business Review, 90(10): 61–67.
Oreskovic, A. (2017, February 20). Facebook generated almost $20 from each of its US and
Canadian users last quarter. Business Insider. Retrieved from http://nordic.businessinsider.
com/facebook-almost-20-per-user-us-and-canada-q4-2017-2?r=US&IR=T
Pentland, A. (2013, October). Data-driven society. Scientific American.
Schneier, B. (2015). Data and Goliath: The hidden battles to collect your data and control your world.
New York: WW Norton & Company.
Symons, T., & Bass, T. (2017). Me, my data and I : The future of the personal data economy. Retrieved
from https://decodeproject.eu/file/158/download
Case C-131/12, Google Spain SL, Google Inc. v. Agencia Española de Protección de Datos
(AEDP), Mario Costeja González, 2014 ECLI:EU:C:2014:317, 100(3) (May 13, 2014).
Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002
concerning the processing of personal data and the protection of privacy in the electronic
communications sector. Official Journal L 201, 31/07/2002 P. 37 – 0047.
132
CHAP TER 6
Managing Digital Media Investments

ERIK MODIG AND M ARTIN SÖNDERGA ARD
In this chapter, we discuss the possibilities and the challenges that come with
moving from traditional to digital media investments. In line with popular
opinion, we discuss whether digital media offers opportunities that are not
available in the more traditional stream. Or whether digital media investments
simply seem more attractive due to people’s mentality that the “grass is always
greener on the other side.” The way in which companies adapt to digital media
is likely to impact company success, it is crucial to understand the rules for
these media investments, especially as these investments continue to increase.
We begin this chapter by highlighting how digitalization has led to new prin-
ciples for planning, designing, and evaluating media investments, and how
companies must work in new ways to handle this change. We then discuss
each area of research and summarise the valuable implications.
NE W MEDIA PRINCIPLES
There has been a dramatic increase in the total spending on digital media in
Sweden since 2010, with the numbers more than doubling in the last six years.
This is in line with international trends. Digital media investment is undeni-
ably increasing, and starting to become a norm in which to invest. One of the
largest changes due to this media digitalization is that it is now possible to
target advertisements down to the single customer. We can customise ads and
let algorithms determine the best way to distribute them to audiences. This
potentially increases marketing effectiveness, as every company can become
more relevant for each customer. This is nothing new: the segmentation and
targeting of different media audiences is one of the cornerstones of media
planning. However, the usefulness of doing this has increased significantly
due to digitalization, which has led media planners to rethink how to work
with reach and frequency: the two traditional parameters of media effect.
133
ER I K MODIG A N D M A RT I N SÖN DERGA A R D
There are potential winnings in a new way of optimising; however, findings

also suggest that it might be harmful for companies to rely too heavily upon
the power of targeting and small target audiences.
NE W T YPES OF E XPOSURE
Besides new ways of evaluating and thinking about media planning, digital-
ization has also led to several changes in the kind of media exposure that
companies can buy for their advertisements: from traditional TV and print to
a plethora of different digital formats. This offers many new possibilities;
however, advertisers really need to be aware that digital media exposure is
usually smaller and shorter, and users tend to process it with lower motiva-
tion. We see a shift from desktop to mobile within the digital channels, and
this increase in mobile spending is mostly what has led to increases in digital
marketing. Banner advertising is disproportionately more prevalent in the
mobile format than on desktops. This means there will be even more of an
increase in smaller ads with shorter exposure time in the future. Advertisers
need to understand and optimise their advertisements based upon these
insights into a different kind of exposure.
NE W METRICS
The transformation into digital media has created the possibility of measur-
ing behaviour in the form of a click rather than through changes in thoughts
or emotions. The click-through rate (CTR) is the standard measure used to
determine whether a potential customer has clicked on any digital marketing
element: such as a link, advertisement, social media post, and so on. This has
led to the possibility of connecting certain marketing activities with company
performance, as clicks can lead to specific transactions that marketers can
trace to specific investments. In this way, marketers can evaluate certain
investments based upon their performance. In digital media channels, 64% of
the budget is spent using a revenue model basing the cost upon the perfor-
mance of the ad, thus, focusing upon the exposure of the ad (measures per
impressions or views) and the number of clicks (Silverman, 2017). As afore-
mentioned, large digital players emphasise the importance of so-called
click-attribution. This, of course, is natural since these players sell clicks.
Advertisers must be aware that, although evaluating a media investment
134
M A NAGI NG DIGI TA L MEDI A I N V ESTMEN TS
based upon clicks may seem the perfect way to account for its effects, several
findings suggest this is not a good strategy for all media investments.
NE W WAYS OF WORKING
The change in the digital media landscape also has implications for the struc-
ture of the marketing department. As more channels emerge, new compe-
tences become necessary; this means that companies need to both recruit new
talent and educate existing employees. The digital landscape in particular has
given marketing departments new opportunities to test and measure their
work. This has opened a more agile way of working, and companies that can
utilise these new opportunities and combine them with older approaches that
still work will increase their chances of winning.
We will now discuss how companies should think regarding reach and fre-
quency. Firstly, we examine the new kind of media exposure the digital envi-
ronment offers and the implications for the kinds of ads that companies can
use. Secondly, we look at how companies should best evaluate initiatives in this
new environment. We will end by going through the implications this has for
the approaches of marketing departments. We would like to emphasise the
importance of these issues for people other than the CMO. Since it is absolutely
necessary for the marketing department to stay relevant and competitive, the
CEO and top management should also know about these changes in order to
better understand the transformation and have the capability to manage it.
Media Planning: Questioning Reach and Frequency

The general opinion among marketers is that new digital possibilities allows
for the smartest to win rather than the one with the largest media budget.
Although a larger media budget will of course still allow for reach and fre-
quency, smart decisions in media planning can offset this to a certain extent,
by asking the right questions when buying access to target groups from pub-
lishers and communicating strategically with the right consumer at the right
time with the right frequency.
SEPAR ATING RE ACH AND FREQUENCY

The media industry has previously discussed impressions, in terms of gross
rating points (GRP) and target rating points (TRP), which generally stands
for audience reach multiplied by the frequency of the advertisement. With the
135
advent of digital media, we can break up reach and frequency into two inde-
pendent metrics instead. Depending upon the strategic objectives for a specific
campaign, the point where the marginal returns diminish will differ, thus,
leading to different optimal frequencies. Obtaining a higher frequency on
digital media might even cancel the possible benefits of previous exposures if
the consumer becomes irritated or gets tired of seeing the same ad. Thus,
getting the right exposure frequency and interval of sending a certain ad to
an individual can have a great impact upon marketing metrics. If a low fre-
quency is the aim for a given strategic objective, then it is better to spend the
money on reach. This leads to two types of strategies when media planning:
either goes for wide reach/low frequency or narrow reach/high frequency.
Low frequency might be the way to go if one is trying to increase brand
metrics; if one is trying to generate sales or communicate a new message,
higher frequencies might be necessary.
Table 6.1.
Wide Reach Narrow Reach
High Frequency Strong media impact: Targeted sales focus:
High investment with a risk of The key challenge is to find the right
over-investing. target audience.
Low Frequency Non-targeted brand building: Weak media impact:
The key challenge is to have relevant and Low investment, with the risk that the
good advertisements that can attract low impact has little or no effect.
a wide audience.
Since no impressions are expended to increase the frequency for a few people,
using digital platforms to obtain greater reach can be cheaper than using
traditional media. Nielsen (2016) found that, when buying TRP, increased
spending was no guarantee of reaching more people. Instead, it often reached
the same people: just with a higher frequency. The bigger the campaign gets,
the harder it becomes to reach new people; however, actually communicating
with these people who are difficult to reach can prove very valuable. This is
because those in the target group who are generally not accessible through
traditional media do not receive communications from competitors to the
same extent, meaning that your message does not have to compete in the
same way.
136
The ability to separate reach and frequency provides CMOs with another
lever to optimise the marketing campaign, thus, creating an even bigger dif-
ference between unskilled and skilled marketers in the results they achieve.
Since no unnecessary spending on the other variable is necessary, campaigns
can also achieve higher reach or frequency with the same budget as before.
Unfortunately, marketers cannot take at face value the numbers publishers
quote regarding target audiences and reach, since many use dubious methods
to obtain these conclusions.
QUESTIONING RE ACH
Different actors within the digital media industry have claimed a greater reach
than is possible: for example, saying they reach more people within a certain
demographic than actually exist. Therefore, with regard to reach, it is impor-
tant not to take at face value the numbers publishers provide since they can be
fraudulent. Not only have they overstated reach; they have also claimed to hit
certain target groups that they, themselves, have compiled using questionable
methods. The same is true for traditional media; however, due to the possibil-
ity of tracking everything on digital platforms, many target groups that pub-
lishers compile seem to obtain greater credibility from advertisers.
There are great differences in the target groups available for purchase,
depending upon which data and assumptions publishers have used to compile
the specific groups. The main differences fall into two categories: how the
publisher identifies and tracks the user through a cookie or by being logged
in, and how the publisher segments the audience, based upon audience, con-
text or behaviour. We begin by discussing how to identify the user, and then
we will dive into how to segment the audience.
The advent of the cookie, which can track what a user has been doing,
raised certain possibilities; its promises, however, go beyond what it can actu-
ally deliver. Many users today have several devices, meaning that it is not
possible to reach those people who have allowed us to gather information on
a single unit or to use the data on their other units to target information to
them using cookies alone. Also, many different people may use the same
computer, meaning consumers may see advertisements that marketers did,
indeed, intend for someone else as they are targeted using cookies. In addi-
tion, many cookies disappear over time; the advertiser will only be able to
reach the consumer for a certain period of time after the consumer engages in
137
behaviour that is relevant to that specific advertiser. Services that let the user
log in instead or use their specific app instead of a website can track users over
many devices, as well as store information for a long period of time. This can
also give marketers better control over ad frequency, as one person will not
be counted as two different people and, thus, will not receive the ad twice as
often as intended.
There are three different types of target group segmentation: audience,
contextual, and behavioural. These follow in the order of least to the closest
to a specific purchase.
The basis of audience segmentation lies in knowing which people are using
publishers’ services; thus, demographic variables are very informative. These
types of segmentation are usually possible when users log in and give mar-
keters information about themselves. This is obviously more accurate than,
for example, using the traditional way of first doing a study of who reads a
certain newspaper and then assuming everyone that reads said paper is a
member of this demographic group. Most readers or viewers of a specific
media outlet are not that homogenous; therefore, we also pay for impressions
of people outside our intended group and cannot target certain individuals of
the reader base; rather, we reach everyone that uses that media. Audience
segmentation in the digital era can be much more specific than traditional
media regarding whom it targets: instead of having a goal of reaching 24–39
year-olds, we can focus instead on 26–29 year-old men who also have certain
kind of education, for example. Using geo-targeting, we can also target people
who are in a certain area with a specific message. This can be very useful
when we are trying to push for a sale.
The basis of contextual segmentation is the fact that the information the
publisher is providing to the visitor has to do with a certain topic. This type
of segmentation is very similar to what we have long since been doing in tra-
ditional media: for example, posting ads about a new car in a car magazine.
The difference here, however, is that we can use information in digital media
that we have gathered later. For example, if a visitor reads an article about
cars, we may conclude the reader is interested in cars. Therefore, we might
later send the visitor an ad for a car. Thus, it is important to keep in mind here
that, although the consumer might be interested in cars, it does not necessar-
ily mean the consumer is interested at that specific time. An ad for a car at
that time might not be as relevant as when it appears together with an article
138
about cars. On the other hand, we get the benefit here that we can re-target
the same person multiple times, which is not possible to the same extent when
the ad has a unique connection to the specific article.
One thing to keep in mind here is that some systems that sell ads online
automatically classify what type of topic the specific site discusses, as mar-
keters can post ads on separate forum threads. Since no human is necessarily
involved, it means this classification can sometimes be incorrect. Another
thing to keep in mind is that customers can look for information about a cer-
tain topic; however, that does not necessarily mean they are ready to buy.
They may just find it emotionally rewarding to read about the topic. This
issue is of course relevant in traditional media as well: just not to the same
extent as it is online, as it is much easier to do, and consumers incur no extra
cost by visiting an automotive site compared to buying a car magazine.
The basis of behavioural segmentation is the user behaviour the publisher
has observed; it is often not as accurate as many advertisers believe it to be.
The type of behaviour can vary in relevance to an advertiser; it of course can
be of especially high interest if the search was very specific. Before purchasing
or using any kind of consumer data, one should ask oneself from where does
this data come and what are the underlying assumptions. For example, how
many times must a consumer engage in a certain kind of behaviour in order
to fit into a given segment? Perhaps one kind of behaviour is more telling
than another, meaning that it is worth more to you as the advertiser. Before
using any data, it is important to take into consideration its source, how old it
is, and how often it is updated.
Some publishers are trying to offer specific audience segmentations by
learning how a certain demographic behaves, thus, inferring that anyone who
acts in the same way must also be a member of this group. Depending upon
the degree of behavioural overlap, the quality of this segmentation differs;
however, it is of course not on par with logged-in users’ information. The logic
might be as follows: We have observed that many executives are interested in
soccer; therefore, anyone who shows an interest in soccer is probably an exec-
utive; therefore, we should advertise to him or her services for companies that
would appeal to executives. That is to say, the behavioural overlap is not
related to the specific advertised product; it is a source of demographic infor-
mation instead and a way of deciding what kind of a person he or she is.
139
Advertisers need to question both size and nature of the target audience,
which different media houses say they can provide. In several cases, they are
overpromising when it comes to the number of real people they can reach as
well as the potential increase in relevance based upon their segmentation
method. Huge benefits are possible if one can find the right audience; how-
ever, it is up to the advertiser to find the right match among the large range of
solutions that are available in the digital landscape.
Industry actors have used questionable methods for achieving reach and
communicating with certain target groups, which is why it is important to ask
the right questions. Has the publisher done the identification and tracking
using cookies or logged-in users? What kind of segmentation has the pub-
lisher done and what assumptions does it underlie? Is it based upon audience,
context or behaviour? Not only have critics raised questions regarding whom
the advertisement reaches; they also wonder what should count as an expo-
sure and how we can guarantee the ad was shown the number of times that it
should have been.
QUESTIONING FREQUENCY
Reach is not only the victim of too ambitious promises of delivery; it is also of
frequency, since it can be hard to know how many times and to whom a cer-
tain ad has been shown. To combat this, different technologies have emerged
to detect possible ad fraud regarding frequency to ensure that real people –
and not bots – are the actual viewers of the ads. This technology is constantly
evolving on both sides: as we get better at detecting fraud, they become better
at faking views.
To better understand what type of ad exposure we are purchasing, it is
important to understand the numbers and reports that publishers return to
us: something that can be rather confusing, since the terms and definitions
they use can differ. Has the ad only been loaded, yet never in-screen; has it
been in-screen, yet only very briefly or has it been in-screen for a longer
period of time? There is still discussion about how long an ad should be
in-screen for it to have the desired effect on the consumer. The time will most
likely differ depending upon the objective of the specific ad. If it is only to
increase familiarity, it will need less time than if the objective is to inform
consumers about a new product. The important thing here is that it is not fair
to say one impression equals another. For example, this is true for viewing
140
time; an ad that has been in view for one second on a desktop is most likely
not as valuable as an ad that has been in view for one second on a mobile,
which has to do with the degree of attention the consumer gives to the ad. As
an impression on one platform might not have the same effect as an impres-
sion on another; it is important to go back to the original goals in the form of
metrics in order to understand their different relative values: for example, by
offering a survey of how the ad has affected the attitude toward the brand.
With the advent of digital media, how we think of reach and frequency has
fundamentally changed. We can now focus upon either reach or frequency:
the implication is good choices can lead to an actor with a smaller budget but
better choices, thus, outperforming one with a larger budget who has not
thought through the way he or she is using the ad. Not only must the CMO
understand which variables to focus upon; the CMO must also ascertain the
actual reach and frequency that he or she has purchased, rather than what
seems at face value to be the case. To conclude, digital media investments
offer new possibilities to optimise reach and frequency. However, marketers
need to stay alert and aware to avoid being fooled when media companies
publish both their own measures and targeting tools.
Media Design: Rethinking Media Exposure

The digital transformation not only has implications for how the public sees
advertisements; it is also true for how marketers design advertisements. The
rules for what makes a good advertisement in the digital media are different
from those of traditional linear television or print. This area is still under
development; some guidelines have already emerged that can help with cre-
ative production. Following these insights will give marketers a better under-
standing of the fundamental principles of digital advertising.
SHORTER E XPOSURE TIME

The change in what works in digital media may come from differences in
attitudes across generations, as well as the context and how the content is
delivered. Since the viewer often has the option of skipping the ad after a
certain number of seconds or scrolling past it whenever he or she wants, it is
important for the commercial to hook the viewer instantly. The user spends
approximately 1.7 seconds for any type of content on a mobile, which gives us
an idea of the timeframe that we are talking about; often decisions on what
141
content with which to engage takes even shorter times than that (Capturing
Attention in Feed, 2017). This decrease in time spent on ads does not necessarily
mean that the effect of the advertising decreases as well. Ads can generate
recall even with very short exposures. We should be able to transmit the
viewer a message or a feeling in a short timeframe, within which, we must
hook viewers to ensure they want to continue watching. This is a paradigm
shift from the way TV commercials generally work: by building up a linear
story, and revealing the brand in the end. Since consumers can click away
from the ad, we must hook them so that they want to see more. We must
reverse the general structure of TV commercials, starting with the most
intensive and attention-grabbing visuals at the beginning, and then building
a context around it. The aspect ratio should differ from that on TV to
increase the likelihood of the ad getting attention when presented in a feed,
such as Facebook’s newsfeed, since we want it to cover as large an area as
possible in order to make it as hard as possible to ignore (Tips for Your Facebook
Video Ads, n.d.). Even when we have caught the viewers’ attention, we should
not have ads that are any longer than absolutely necessary; consumers are
much more likely to watch an entire ad if it is 15 seconds or less.
There is currently a discussion in the media about whether or not it is possi-
ble to get 30 seconds’ of exposure that users cannot skip, as was and is the case
with television. This is a big problem, since many no longer know where to put
the main messages of their campaigns. Perhaps they should rethink instead the
underlying assumption that they actually need this type of exposure: it is quite
possible to do the same job with a five-second video ad if they rethink the gen-
eral structure of the video ad according to the viewpoints presented herein.
Advertising presented online needs a different design to what works in
traditional media. Since the attention span is much shorter here, it must offer
the right feeling, thought, association, and so on much quicker than ads on
linear television. This does not necessarily mean the advertising is less effec-
tive because consumers spend less time on it; it just means that a different
design is necessary to influence the consumer. For example, marketers must
invert the logic they use for constructing commercials for television: saving
the best for last. Since consumers can now skip ads or just open another tab
and wait for the ad to end, we must immediately pique their interest. As we
will see in the next section, ads not only need to be shorter; they must also be
142
easier to understand, as ads that consumers see many times are shown in a
context, where consumers take in many other impressions at the same time.
NO SOUND
Even though digital offers possibilities to communicate with sound, only a
limited portion of the audience consumes digital media with sound: such as
news or social feeds. We cannot assume that all viewers have their audio
turned on, and they may be in a loud environment; therefore, video ads can-
not rely upon audio to make sense: they should still have an impact even
without sound; the audio should only be an extra possibility to spice up the
ad even more. One way of doing this is to utilise captions or subtitles, or to
make sure the visual cues the ad uses are easy to interpret.
COGNITIVE LIMITATION
An ad is presented many times in a context where consumers are not pre-
pared for a new message, meaning that it can take the viewer some time to
understand what is going on in the ad. In one way, this is different from the
way television presents commercials: where many are clustered together, so
the viewer is somewhat prepared for the presentation of an entirely different
concept. This means that it is essential for the ad to be easy to understand, not
only since viewers’ attention is limited; it is also because viewers are generally
not very motivated to understand. Since the viewer is often unengaged in the
ad itself and not paying much attention, it is also more effective if the product
is present in at least half of the ad. Ads with lower cognitive loading have a
more positive effect on ad recall (Closing the Creative Loop, 2016), as well as other
traditional marketing metrics. To keep the cognitive loading low, we cannot
work on all the important variables in the same ad, as this will lead to cogni-
tive overload and, thus, no improvement upon any of the variables. We
should instead use one ad specifically to increase its appeal and another one
to inform about a new aspect of our products.
To a certain extent, these guidelines go against what many creators strive
for in their ads: namely an awareness that requires thought to understand, and
once the viewer grasps the point, the concept makes sense. Instead, a good ad
should be short, attention grabbing, and as easy to understand as possible.
Ads in the digital age need to be easy to comprehend due to decreased moti-
vation and attention, so we need to simplify the ads and focus upon specific
143
variables. In turn, this requires a re-definition of a good marketer. This is not

about making the cleverest ads; it about understanding how to make them
short, attention grabbing, and easy to understand. One way to increase the
motivation for viewers to process an ad is to increase its relevance, which is
where digital media offers new possibilities as well as pitfalls.
RELE VANCE
Relevance can be discussed in relation to the media platform as well as to the
individual. Realising that different platforms satisfy different needs within
the viewer is important to consider, meaning that what is relevant and valu-
able in one context might not be so in another one (Why Creativity Matters in the
Age of Mobile, 2017). Different platforms offer value to the consumer in a vari-
ety of ways. For example, many use Facebook to feel a sense of belonging and
to show others who they are, whereas Instagram is more a source of inspira-
tion. The same difference occurs in the traditional media, such as print: for
example, the difference between lifestyle magazines and newspapers. Not
only are consumers looking for different things in different channels; they
might also perceive an ad differently depending upon the context in which
they see it. For example, an ad on Instagram might seem to have a higher
aesthetic value (Why Creativity Matters in the Age of Mobile, 2017). Not only
should brands adapt their ads depending upon the exact channels in which
they are presented; they could also utilise the possibility within digital media
to offer different ads depending upon the segment to which they believe the
consumer belongs; in doing so, they can increase the perceived relevance for
the individual further. Large differences are not required: for example, it can
be a good idea to show products that are actually relevant to the viewers when
advertising a second-hand portal on which people can sell and buy whatever
they want, in order to ensure they make the connection that they can buy and
sell what interests them. This is of course obvious once they think about it;
however, we must make this job as easy as possible for them since we cannot
expect great viewer engagement.
As advertisers, we need to think about digital not as one unified channel,
but rather as many different platforms that satisfy many different needs. To
be relevant in relation to the customer, we must consider what benefits cus-
tomers seek in different outlets as well as try to personalise the ad using the
available data.
144
Due this new form of advertising, ads must communicate their messages
more quickly than traditional media. When done well, this does not mean that
the shortened exposure reduces the effectiveness of ads. Not only does the
viewer process the ads more rapidly; they also use less attention and require
fewer cognitive resources. The bottom line is that ads must be easy to under-
stand. One way to increase the cognitive resources and the time consumers
spend processing the ad is to make it more relevant, which involves under-
standing why the consumer visits a certain media outlet, as well as utilising the
methods previously described to target certain variants of the ad to specific
customers. Even though much remains to be explored within digital market-
ing, these are some established ways of thinking about ads in this new context,
which can lead to more effective advertising. Overall, marketers need to pay
more attention to the specific design requirements for digital exposure. Today,
several marketers spend money on ads that perform much worse than they
would if only marketers prepared them properly for the digital landscape.
Media Evaluation: Questioning Click Metrics

When the internet was a new phenomenon, consumers had a completely
different attitude to digital ads, and they clicked on ads to a much greater
extent than they do now. Many digital companies, such as Google, were
driving forces behind this logic, as they themselves profited from the focus on
CTR. This new currency was easy for other actors within the industry to
adopt, since it is easy to understand intuitively and to measure. These factors
together explain why the issue is in focus today.
Over time, however, the general CTR dropped due to changed attitudes
and increased ad density. This did not mean that the effectiveness of banner
advertising dropped; just that instead of driving a direct behaviour such as a
click, it led to more traditional effects, such as changed attitudes toward the
brand or increased purchase intention. Despite this, many are still evaluating
ads in the digital media on how many clicks they receive. Most people under-
stand the problem of solely focusing on clicks, but they still choose to continue.
What are the underlying reasons for this, and how can they be changed?
TOO E ASY TO ME ASURE AND UNDERSTAND

The explanation you usually get when asking why marketers evaluate a cam-
paign by clicks is usually one of the following: It is a universal measure with a
145
clear standard. It does not require additional bets on panels or surveys to get
an answer. It is easy to communicate upwards, as it is a clear behaviour that
everyone can understand. It often happens close to a purchase. With all these
arguments available, one can still understand why the measure is tenacious,
since all these arguments describe valuable things. The problem lies in the fact
that if the metric does not tell us anything of value, all the above-mentioned
arguments become irrelevant.
THE CLICK’S SIGNIFICANCE IS OVERR ATED

It is partly the nature of the metric that makes CTR problematic. Instead of
measuring how the average has changed, through a survey for example, you
measure those who have been influenced so much that they are very inter-
ested right now in buying or knowing more about the product, which is a very
small percentage. Not only is it problematic just to measure those who actu-
ally engage in behaviour at the moment of exposure, but it is also the case that
many of these clicks are mistakes. If we look at the difference between mea-
sures of a good and a bad campaign, 0.15% and 0.05%, we see that they differ
by 100 clicks per 100,000 impressions. In addition, many clicks on the phone
are mistakes. The intentional clicks are not even an even a proportionate
sample of the people marketers target. Many times, a small percentage of all
visitors make the majority of the clicks, and these visitors tend to be younger
and have lower incomes than average. With this in mind, one can really
question how valuable a click is to the brand.
OTHER BEHAVIOUR VARIABLES ARE MORE IMPORTANT THAN CLICKS

Already in 2003, research showed that the number of clicks was a bad mea-
sure of the effectiveness of a banner (Drèze & Hussherr, 2003). What instead
should be used is the classic brand-equity metrics that have been in use for a
long time in many other media outlets. Even before choosing the channel and
designing the ad, one should decide on the exact goals of the campaign and
how to evaluate the results. If the goal is to increase the brand awareness, it is
important to keep in mind the negative relationship between classic brand
equity measurements and CTR shown in Schibsted’s study of mobiles in
collaboration with Lund University (Olivensjö & Sundberg, 2015), showing
that clicking on the ad led to a worse view of the brand.
146
There is evidence that classic brand-equity metrics have a greater correlation

with actual conversion than CTR has. For example, one study has shown (Pre-
target & ComScore, 2012) that the correlation between having the mouse over
the ad and actual conversion was 0.49, whereas the same correlation for CTR
was 0.01. Not even in the cases where we want the consumer to carry out an
action directly on the mobile phone is CTR the preferred option, as we would
rather look at how many have performed the action and then use cost per action.
If we want to make a fair measurement and evaluation of the investment,
there is unfortunately no single measure that can capture all possible dimen-
sions that are relevant to our brand. Instead, it is necessary to reflect on and
understanding what different metrics in marketing measure.
Consumers’ attitudes toward ads on the internet have changed signifi-
cantly since ads began; yet the ways many brands evaluate their media invest-
ments online have not. Using CTR became the standard for evaluating digi-
tal ads, as some actors within the industry profited from its focus and because
of the ease of measuring and understanding it. The importance of CTR is
overrated, as it only measures the small number of people who actually
engage directly in behaviour, and neglects the fact that many of these clicks
are mistakes. Those who click with intent are not a representative sample of
the visitors to the site; instead, many who do are younger people with below
average incomes. To better measure the effectiveness of online ads, marketers
should use traditional brand-equity metrics and, when they are not available,
other measurements that can be collected automatically can prove better:
such as clicking a mouse over ad or an action that has been taken.
The New Marketing Department

A new way of organising the marketing department is required in order to
capture all these new possibilities and to cut costs using the new technology.
These new ways of marketing require new processes and the modification of
old ones. For these processes to run smoothly and to generate the maximum
value, employees need new competencies and skills. Since we are still in the
early stages of this development, it is not yet clear which systems are the best;
with no established standards, it is hard for the CMO to know in which new
initiatives to invest. Many CMOs feel they face never-ending possibilities;
they are fearful of the possibility that their competitors might be making
better use of them than they are. When there are no established solutions, big
147
media spending is no longer the concern; instead, it is hiring the best talent to
capitalise on these new opportunities. Having the right employees will
become even more important to identify the processes companies that can
standardise and upon which they should prioritise and focus, as staff under-
stands and learns the appropriate skills.
RE-E VALUATION OF M ARKETING PROCESSES

Some processes require an entirely new division, such as search engine opti-
misation; creating ads for digital platforms require modification of existing
processes. Some jobs or processes require entirely different kinds of people,
whereas others require re-educating existing employees. The number of dif-
ferent processes, however, will increase; therefore, it is crucial for the market-
ing department to handle a larger number of different procedures. Some
methods should become more automated than many believe is possible. In
many cases, this does not require new technology; rather, it is the standardi-
sation of some processes is vital for companies, so they do not repeat them
unnecessarily. Marketing departments need to sort out which processes they
can standardise, and focus upon those that require the greatest effort. In
doing so, they can experiment more in the more relevant areas, thus, finding
the recipe for success that will allow employees to be even more creative at
work, and experiment with new concepts.
SET UP E XPERIMENTS FOR CONTINUOUS LE ARNING

With the possibilities that the new digital environment offers, it will be possible
for companies to try things out to a larger extent without incurring much extra
cost. They might run some ads simply to see if a certain hypothesis is correct
regarding what drives growth, and later use that information to improve other
campaigns. The emerging technology of machine learning can automatically
test different layouts and combinations of elements in digital channels and find
the combination that works best. The principle of continuous learning through
experimentation and incremental improvements year after year will become
even more relevant as the professionalisation of the marketing trade continues.
The starting point should be that a large majority of the marketing budget and
campaign will stay the same, spending most of the budget upon what already
works quite well, while using a smaller portion to try out new ideas. By instill-
ing a culture of slow but steady improvement using feedback, the marketing
148
efforts will improve over time. Reacting more quickly to the things that matter
when most of the processes are stable will also become easier to do.
MOVING FAST
Information moves more quickly in the new marketing environment; therefore,
it is even more important to be fast in order to capitalise upon opportunities
and to negotiate threats. Information is also more spread out and more democ-
ratised than it was in the past. A portion of the marketing department for many
companies should focus upon interactions with customers in real time: skills
that are similar to what customer relationship management divisions have been
doing for a long time. The possibility for people to interact with each other in
real time is something in which companies should partake as well. This means
another type of marketing material is necessary. In these contexts, it is no lon-
ger about a small amount of content of high production quality, but rather a lot
of continuous content at a lower production quality. The content can be more
personal, and it does not require the same professional feeling, as the goal here
is to build a more intimate relationship with the customer. This requires not
only a new type of marketing department; different structures in other parts of
the organisation are also necessary, as is a quick clearance to respond to events
in real time. A media budget specifically for capitalising upon current trends
might be necessary to create awareness, which is a great way to obtain more
attention for your brand, as it increases the perceived relevancy: very useful
when consumers receive floods of new messages. A risk voiced by CMOs about
restructuring their marketing departments in accordance with these sugges-
tions is that they will have difficulty communicating one uniform message.
This will not necessarily have as much of a negative impact as they believe.
Being in tune with campaigns that run in the traditional media is, therefore, not
as important for new departments or those that have recently increased in size.
DO NOT DROWN IN THE WRONG NUMBERS

Decisions in marketing departments are increasingly being made using the
new possibilities of measuring marketing impact. However, just because we
can measure something does not mean that it is a relevant number that can tell
us something useful. A structured approach based upon consumer psychology
and research methodology is necessary in order to discern the relevant mea-
surements and the type of decision for which they are required. A quantitative
149
rapport is not worth much if the underlying assumptions and data upon which
it builds is not of high quality. When we need to measure the initiatives in
different media outlets against each other, we must look at what they have in
common: the person who is actually receiving and responding to the commu-
nication. An understanding of consumer psychology is necessary to grasp
what types of metrics are relevant to a specific brand or initiative and how
they compare with each other. Having good knowledge of research methodol-
ogy is necessary to measure these metrics in the fairest way. When focusing
upon measurement and data, it is important to keep in mind that just because
a variable is easy to measure, does not mean that it is relevant; in fact, the
relationship between the two might actually be the opposite. Many of the most
important variables in success are hard to quantify, yet are still very important
to marketing: such as unconscious emotions toward the brand. By forgetting
about these more abstract and esoteric metrics, brands can easily destroy their
brand equity by only looking at the variables they can easily affect, that are
easy to measure, and that lead directly to sales.
The new opportunities and threats in digital marketing require a restruc-
turing of marketing departments to adds new processes and skills. This rap-
idly changing environment makes it hard for the CMO to know upon which
areas to focus. Companies should standardise other processes and marketing
investments to a greater extent in order to allow for the allocation of time and
resources to the respective focus areas. They should reinvest the larger part
of the marketing budget in what already works to allow for a smaller part of
the budget for experimentation and to discover new initiatives that might
work. They should reserve another part of the budget to capitalise upon
emerging trends, as this will allow the marketing of the company to stay rel-
evant. When evaluating all these new initiatives, the focus should be on the
most relevant numbers rather than those that are the easiest to measure.
Surviving The Shift

Transforming to the digital media landscape will not be easy for companies.
They will have to adapt. What we see now, however, is that many companies
are running much faster than their marketing department can handle, and
probably much faster than is necessary. This new technology offers as many
opportunities as it does ways of going in the wrong direction. We have high-
lighted four of the concepts companies need to understand in the digital
150
media landscape in order to optimise their media investments: questioning

reach and frequency, rethinking media exposure, the new marketing depart-
ment, and questioning click metrics.
We began by showing how the separation between reach and frequency
has given the CMO more levers with which to optimise the marketing cam-
paign. When utilising this new opportunity, CMOs need to be careful since
many of the actors within the industry use dubious methods to determine
whether or not someone is a part of the target group. Armed with the right
questions, CMOs can avoid many pitfalls. They also need to investigate fre-
quency to ensure the ad has received as much exposure as the publisher
claims it has. Before buying exposure, it is also necessary to understand
which type of exposure is necessary for the ad to have the desired effect.
We then discussed how companies should design ads for digital media.
They need to communicate their message much faster than in traditional
media, and to use fewer of the consumer’s cognitive resources. They can often
construct ads more effectively, yet this sometimes requires focusing upon only
one thing, rather than many different ones. Ads should be relevant for them
to be effective. Marketers need to consider the platforms on which they will
publish ads and the benefits of using these media outlets, as well as personalis-
ing the ads using information garnered from the user.
The standard of using CTR to measure the success of ads has arisen not
because of its effectiveness in measuring ad impact; rather, it is gathered auto-
matically, thus, it is easy to understand intuitively and the big actors within
the industry have pushed it. To evaluate initiatives in digital media, the CMO
cannot rely upon click metrics, traditional brand-equity metrics should be
used instead.
Discussing the new opportunities and threats due to digital media for the
brand, we saw that the marketing department might have to restructure to
accommodate for these new needs. They must standardise some processes to
a greater extent, whereas others will benefit from more experimentation. The
marketing budget should match these structural changes, with the main part
going to reinvesting in what works and a smaller part to experimenting and
to following up on trends.
We hope this information will arm forward-looking CMOs with insights
that will enable them to survive one of the largest shifts in paradigms since
marketing’s inception.
151
References
Capturing Attention in Feed: The Science Behind Effective Video Creative. (2017, April 20). Retrieved
from https://www.facebook.com/iq/articles/capturing-attention-feed-video-creative
Closing the Creative Loop: Informing Art With Science. (2016, June 13). Retrieved from
https://www.facebook.com/iq/articles/closing-the-creative-loop-informing-art-with-science
Drèze, X., & Hussherr, F. X. (2003). Internet advertising: Is anybody watching? Journal of
Interactive Marketing, 17(4), 8-23.
For Display Ads, Being Seen Matters More Than Being Clicked. (2012, April 24). Retrieved from
https://www.comscore.com/Insights/Press-Releases/2012/4/For-Display-Ads-Being-Seen-
Matters-More-than-Being-Clicked
Nielsen. (2016, July 28). Bigger is Not Always Better: Using Reach Efficiency to Run Smarter Digital
Campaigns. Retrieved from http://www.nielsen.com/us/en/insights/news/2016/big-
ger-is-not-always-better-using-reach-efficiency-to-run-smarter-digital-campaigns.html
Olivensjö, A., & Sundberg, G. (2015). Evaluating the Effect of Mobile Display Advertising—Guidelines
on How to Advertise in the Mobile Channel. Retrieved from URL
Silverman, D. (2017). IAB Internet Advertising Revenue Report. Interactive Advertising Bureau.
Retrieved from URL
Tips For Your Facebook Video Ads. (n.d.). Retrieved from https://www.facebook.com/business/m/
facebook-video-ads
Why Creativity Matters More in the Age of Mobile. (2017, March 28). Retrieved from
https://www.facebook.com/iq/articles/why-creativity-matters-more-in-the-age-of-mobile
152
Re-Organisation
in Order to Bridge the Gap
to Digital Customers
CHAP TER 7
Digitalization of Professional Services:

The Case of Value Creation
in Virtual Law Firms
TALE SK JØLSVIK, K ARL JOACHIM BREUNIG, AND FRIDA PEMER
Introduction
Virtual firms are already here, and are proving to be successful. These firms provide an
opportunity for independent practising lawyers to act as large law firms without giving up
their autonomy... (Lawyer, Virtual Law Firm, US)
From simple information and communication technology (ICT) to advanced

artificial intelligence (AI), using digital technologies is becoming an increas-
ingly integrated part of everyday life in organisations. The use of digital
technology can provide significant business opportunities, yet can also poten-
tially undermine firms’ business models and competitiveness (Christensen,
Wang, & van Bever, 2013). Furthermore, digitalization could change organisa-
tions and their value creation strategies more fundamentally and faster than
any other technological change we have previously witnessed ( Jesuthasan,
Malcolm, & Zarkadakis, 2016). So far, the use of digital technologies has
largely influenced work-intensive firms in production, leading to the automa-
tion of standardised and routine tasks (Frey and Osbourne, 2013). Recently,
the emergence of new digital technologies may also lead to far-reaching
changes for professional service firms, which offer expertise-based and tai-
lor-made knowledge-intensive services (Davenport & Kirby, 2015). In fact,
professional services are thought to be one of the industries in which the effects
of digitalization will be most prominent (Manyika, Chui, Bughin, Bisson, &
Marrs, 2013). So far, however, very few empirical studies supporting this state-
ment exist, and we do not know the way in which professional service firms
155
TA LE SKJØLSV I K, K A R L JOACH I M BR EU N IG, A N D F R I DA PEMER
will change or what consequences digitalization will have for them. Therefore,
we present findings in this chapter from an interview and a media study of the
effects of digitalization on a particular type or professional service firms: vir-
tual law firms1. Our aim is two-fold: First, we take a processual view and
describe how virtual law firms have developed over time. Second, we discuss
the consequences of digitalization for professional service firms operating
within the legal sphere.
The chapter is structured as follows: After describing the characteristics of
professional service firms and our methodology, we provide an overview of
how virtual law firms have emerged. We then discuss consequences of digi-
talization, before concluding the chapter by looking ahead and outlining what
the future might hold for professional service firms and their clients.
Characteristics of Professional Service Firms

Professional service firms are characterised by being highly dependent upon
the professional employees they attract (Malhotra & Morris, 2009). These
employees contribute with competences, experiences, relations, professional
judgment, and networks to their clients (Greenwood, Li, Prakash, & Deep-
house, 2005), and follow professional norms (von Nordenflycht, 2010) while
considering their clients’ specific needs and interests (Løwendahl, 2005).
Employees in professional service firms generally prefer autonomy in their
work and dislike control, standardisation, and formal organisational pro-
cesses (Alvesson & Karreman, 2006). The basic deliverable in this type of
organisations has been characterised thus far by offering tailor-made services,
based upon a professionalised knowledge base and trustful client relations.
ICT has increasingly become an integral part of the work professional ser-
vice firms perform over the past 25 years, and tasks previously performed by
humans are being computerised (Chui et al., 2012; Manyika et al., 2013). As
part of this development, the understanding for how technology can be inte-
grated in knowledge-work has deepened (Christensen et al., 2013; Susskind &
Susskind, 2015). Today, it is difficult to imagine a lawyer or consultant work-
ing without access to important information on the internet, or performing
analyses or presentations without using computers. Documents are shared on
1 A virtual firm is an organisation that involves dispersed entities that need ICT to support joint work
and communication (Hedberg, Dahlgren, & Olve, 1997).
156
DIGI TA LI ZAT ION OF PROF ESSIONA L SERV ICES
servers and communication with colleagues and clients take place via email,
Skype, and social media.
In addition to using technology as support in the daily work, digital tech-
nology can also impact how professional services earn money, organise their
work, and collaborate internally and with their clients. Thus far, the techno-
logical development has revolutionised how individuals communicate and
share knowledge across national and organisational borders (Breunig, 2016).
The emerging AI technologies are likely to influence value creation by being
used to mass-produce or productify knowledge-intensive and professional
services (Sawhney, 2016).The use of AI is suggested to be a potential source of
increased productivity (Chui et al., 2012), as increased economies of scale and
possibilities to standardise services can lead to innovation of new value cre-
ation models (Breunig, Kvålshaugen, & Hydle, 2014), and new organisation
types, such as virtual firms (Breunig & Skjolsvik, 2016).
VALUE CRE ATION IN PROFESSIONAL SERVICES

The concept of value creation is often used to describe what an organisation
produces, what it wants to achieve, and what it is achieving by delivering the
best customer value at the lowest use of resources possible (Skjølsvik &
Voldsund, 2016). While it is often described as a function of what the customer
perceives to be valuable, measured in monetary terms (Hoopes, Hadsen, &
Walker, 2003), value creation is not restricted only to the customer. Rather,
firms also create value: for example, to their current employees, potential
employees, potential customers, investors, and owners. Thus, value creation
can be thought of as aiming to create perceived value among the firm’s stake-
holders and, in doing so, using its resources in a most efficient manner and at
a cost competitive level. The value creation in professional service firms is
linked to tailor-made services and problem solving based upon professional
expertise. The business model underpinning the service deliverable has
changed very little over the last century, and is based upon bundled services
billed by the hour (Christensen et al., 2013). The role of professional service
firms has also been linked to other stakeholders, as this type of firm have been
described as knowledge disseminators that leverage important knowledge
among several actors in society (Starbuck, 1992).
157
Methodology
We have chosen virtual law firms in order to study how digitalization influ-
ences value creation in professional service firms. Virtual law firms is an
internationally established type of organisation (Gordon, Shackel, & Mark,
2012) that forms an illustrative example of how digitalization impacts an
industry that traditionally has been conservative and knowledge-intensive.
The results discussed below build upon an explorative interview study with
20 informant related to virtual law firms, and a longitudinal media study of
news articles on virtual law firms, published in the Factiva Dow Jones data-
base: 2006–2017.
The interview study aimed to provide contextual information and insights
into how virtual law firms use digitalization to improve their value creation.
Our identification of relevant cases and informants started quite broadly and
followed a snowballing logic (Noy, 2008). Initially, we contacted two high-tech
industry specialists based in Silicon Valley with whom we had previous rela-
tions. The first was a COO of a major internet corporation and the second
was an Intel retiree with a 40-year history in Silicon Valley. Subsequently, we
also approached two individuals that work with, invest in, and facilitate scal-
ability of new web-based ventures. We approached these venture capital and
innovation incubator communities to learn more about the market conditions
and latest trends of the high-tech innovation industry within the context of
law firms. In addition, we contacted two professors at the Stanford Law
School and were introduced to their initiative CodeX: The Stanford Center
for Legal Informatics, with particular emphasis upon the intersection between
new ICT and organisational developments for the law firms of the future.
Our first interviewees also introduced us to a former mayor of Palo Alto, now
working as an advisor to tech start-ups, and to the leader of the Palo Alto Bar
Association. During these initial interviews, we were able to identify several
different firms using new technology to innovatively offer legal services that
had started within the last 10 years. Subsequently, eight of these firms based
in the US and UK were contacted and interviewed via Skype in 2015 and
2016. Each interview lasted between 1-2 hours. Those interviewed also pro-
vided us with internal documents and information about their firms, which
was used as background information to better understand the organisations.
An overview of the type of interviews conducted can be found in Table 7.1.
158
Table 7.1: Overview of Participants in the Semi-structured Interviews
Informants #
High-Tech Industry specialists 2
Venture capitalists/Innovation incubators/local municipality officials 3
Silicon Valley based researchers with knowledge of the legal industry 2
Legal professionals related to high tech start-ups 5
Attorneys/Partners in virtual law firms 8
Total 20
The interviews were recorded, transcribed, and analysed in several steps. In

particular, the collected data was analysed using data reduction methods and
an inductive approach (Gioia, Corley, & Hamilton, 2013). First, an inductive
approach was taken to develop relevant first order categories that, in turn,
were compared to existing theory. In turn, these categories were grouped into
larger subsidiary and main dimensions of the business model framework (as
shown in Table 7.2). Memos in Word were used as core properties of the
exploratory categorisation emerged. Subsequently, we chose a more deductive
approach re-coding our data comparing it with different core dimensions of
business models (Osterwalder & Pigneur, 2010), especially with attention busi-
ness model innovation in professional service firms (Christensen et al., 2013).
We used the database Factiva Dow Jones in the media study and the search
term “virtual law firm” to identify articles published in 2006–2017. The
search resulted in 182 relevant articles (see Table 7.2). The findings from the
two studies will now be described in more detail.
Table 7.2: Overview of Articles per Year in Factiva Database. *Only for Half of 2017 Included.
Year 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017*
Number of 16 18 10 20 21 32 21 21 24 34 23 18
total articles
Duplicates/ 5 4 2 2 5 10 4 3 6 23 5 7
beyond scope
Final number 11 14 8 18 16 22 17 18 18 11 18 11
of articles
159
The Development of Virtual Law Firms: 2006–2017

A close analysis of the articles revealed that the emergence of virtual law
firms had moved through three phases: 1) virtual law firms as competitive
equals: 2006-2010; 2) dissemination: 2011-2014; and 3) transformation: 2015-
2017. The three phases will now be described in more detail. As the develop-
ment gained momentum – particularly in the third phase – and, as we are
witnessing a large variety of virtual law firms today, we will give extra atten-
tion to this phase.
2006 –2010: VIRTUAL L AW FIR MS AS COMPETITIVE EQUALS

Following the financial crises, thousands of lawyers and support staff lost
their jobs (see, for example, The Legal Intelligencer: October 16, 2007). Many law
firms also had to cut down employee salaries to survive. In such a market, the
traditional business model was increasingly regarded as being “overweight”
(ibid.) and virtual organisations became a means to realise cost savings.
Previously, the virtual law firms had been small and local, mainly focusing
upon private law. After the financial crisis, however, many large virtual law
firms were established, having extensive geographical reach and directed
towards large client companies. Virtual Law Partners (VLP) and Rimon are
examples of such large virtual law firms that were established during this
period. The business model in this type of firms was to hire lawyers at part-
ner level in traditional law firms who had an established client base. These
firms aimed to provide both clients and employees with better offers than
traditional law firms in order to reduce costs, offer reasonable services, and
let the lawyers keep the main part of the revenues. The lawyers were not
physically present in the virtual law firms, yet were assisted by invoicing
systems, IT support and marketing and recruitment, and collaborated via
software solutions and. The offering of new technologies supported this
development for law firms (see, for example, Directlaw and Virtual Law
Office Technology).
2011–2014: DISSEMINATION AND E XPERIMENTATION

In the second phase, an increasing number of law firms realised the possibil-
ities that the use of available digital technology could bring, and the tradi-
tional law firms started to mimic the virtual law firms. For instance, they
160
started to offer more flexible work hours and alternative career paths to their
employees (see, for example, The Guardian: July 2, 2011). An important new
technological development in this period was the automation tool for docu-
ments (see, for example, Epoq Legal, LegalZoom and Rocket Lawyer). New
management tools for smaller firms were also developed: such as Total Attor-
neys, Clio, Rocket Matter, and MyCase. Yet another important development
was the use of apps and software enabling the law firms to better tailor their
services to their clients and support the development of client relations2.
Virtual law firms were mostly an American or British phenomenon up
until 2010. A game-changing event was when Axiom entered the Financial
Times’ list of the best European law firms. In the end of the second phase, the
phenomenon spread steadily to several cities and countries, and simple vir-
tual law firms such as Axiom sought to build a global presence. This develop-
ment illustrates how organisational advantages associated with the virtual
model enable fast internationalisation.
2015 –2017: TR ANSFOR M ATION

The large number of new virtual law firms was described in the articles dur-
ing this period, as being challenging the legal industry. A first reason was that
lawyers at large law firms started to view virtual law firms as an attractive
employment alternative. As was illustrated in the Australian Financial Review
on 12 August 2016:
So a lot of lawyers are now saying, ‘You know what? It’s just not worth it’. With technology,
you can go and set up your own business, make a name for yourself and do financially better.
Secondly, there was also an increased interest in the legal industry in

LegalTech, and awards were instigated for technology-related legal innova-
tions, thereby, contributing to improving the status of virtual technolo-
gy-based law firms.
Third, new technology related to AI and cloud base computing was dis-
seminated among law firms. During this period, even traditional legal firms
started to explore digital solutions, such as The Robot George tested out by
The Conveyancing Shop Lawyers in New Zealand:
2 See e.g. https://blog.highq.com/news/new-swedish-firm-synch-advokat-selects-highq-for-extranets
161
George is a tele-presence Robot from Double Robotics that includes a tablet screen as a “head’
and Segway like “ feet’ to get around. George can move, back, up, and turn. This is all
directed by his controller – one of the Conveyancing Shop’s specialist lawyers, a client, or a
management team member. (Scoop, 30 July 2015)
The dissemination of the virtual law firm as a business model across geogra-
phies continued during this period, particularly in Australia. While larger
firms in the US, such as FisherBroyles, Potomac Law Group, Rimon and VLP
Law Group were prospering and growing extensively, and the value proposi-
tion of virtual law firms was increasingly recognised (Daily Breeze: 26th March
2015). Media and experts suggested that the virtual platform could enhance
and support the client-lawyer relationship (Business Wire: 28th August 2015),
and that the client-lawyer relationships had become a key focus of these firms:
Many virtual attorneys go out of their way to be more responsive to their clients to counteract
any concerns that a virtual attorney is ephemeral or transitory. This renewed primacy of the
attorney-client relationship in many ways marks a return to very traditional values of the
practice of law. (Inside Councel, December 2nd, 2015)
However, difficulty was also recognised in establishing and running a virtual

law firm (for example, American Lawyer: November 1st, 2016). Failed virtual
law firms, such as the faulter of virtual law firm, Clearspire caused this (ABA
Journal, 1st July 2017). Larger virtual firms were also likely to dominate the
legal service industry and take market shares from medium-sized and small
virtual firms.
We witness a decrease today in the use of the term “virtual law firm”. This
is partly caused by an increased discussion of alternatives to the traditional
law firm model, of which in-house counselling and outsourcing are two
extreme alternatives. Axiom is leading this transformation, which is recur-
rently described as being a key player in offering client advice. There also
seems to be an interest in shifting the terminology away from the notion of a
“virtual” to a “cloud” based firm. This indicates that law firms want to asso-
ciate themselves with more modern, up-to-date, cloud-based technologies that
augment their value offerings to their clients (for example, the collaboration
between the Swedish law firm MAQS Advokatbyrå and the legal technical
solutions firm Virtual Intelligence, VQ 3). The companies also seem to want
3 https://maqs.com/en/news/news/maqs-vq-develop-pioneering-artificial-intelligence/
162
to communicate to their clients that they are not just virtual; they also have a
physical presence. In turn, this suggests that, while technology is an impor-
tant enabler for improving efficiency and accuracy, the personal client-lawyer
relationship is still regarded as important for value creation.
Consequences of Digitalization in Professional Service Firms

In developing a better understanding of digitalization in professional service
firms, we further use the interview data together with the media study to
focus on three main areas: the nature of the digitalization, the organisation of
the virtual law firm, and the impact of digitalization on value creation in this
type of firms.
THE NATURE OF DIGITALIZ ATION IN PROFESSIONAL SERVICE FIR MS

A main finding from the interviews was that the technology used in many
cases was not particularly sophisticated or unique. An interviewee explained
this in the following way:
“The systems we use really are on-the-shelf systems. The important thing is (to)
find out what is already out there and take advantage of it. We don’t need expen-
sive, fancy, or tailor-made systems. We use The Cloud for sharing documents,
online video conferences, social networks to collaborate, and LinkedIn for
marketing.”
Thus, it does not seem that becoming a more digital or virtual firm necessar-
ily implies extensive investments. Only a decade ago, only large firms could
afford investments in advanced ICT tools. Currently, we witness that digital-
ization provides opportunities for small and medium-sized firms as much of
the applied technology is inexpensive and standardised solutions. As many of
the firms combine technology, they develop themselves with existing openly
available technology, and success seems to stem more from their business
model and their willingness to apply technological tools than from the tools
themselves. cloud computing, automation and AI were key technologies that
were emphasised in the interviews and in the examined news publications.
Each of these will be discussed in the following section.
Cloud Computing: The increased use of cloud computing offers many benefits
to law firms and is one explanation to the growth of virtual law firms:
163
The rise of cloud computing in particular has made it possible for attorneys to keep all the
technology tools they need to practice on hand at all times and eschew direct client contact and
office space if so desired. (Broward Daily Business Review, 31st January 2017).
Many firms have particularly developed internal platforms that rely upon
cloud-based solutions. The platforms form the backbone of the virtual firms
and support communication and collaboration. As one of those interviewed
explains:
“…The entire firm is built in the cloud. We log into the platform that forms the
basis for the internal social network; it resembles Twitter and makes it possible
for us to post information on new clients or legal updates. We can see who is
logged in, and in order to communicate with each other, we can just click and
then immediately have a chat or a video conference.”
Automation and AI. Technological developments in AI and automation are the

second main influencer in the legal services industry. There are a number of
AI programs targeted to the legal industry. Examples are ROSS, which is
built upon IBM’s Watson: a software focusing upon AI and machine learning
and Luminence, which is used in due diligence. Having access to AI currently
demands large investments from law firms, since it takes time before the AI
has been trained to work well within different legal areas. Despite the costs,
most of the larger law firms are using resources to develop AI solutions. As
costs decrease and the software becomes easier to use; however, it is likely
that it will become more accessible for more law firms, with far-reaching con-
sequences for the industry. In a Swedish setting, the language will probably
delay this development as it takes additional time for the software to learn to
decipher Swedish legal texts. In 5–10 years’ time though it may be possible
that AI will perform many of the tasks that law firms do today. Some infor-
mants, however, argue that technology would not change the core of what
lawyers do. They claimed new technology will not influence the professional
knowledge, and that the assignments they were given would not change,
although the way in which they were performed would be somewhat differ-
ent. This is illustrated in the following quote:
“Technology won’t replace anyone; it is a tool that makes it possible for us to do

things differently and (be) more efficient. Individual professional expertise
cannot be replaced by technology.”
164
After having seen what tasks AI can perform in relation to searches for previ-
ous legal cases and court decisions, an interviewee expresses his concerns:
“...It may well be that, in the future, one gets sued for misconduct if one doesn’t
use this type of systems when preparing cases.”
Thus, it might be that the use of technology might not only be a source of
effectiveness; it may also become a prerequisite over time. Also, the fact that
outsourcing of law services seems to be of increasing interest to clients, the
application of AI might happen more rapidly than if law firms were to do it
themselves, as illustrated in ABA Journal:
Most importantly, outsourcing providers can use technology that a law firm cannot afford.
That might include artificial intelligence, contract management, process mapping and
workflow technology. ( July 1, 2017)
This takes place not only through innovation in the business models – as is
the case for the virtual firms; it is also related to basic tasks previously per-
formed by junior lawyers. As The Straits Times illustrates:
Artificial intelligence (AI) is shaking up the law and accounting sector, as companies embrace
the use of increasingly smart machines to perform mundane tasks that have traditionally been
the preserve of junior employees. (September 28, 2016)
Machines with equally good or even better results can now can perform
many of the tasks that was once performed by juniors as well as administra-
tive staff in the past: anything from analytical tasks, data gathering or assess-
ments in individual cases. While the use of AI and automation might improve
value creation by helping to reduce costs, for example, it is also believed to
have drawbacks. The application of AI to routine tasks for junior associates
means that they miss important opportunities to learn how to perform more
basic searches and analyses, as well as get experience and become socialised
into important professional values and norms. As one of the informants
describes:
“…. If robots can perform tasks we normally let junior associates do, then I find
it difficult to see how we will develop the knowledge necessary, and also provide
our new employees with norms and values that are fundamental to become a
good lawyer.”
165
THE ORGANISATION OF THE VIRTUAL L AW FIR M

Interviewees described that the virtual law firm differs in many ways from
the traditional law firm. While the physical building is regarded as being
highly important in traditional law firms, - not the least from a marketing
perspective - this is not so in virtual law firms. One interviewee explains:
“Nowadays having a large library or support staff is highly redundant, and I find
it more efficient to do my searches online myself. The physical office has become
irrelevant. We used to need a large office to share resources, libraries, and
administrative staff – in fact, one of the main motivators for becoming a member
in a large law firm was to get access to these resources. In our company, it is very
different, now it is the web site that is the law firm, not a fancy office building.
The large law firms are ignoring these changes.”
The lack of physical presence was even described as an advantage for the
virtual law firms, as it made them more flexible and reduced their fixed costs.
As explained in ABA Journal:
The virtual law firm-an office has no mahogany-walled waiting room, no expensive
downtown location and no expensive overhead? Well, how about ditching the office altogether
and using technology to communicate, share information and service clients? That’s the idea
behind the rise of the virtual law firm… ( June 1, 2015).
The digital technology was also said to enable law firms to develop networks
that enable them to access research and information previously inaccessible
to them. This influences both how law firms can work with better resource
allocation and lower their costs through a cost-efficient organisation. For
instance, the virtual law firm is built upon a network model that connects
dispersed lawyers, allowing them to portray themselves as being part of a
larger firm, communicate with peers, leverage up-to-date technology and
support administrative needs: such as billing and accounting. As the network
model has many advantages, many large law firms are currently experiment-
ing with tying up resources in external networks: for example, via Upwork or
Freelancer in order to access new expertise and complement their internal
resources. While using network-based resources might not change how the
core tasks of lawyers are performed, it certainly affects the context of value
creation within legal services. Those interviewed emphasised that the new
digital technology enables firms to organise their work in new ways. They
166
also described how the key to successfully creating value in law firms has
changed due to the introduction of digital technology:
“If technology changes our practices? But of course! We would not be able to
take advantage of the new possibilities the technology brings unless we changed
the way we organised our work”.
As aforementioned, professional service firms are defined by the professional

autonomous role the employees have in their work. This ideal has been under
pressure in many professions as professional service firms have tried to pro-
fessionalise their management and move toward more business-oriented
organisations in combination with the traditional partnerships. However, the
virtual organisation allows the professionals to maintain their autonomy. As
one interviewee described:
“The effects [of virtual organisations] on the professional autonomy? It improves

it. The technology frees us from hierarchic and bureaucratic structures”.
In a virtual organisation, the structural features are of less importance.

Rather, culture is described as being central for how to organise the firm, the
practices, and for internationalisation. This is also regarded as important
glue, as well as selecting which persons to recruit and retain. Another inter-
viewee explains:
“We aim to build a culture in which respect for each other and knowledge-shar-
ing are important values. A central tool to achieve this is to give your co-workers
as much freedom as possible. An important focus in recruitment is to get on
board competent people with a viable portfolio, but who also want to collaborate
closely with colleagues. This requires both basic IT competence and an extro-
vert personality. We need people who can contribute to building a company by
sharing and collaborating with colleagues. “
THE INFLUENCE OF DIGITAL TECHNOLOGY ON VALUE CRE ATION

The costs and value that is offered to the customer and other stakeholders of the
firm is at the core of value creation. The virtual organisation forms the basis for
the realisation of the value creation potential. Improvements in value creation
relate to how virtual law firms are organised; it is their way of organising, which
167
enables them to save costs. A virtual structure also enables them to deliver bet-
ter services. As The Daily Breeze explains:
We have three major value propositions: top quality work; we’re extremely responsive, and
people can save up to 50 per cent on their legal bills… (March 26, 2015)
This illustrates how the virtual organisation has had a large impact on the
value creation by removing structures and making redundant much of the
traditional resource base such as libraries and support staff.
Reputation was another important feature strongly emphasised by the
informants. While it used to be common to build a reputation through per-
sonal contacts, a new trend among law firms seems to be focusing more on
the internet to build a reputation online. One interviewee explains how the
online presence could create value for the clients, as it gave them quicker and
better access to information:
“Visibility and building reputation on the internet has become very important.
Most well known lawyers now have weekly blogs, or they retrieve information
from other lawyers’ blogs or LinkedIn updates. This increases the speed: you
can now read online about new legal updates every morning.”
In addition to contributing locally with value creation, the virtual organisa-

tion also enables much faster the international scaling of its operations. One
interviewee explains this in the following way:
“Having an international presence is steadily becoming a more important

competitive factor. We hire specialists from all over the world. So we have a
distributed model, which takes advantage of the benefits of having partners in
the cities where our clients work. Our virtual model enables us to establish our
firm internationally much faster. I would say that international expansion
occurs when a client asks for a long-term representation in a specific geographic
area. We saw an example of this in Tel Aviv recently. We could open an office
there in one week’s time. First, we identified a lawyer running his own practice
there, who was interested in being part of a broader international collaboration.
So, we just connected the lawyer to our platform. Geographic distance is not as
important as time zones or languages. For me, only 50% of my clients are located
nearby – 30 years ago that would have been totally impossible.”
Internationalisation has traditionally been associated with high capital com-

mitment limited to large firms. However, the increased digitalization also
168
enables small and medium-size firms to internationalise and utilise “off-the-

shelf” and inexpensive technology to obtain competitive advantages. This
change is important, and emphasises how a clever use of inexpensive new
digital technology in many situations can be more valuable to firms than
merely focusing upon the high end of technological complexity.
Apart from creating value for their clients, virtual law firms also create
value for their employees. The way of organising the work makes it possible
for lawyers to find a better work-life balance: for example, having time for
one’ family, especially with young children, or taking up a hobby. This type
of value is important for the virtual law firms when recruiting highly compe-
tent employees, and it also appeals to potential clients:
“Our employees are our most important resource. Our clients don’t hire the
firm; they hire the people working here. If we are able to recruit the right people,
well, then we will earn good money. So it is important to us to create a good and
attractive place to work. One way of doing so is to encourage flexible work hours
and the possibility of working from home. We want to create a better work-life
balance and make it attractive for our lawyers to work here.”
Looking Ahead: The Future of Professional Service Firms

The study indicates several broader future potential changes in general to
professional and knowledge- based services, particularly to the legal service
industry. A first observation is that the digitalization of the legal industry is
not predominantly driven by increased technological complexity. Rather, it is
concerns how the new digital technology is applied and utilised by law firms
to improve their competitiveness. As we previously discussed, law firms are
adopting and integrating systems based upon AI. The media has discussed
how software with AI might influence how the very core of legal tasks is
performed, thus, fundamentally challenging the industry. The legal industry
is also apparently under strong price pressure, and one strong incentive to
automate tasks is cost reduction. The following quote from a virtual law firm
operating in the US market illustrates this:
The US is a legalistic society, however only 20% of its population can afford legal services.
This is a market ripe for disruption. The ones that manage to provide affordable services at
a quality that suffice – will take the grand prize.
169
Based upon our findings, we suggest that the increased use of automation
and AI in legal services will have the following consequences:
First, automation of time-consuming tasks related to data collection and
analysis can produce organisational slack and pose an opportunity for law
firms to engage in more strategic matters, such as market development and
expansion. Automation might also provide opportunities for lawyers to focus
more upon tasks that machines cannot easily performed – such as representa-
tion in courts or negotiations – as they build upon creativity, human judg-
ment, and empathy. However, this might also lead to changes in the value
creation process in law firms. Law firms have traditionally offered high-
priced bundled services that are billed by the hour. Today, an increasing
number of law firms are seeking to identify tasks that are scalable, which can
benefit from automation. The automation makes it possible for law firms to
unbundle their services, thus, performing them more efficiently. It also opens
up a possibility for competitors to invest in technology as an entry strategy
into the market of legal services. There are currently several examples of
solutions to solve legal issues offered by firms outside the legal industry.
These firms do not have the same regulatory requirements, such as being
restricted by the equity clause: for example, divorce settlements, real-estate
transaction or incorporating a firm are increasingly offered as online services.
Other examples are increased in-house legal expertise, software to support
certain services such as Due Diligence offered by Luminence, and firms such
as Axiom and Burton law providing strategic services to enable more profes-
sional purchasing of legal services. Thus, it is plausible that the legal industry
will experience more intense competition from actors presently not part of the
market for legal services, and that the current core offerings by law firms will
be less relevant in the future. One interviewee addressed this accordingly:
“An analogy could be predictions about self-driving cars and the future of
driving instructors. Acquiring a digital strategy will not suffice for the driving
instructor if the market for driving licenses are drastically reduced. Within the
legal industry there might be several such examples of potential big changes in
marketability of certain services – becoming irrelevant: for example, due to
blockchain technology predicted to revolutionise contracts and the need for
third-party involvement in transactions.”
170
Secondly, the increased use of digital technologies in law firms will create a
need for new types of competences. While law firms have built upon the
lawyers’ strong legal knowledge, they now need to recruit or develop people
with technological competences. In turn, this poses new challenges for law
firms, relating to questions such as: i) How to recruit and retain highly tech-
nologically-competent employees; ii) if new career paths are needed: for
example, different paths for lawyers and technical experts; iii) How the power
balance and status of lawyers and technical experts will play out internally;
and iv) Whether the law firms instead should outsource all technical issues.
In addition, law firms need to develop strategies for how to use the new digital
technology and transform it into new business opportunities. To do so, law
firms might need to complement their legal competence with strategic compe-
tence. This, in turn, opens up for changes in the traditional professional
partner-structure in law firms, and the development of new managerial levels
not necessarily built upon legal competences.
Third, while virtual law firms reveal many advantages, such as being
flexible, agile, having low-fixed costs, and the ability to draw upon expertise
on an ad-hoc basis, they also have disadvantages. One such potential disad-
vantage is the lesser degree of collaboration and knowledge development
among the lawyers in virtual law firms. While law firms have traditionally
played an important role in training through professional apprenticeships,
this might not be as common in the future. The fact that virtual law firms
mainly recruit lawyers with long experience and do not hire or spend time on
training junior lawyers partly indicates this, as does the increased use of
automation and AI. Therefore, the legal service industry needs to ask itself
how future lawyers will be trained and given the opportunity to develop into
experienced and knowledgeable senior lawyers.
Transforming a law firm’s value production process into an automated
services production flow, however, will require investments, and might require
external capital – now prohibited due to the equity clause. This might also
have consequences for the clients to law firms. Using automation and AI will
make it possible for law firms to deliver better and more tailored expertise
faster and easier to their clients. For the clients, this means that they might get
more value for their money. The unbundling of services also makes the val-
ue-creation process more transparent, and easier to compare between law
firms: for example, via online references. The reduced information asymmetry
171
and opacity of the services’ qualities is likely to shift the power balance
between lawyers and clients in the clients’ favour; it will be interesting to see
whether the increased use of automation and AI will lead to the development
of new standards in the legal services industry, or if law firms will create
models that make them retain some of the value.
172
References
Alvesson, M., & Karreman, D. 2006. Professional service firms as collectivities. In R.

Greenwood, & R. Suddaby (Eds.), Research in the Sociology of Organizations: Professional Service
Firms, Vol. 24: 203–230. Oxford: Elsevier.
Breunig, K. J. 2016. Limitless Learning: assessing social media use for global workplace
learning. The Learning Organization, 23(4): 249-270.
Breunig, K. J., Kvålshaugen, R., & Hydle, K. M. 2014. Knowing your boundaries: Integration
opportunities in international professional service firms. Journal of World Business, 49(4):
502-511.
Breunig, K. J., & Skjolsvik, T. 2016. The Virtual Future of Professional Service Firms. Academy
of Management Proceedings, 2016(1).
Christensen, C. M., Wang, D., & van Bever, D. 2013. Consulting on the Cusp of Disruption.
Harvard Business Review(October).
Chui, M., Manyika, J., Bughin, J., Dobbs, R., Roxburgh, C., Sarrazin, H., & Westergren, M.
2012. The social economy: Unlocking value and productivity through social technologies:
1-170: McKinsey Global Institute.
Davenport, T. H., & Kirby, J. 2015. Beyond Automation. Harvard Business Review, 2015( July).
Gioia, D. A., Corley, K. G., & Hamilton, A. L. 2013. Seeking qualitative rigor in inductive
research notes on the gioia methodology. Organizational Research Methods, 16(1): 15–31.
Gordon, T., Shackel, R., & Mark, S. 2012. Regulation of legal services in the e-world: a need to
short circuit hot spots in ethics and novel practices? International Journal of the Legal
Profession, 19(1): 55-87.
Greenwood, R., Li, S. X., Prakash, R., & Deephouse, D. L. 2005. Reputation, diversification,
and organizational explanations of performance in professional service firms Organization
Science, 16(6): 661-673.
Hoopes, D. G., Hadsen, T. L., & Walker, G. 2003. Guest editors’ introduction to the special
issue: Why is there a resource-based view? Toward a theory of competitive heterogeneity.
Strategic Management Journal, 24(10): 889-902.
Jesuthasan, R., Malcolm, T., & Zarkadakis, G. 2016. Automation Will Make Us Rethink
What a “Job” Really Is. Harvard Business Review(October ).
Løwendahl, B. R. 2005. Strategic Management of Professional Service Firms (3rd ed.). Copenhagen:
Copenhagen Business School Press.
Malhotra, N., & Morris, T. 2009. Heterogeneity in professional service firms. Journal of
Management Studies 46(6): 895–922.
Manyika, J., Chui, M., Bughin, J., Bisson, P., & Marrs, A. 2013. Disruptive technologies:
Advances that will transform life, business, and the global economy. San Fransisco, CA,
US: McKinsey Global Institute.
Noy, C. 2008. Sampling Knowledge: The Hermeneutics of Snowball Sampling in Qualitative
Research. International Journal of Social Research Methodology, 11(4): 327–344.
173
Osterwalder, A., & Pigneur, Y. 2010. Business Model Generation: A Handbook for Visionaries,
Game Changers, and Challengers: John Wiley & Sons.
Sawhney, M. 2016. Putting Products into Services. Harvard Business Review, 94(September):
82-89.
Skjølsvik, T., & Voldsund, K. H. 2016. Forretningsforståelse Oslo: Cappelen Damm.
Starbuck, W. H. 1992. Learning by Knowledge-Intensive Firms. Journal of Management Studies,
29(6): 713–740.
Susskind, R., & Susskind, D. 2015. The Future of Professions: How Technology Will Transform the
Work of Human Experts. Oxford, UK: Oxford University Press.
von Nordenflycht, A. 2010. What is a professional service firm? Toward a theory and
taxonomy of knowledge-intensive firms. Academy of Management Review, 35(1): 155–174.
174
CHAP TER 8
Robotisation of Accounting
in Multi-National Companies:
Early Challenges and Links to Strategy
M ARTIN CARLSSON-WALL AND TORKEL STRÖMSTEN
Introduction
Multi-national companies are complex organisations (Busco et al, 2008; Bartlett
and Ghoshal, 2002). This complexity comes in many different forms. A central
aspect is that operations take place in different geographical locations, which
leads to challenges when it comes to controlling at a distance. The ways in
which activities and behaviour are controlled also differ between companies.
Some organisations rely largely upon calculative numbers and control, while
others use a norm-based control regime. In this chapter, we are interested in
exploring how the finance functions in multi-national companies use digital
robots, more commonly referred to as robot process automation.
Robot Process Automation (RPA) is a software that performs administrative
tasks and activities that otherwise humans and knowledge workers (such as busi-
ness controllers) would perform. The reasons why RPA is often referred to as
“robots” is because it can take over relatively simple and routine tasks such as
moving data from one system to another. As Lacity and Wilcocks (2015, p.2) write:
“Knowledge workers consistently tell us they want to be liberated from such

highly-structured, routine, and dreary tasks to focus on more interesting work.
Some are actually getting that wish, thanks to a new approach known as Robot
Process Automation (RPA)”.
As of the autumn of 2017, many companies experiment with various forms of

RPAs. The hopes for robotisation are high since one “robot” (often equal to
one software license) can do work that normally takes several individuals.
175
M A RT I N CA R LSSON-WA LL A N D TOR K EL ST RÖMST EN
Still, fear also exists that skilled jobs will be replaced by RPAs. This might
start with help doing dull high-routine work; however, as the RPAs get better,
we do not know what the future will hold.
Despite the interest in RPAs, there has been surprisingly little discussion about
the pros and cons of robotisation in administrative tasks (Isaksson and Wennberg,
2016). The invisible activities of book keeping and paying bills are seldom given
much attention when researchers and practitioners discuss digital transformation.
To remedy this, we have conducted an exploratory case study of three multi-na-
tional companies and how they invest in RPAs within the finance function1.
Even if the results are preliminary, our empirics indicate that investments
in RPA technologies are related to the companies’ choice of internationalisa-
tion strategy and the challenges that different types of strategies create.
This chapter is organised in the following way: First, we briefly review the
literature on accounting and digital transformation. Even though the term digi-
tal transformation is rather new, it will be shown that current research challenges
can be linked to earlier research on accounting and ERP systems and account-
ing and functional IT. Secondly, focusing upon multi-national companies, we
introduce a framework that discusses three types of internationalisation strate-
gies: ethnocentric, polycentric, and geocentric (Perlmutter, 1969; Hedlund,
1986). These strategies differ depending upon the scope and scale of internation-
alisation and we will use them to explore how the choice of internationalisation
strategy is linked to the investment in RPA technologies. We then present three
illustrative case studies. Through these cases, we propose some tentative prop-
ositions and patterns for how RPAs have been introduced in the finance func-
tion. We end the chapter with a discussion and concluding remarks.
Accounting and Digital Transformation:

Three Distinct Phases
One can distinguish between three distinct phases when reviewing the litera-
ture on accounting and digital transformation. Beginning in the 1970s, we
had a first phase called “accounting and functional IT”: where the digital
1 In line with an exploratory study, the research process has been highly iterative. Both authors have
had formal and informal discussions with CFOs, controllers, and IT-experts around digital
transformation for a long period of time. The topic of RPAs started to emerge during the spring of
2016. However, it was not until one year later, in 2017, that we started to do more systematic interviews.
This chapter is based upon < total of 30 structured interviews and informal discussions with
individuals from telecom, manufacturing, IT, bank, airline, hotel and forestry industries.
176
ROBOT ISAT ION OF ACCOU N T I NG I N M U LT I-NAT IONA L COMPA N I ES
transformation focused upon developing specific IT programs to keep track

of accounting transactions or to improve the profitability calculations (New-
man and Westrup, 2005). This phase gave the accountants a relatively large
autonomy, since systems were developed and used within the finance func-
tion. As a consequence, there were few tensions and battles with the IT
department (Newman and Westrup, 2005).
Tensions with the IT department grew stronger during the second phase,
which began during the mid 1990s. This was the era of the ERP systems.
These Enterprise Resource Planning systems brought a new type of digital trans-
formation to the foreground: integration (Hyvönen et al, 2008). To remedy
the large archipelago of functional islands, ERP systems focused upon the
integration of different functions within the company to promote a more
process-oriented and cross-functional culture (Quattrone and Hopper, 2005).
Since the era of the ERP systems occurred at the same time, there was a
strong corporate trend toward a shareholder value orientation, and the
finance function became a powerful player. In a sense, one can say that many
Chief Financial Officers (CFOs) became “captains of the ship” and controlled
both the design of the ERP system as well how it should be implemented. In
many stock listed companies, it was not unusual that the IT-function became
a sub-function within global financial organisations.
Table 8.1: The Accounting Literature on Digital Transformation
Accounting and Accounting and Accounting and

Functional IT ERP Systems Digitalization
Primary Time Period 1970s to mid 1990s From mid 1990s to 2010 2010–present
Type of Digital Digitalize specific Develop large Integration of ERP systems
Transformation accounting tasks within the ERP systems with digital robots, artificial
finance function intelligence, and customised
CRM systems
Main Accounting Keep track of accounting Integrate accounting Integrate accounting
Function transactions and do internal information with other information with both
calculations functional systems to internal and external data to
leverage synergies leverage synergies, as well
as question business model
Power Battle With IT Limited accounting and Strong; however, Potentially strong, where IT
IT lived separate lives accounting strengthened could be the winner since
its role since the CFO many accounting tasks could
often became responsible be automated
for the ERP system
177
During the past years, we have seen the emergence of a third phase of digital
transformation (Schäffer and Weber, 2016). This can be seen in the terminol-
ogy. Instead of talking about accounting and IT, we now talk about accounting
and digitalization to highlight how this third phase is even broader: focusing
upon both front-office digitalization targeting customers and new business
models, as well as back-office digitalization with a more traditional focus
upon automation and robotisation. From a finance perspective, this third
phase entails both opportunities and risks (Quattrone, 2016). On the one
hand, line managers have strong needs for controllers to become “trusted
business partners” in order to make sense of big data and use new IT tools.
On the other hand, with new robots and artificial intelligence, there is a large
risk that the finance function will be considerably smaller, in terms of head-
counts. Some IT directors to whom we have spoken even foresee that “finance
will most likely be there, but they will be part of (my) organisation in the
future”. Thus, recent developments are very interesting because we do not
know what the future finance function will look like, what competences it will
have, and where in the organisational hierarchy it will be located.
With this historical background, it is now time to dig deeper into the strat-
egies of multi-national companies. As we highlighted in the introduction, this
is important since we have discovered a pattern where the type of robotisation
implementation seems to connect to the internationalisation strategy.
Internationalisation:
Three Strategies for Multi-National Companies
According to the literature on multi-national companies, there are three fun-
damental strategies for internationalisation: ethnocentric strategy, polycentric
strategy, and geocentric strategy. Howard Perlmutter (1969) coined these
concepts, and conducted research on how multi-national companies organise
them and how the company headquarters controlled the subsidiaries. In a
later article, Gunnar Hedlund (1986) developed the concepts and added a
fourth type: the heterachichal organisation. We will briefly discuss this type in
our concluding section.
Ethnocentric companies are companies with a strong and clear home base. All
companies have first started in a home market and then gradually moved into
new markets to explore and exploit its specific advantages. The relationship
178
between headquarters and the subsidiaries in ethnocentric companies can be

characterised as a hierarchical one. Subsidiaries’ role is to implement the strat-
egy formulated by the headquarters. Following from this, the interdependen-
cies that can be identified in these types of companies are typically sequential.
Often production is conducted in the “home country” and then shipped to the
subsidiary that is responsible for sales activities. The forms of control used in
these companies are characterised by a control that is strongly linked to the
parent or home country unit. Since the company originates from a home base,
the control is often classified as normative or cultural based. Basically, the
company is managed by “home country people”. In addition, the technologies
in these companies allow for a more calculative and diagnostic type of control.
Polycentric companies are characterised by their independence. These compa-
nies have subsidiaries in many foreign markets and, due to the distance as
well their operations, will be independent in relation to other parts of the
company. Hence, the term multi-national: there are multiple units that inde-
pendent from each other; therefore not much interaction or transactions takes
place between the units. To a large extent, the role of the headquarter will be
to create a sense of an organisational belonging. The issue of keeping some
activities standardised to draw upon the potential economies of scale will be
important; however, the units will continuously challenge this, as they strive
for even more independence as long as they do not see the value in standard-
isation and centralisation.
Lastly, geocentric companies are more complex than both ethnocentric and
polycentric companies. These companies try to combine the local intimacy
with global presence, which means the relationship between headquarters
and subsidiaries will be more complex. An important idea with geocentric
companies is to develop “centres of excellence” that take advantage of the
scale, so less double work is conducted in the group. Hence, there will be, or
can be, many centres as well as internal transactions in these types of compa-
nies. The forms of control that are exercised in the geocentric companies are
often normative and coercive controls.
In the next section, we present our illustrative cases. The focus will be to
describe how companies with different internationalisation strategies engage
with investment in RPA technologies.
179
Illustrative Cases
FORESTCO: ROBOTISATION FOCUSING UPON KEEPING TR ACK OF SAL ARIES

Our first case company is ForestCo. During the past years, ForestCo’s finance
function has received new requirements to become more efficient; “more
lean”, as one of the interviewees put it. ForestCo has five production units in
Sweden and multiple sales units around the world. The company ships prod-
ucts internationally to the different units, where sales offices are located.
Technical problems in printing facilities or converting processes can arise in
the forest and pulp and paper industries. When this happens, it is appreciated
if the supplier can assist. The organisational backdrop in this case is that
headquarters decides and the subsidiary follows.
ForestCo has decided to do a pilot with RPAs for two administrative pro-
cesses within the finance function. The first is foreign payment. This process
is administrative-heavy and requires detailed information to be accurate.
Hence, ForestCo wants to robotise the task of getting the bank account right
and the right IBAN number.
The second process relates to salaries. Both the actual payment as well as
the automatised process should create information that shows deviations that
management can identify and act upon easier than before. The subsidiaries
are often sales offices where administrative processes with customers are
handled. Salaries are the main cost element in the subsidiaries. To prevent
un-authorised employment and/or unplanned cost increases, the automatisa-
tion of salaries will make this process even more transparent than before.
The IT organisation was the department that had the initial say in this
robotisation process. They soon realised, however, that this required new
thinking and problem solving. For example, one issue that emerged was
access to the IT systems. Who should have access? Suddenly, outside partners
had access to ForestCo’s internal IT system. As soon as this was identified,
security issues started to be discussed in the company. The need to identify
and define new areas of accountability was also something that came out of
the process. As one employee said: “The IT organisation needs an internal
counterpart that is not present today. Just like there are systems for HR and
production, there is a need for a finance system that IT can speak to and
where questions are channelled.”
180
Another consequence of the robotisation of accounting within ForestCo

was the identification of new roles and professions. ForestCo had centralised
its transactional accounting to a shared service centre. However, it became
clear during the robotisation project that there was a need for someone who
was responsible for robotisation or “someone who owned the question”, as an inter-
viewee described it.
Another challenge that emerged was that employees felt uneasy about
becoming redundant. Management had to spend time explaining that this
was of no concern to them; no one would be laid off due to the robotisation
project. One director said: “What will this lead to? In the end, I am not sure
that we will be fewer people working here. But I am rather sure that we will
have to replace people, we will need other types of competences.”
In relation to the process to implement robotisation in ForestCo, the general
impression from the company is that this is a project that has taken longer time
than expected. The main reason has been that ForestCo has not specified the
processes that it wanted to automatise. “Our processes were not described
specific enough”, one ForestCo manager said. The implementation of RPAs,
therefore, has started to problematise the processes rather than just make them
more efficient, which was the initial ambition when the pilot projects began.
M ANUCO: ROBOTISATION FOCUSING UPON ACCOUNTS PAYABLE

ManuCo is our second case company. ManuCo has operations in 170 coun-
tries and is organised in separate divisions, each with its own financial
responsibility and independence. Due to the strong decentralisation and
independence, there are very few internal transactions between the different
divisions. Headquarters provide some infrastructure for the divisions: such
as Global HR, finance, and marketing.
ManuCo has initiated a RPA pilot project that is located in one of the
company’s units in Warzaw, Poland. This project takes place on the corporate
level within ManuCo. The project background is organisational. ManuCo
initiated a Shared Service solution in 2000 within the company. All units that
were conducting transactional accounting services were then centralised into
the Shared Service Center in Warzaw. The operations were later outsourced
to a third party in India.
181
Indian supplier, ManuCo’s Share Service

payments of Centre in Warzaw, Poland
accounts payable
Operational work flow

Problems
ManuCo factory in Boston, USA
American Customers,
utility supplier all around
of electricity the world
Figure 8.1: Illustration of problems with Indian supplier and American factory.
The robotisation of the accounting transactions focuses upon the company’s

accounts payable. This might seem like an easy and non-strategic activity;
however, ManuCo’s experience with the outsourced Shared Service operation
in India ended in situation that turned out to be very strategic. One day when
management in Boston came to the production facility, it was closed and no
lights were on. Nothing worked and the facilities looked abandoned. The
Shared Service supplier in India had ultimately failed to pay the bills to the
electricity company and had not done so even when the utility company
repeatedly had tried to make ManuCo pay the bills.
In the end, electricity was shut down and ManuCo’s unit could not run its
operations.
The robotisation project of the accounts payable should be seen in the light
of this. The plan is now to bring the outsourced transaction services in-house
again. Even “simple” transactions could be “strategic”; therefore, the accounts
payable will probably follow the decentralised structure of the company and
be performed in the different divisions and subsidiaries. This makes this case
interesting; not even very standardised transactions can be handed over to
the outsourced party since, if they are badly managed, far-reaching and stra-
tegic consequences can result.
Furthermore, ManuCo runs a very decentralised business where the divi-
sions make decisions independently and with a strong financial accountabil-
ity. Hence, the idea with the RPAs is to bring them back into the divisions, so
that decentralisation also concerns the transactional accounting activities.
182
Therefore, the introduction of RPAs seems to facilitate the idea with decen-
tralised decision-making and accountability in ManuCo.
GLOBAL: ROBOTISATION FOCUSING UPON INTER-COMPANY TR ANSACTIONS

Global is one of Sweden’s largest companies with operation all over the world.
The company is active in an industry that was once heavily regulated; how-
ever, a wave of deregulation since the 1990s has changed the business land-
scape for Global and the way its business is conducted.
Global’s focus when it comes to RPA implementation has been what is
known as “inter-company processes”. Since there are significant interdepen-
dences between the different units within Global, the internal transactions
must be handled with care. The internal transactions amount to over 10
million transactions on an annual basis, which can and, in fact often do, cre-
ate a headache for management. The reason for this is that there are often
mismatches between what an internal unit book as sales and what an internal
buyer books as purchased items. For example, the seller invoice can state that
goods for 100 items have been shipped. The buyer, on the other hand,
receives goods for 80, thus, creating a variance of 20. This happens because
of different systems in the buying and selling units, which creates unbalances
in Global’s accounting system. The local unit’s balance sheet and income
statement differ in relation to Global’s financial reports. Solving these unbal-
ances takes a lot of time and energy. The solution is a robot that looks at the
purchasing order that is sent to the other local unit(s) within Global. The
purchasing order is cleared in relation to the selling order and ensures that the
same amount is ordered and shipped. This will lead to no variances and the
orders are cleared on the appropriate level. Payment is accurate and the bal-
ance sheet and income statement make sense.
The challenges so far are that Global’s local units use different ERP sys-
tems, even if the main ERP supplier is the same in the whole company. All the
local units have different codes and, in effect, different ERP systems. Another
challenge is to give the robot access to the local units’ accounting system.
When this works, the amount of manual work in the local units will signifi-
cantly decrease. Head count will decrease as will the risk for errors. For
example, sales and the cost of sales will be accounted for in the same year:
something that previously was far from clear.
183
Problems can also occur due to local tax legislation. In some cases, coun-
tries have very strict laws and regulation in relation to cross-country pay-
ments, especially when the goods that have been bought are an IT-related
service. This happens on a regular basis since Global’s units in the different
countries are dependent upon IT services from another country.
Global has had two Shared Service organisations: one unit has been
responsible for accounts payable and another unit for accounts receivables.
These units have been working with the financial accounting for external
stakeholders; however, the internal accounting has had the same type of
organisation. As a consequence of the robotisation, the internal accounts
payable and receivables will be moved to the same unit. The external
accounting does not have to match in the same way, as the internal transac-
tions need to do, hence, the merger of these transactions.
Discussion and Concluding Remarks

This chapter has explored the use of Robot Process Automation (RPA) in the
finance function of multi-national companies. The motive behind the chapter
is the relatively silent role that digital robots play in the administrative pro-
cesses of multi-national companies. During the interviews that we conducted,
a pattern emerged that indicated the internationalisation strategy of the com-
pany influenced the initial projects that the companies had pursued in rela-
tion to robotisation.
First, ForestCo illustrated a company with an ethnocentric strategy. Ethnocen-
tric companies often have a clear home base and internally sell their products
to sales organisations around different geographical markets. In the case, it
was possible to see how ForestCo’s finance organisation prioritised to auto-
mate salary process in order to streamline this process. A complementary
reason was also to keep track of how salaries develop over time and to iden-
tify deviations in the subsidiaries early on. Therefore, the control of subsid-
iaries increased with RPAs; it will be easier to take action if salaries increase
for some reason. This control goes hand in hand with the way ethnocentric
companies are controlled through a hierarchical governance mode, which
also can be classified as coercive.
In our second case, ManuCo illustrated a company with a polycentric strategy.
In this case, the accounts payable RPA pilot was initiated on a corporate level,
yet concentrated on shared activities and resources. Initially, accounts payable
184
was seen as non-strategic and was outsourced to external parties. However,

even a seemingly simple activity created problems. More specifically, due to
inadequate control, the Indian supplier did not pay the bills for the factory in
Boston. After repeated calls from the utility company, the electricity was shut
down and a costly production grinded to a halt. To handle this problem,
ManuCo wanted to regain control and in-sourced the accounts payable. How-
ever, it decided to experiment with RPAs to ensure a cost-efficient process.
This was also a robotisation initiative that made sense from the perspective
that ManuCo’s divisions were responsible for driving its own digitalization
projects. Those projects where the corporate level is involved must be related
to less strategic issues for the divisions. As the outsourcing also showed, how-
ever, even less important activities – such as account payables – might inter-
rupt and harm important activities, thus, creating problems in critical cus-
tomer relationships.
Table 8.2: Summary of Illustrative Cases
ForestCo ManuCo Global

Internationalisation Ethnocentric Polycentric Geocentric
Strategy
Type of Accounting Keeping track of salaries Accounts payable Inter-company
Function Being transactions
Robotised
Problem/Solution Quickly identifying deviation From centralised to decentra- Managing the internal
in salaries in order to lised handling of accounts market through robots in
maintain the headquarter- payable is facilitated by the a more efficient way
subsidiary control use of robots
The Global case shows how a multi-national company with a geocentric strategy
needs to manage inter-company processes. The amounts of internal transac-
tions make this a natural first step for Global, and the amount of money and
energy of the employees that can be saved are substantial. In this case, we
could see that the RPAs reinforce this organisational structure and make the
structure run much more smoothly with less tension than before.
To conclude, these illustrative cases have shown that the organisational
strategy of multi-national companies tends to influence their RPA initiatives.
In ethnocentric companies, such as ForestCo, we can see that the robotisation
enforces the hierarchical control over the subsidiaries, while in the ManuCo
185
case, the corporate level initiative to run the accounts payable through a robot
will lead in the end to every division having a robot that can be independently
used in relation to the specific needs of that division. Lastly, geocentric com-
panies, where Global is a very good example, must handle internal transac-
tions in a smooth way. Here, robots will play a key role in order to create an
efficient internal market.
If we go beyond the issue of the robotisation of accounting, it is interesting
to return to Table 8.1 and the third phase of “accounting and digitalization”.
As we described, when digitalization focuses upon both front-end and back-
end digitalization, it is unclear how the finance function will develop in the
future. On the one hand, an optimistic scenario can be that accountants and
controllers will become trusted partners in a world of big data. In this scenario,
robots and artificial intelligence are important tools to complement the neces-
sary judgment that most likely is needed to make reasonable business deci-
sions. As was illustrated with the ManuCo case, one never knows when a
seemingly standardised activity becomes strategic. On the other hand, we can
also see a pessimistic scenario where robots and artificial intelligence becomes
substitutes for accountants and controllers. In this scenario, IT directors or
perhaps “the Corporate Digital Office” becomes the king/queen. Our talks
with IT directors have shown that many seem to believe it is now possible to
regain the power that was lost during the ERP-system era. However, the cases
– especially the ForestCo case – seem to indicate that some potential organisa-
tional tensions arise between the IT department and the finance function.
RPA is unchartered territory in some multi-nationals, and it is not clear where
the responsibility of the processes will reside.
For us, as scholars in accounting and financial management, an interesting
path to follow is certainly the role of the business controller in this new era.
What role will the controller play in the digital world? Schäffer and Weber
(2016) point out some interesting and potential important areas to follow. The
controller community has long aimed to be more of business partners, rather
than only “bean counters”. In order to be a business partner in the new digital
landscape, the digital business model must be translated so it corresponds to
the business model of the company; otherwise, the controller risks supporting
the wrong processes and, in the end, supporting the wrong decisions as a
function will also become obsolete. More than that, Schäffer and Weber point
out that controllers need to integrate analytics and, thereby, also need to gain
186
new knowledge and skills (such as statistics) in order to become valuable for
the business.
Still, the limits of systems can be problematised. For example, Professor
Paolo Quattrone recently offered a critical view of digitalization in the field of
accounting research, remarking upon the consequence it can have, if we are
not careful (Quattrone, 2016, p.120):
The digital revolution has the opportunity to challenge the tyranny of transparency and this
modern divide because the entire edifice of measurement could potentially be disrupted by a
tweet or an internal e-mail…. If I had to bet on what big data will do for decision-making,
I would say that it will make people take wrong decisions much more quickly than before, with
even less room for the exercise of wisdom beyond the increasing compliance that affects various
realms of decision-making, from finance to risk management.
Quattrone pictures a somewhat dark future. We acknowledge that the infor-

mation will be available to more functions and individuals in large multi-na-
tional companies than ever before that will perhaps lead to issues we cannot
foresee today. One potential path is that digitalization will create even new
organisational forms about which we have little knowledge. One such organ-
isational form could well be the heterarchy that Hedlund (1986) sketched out
in an analytical and conceptual article in 1986, where he argued that there
would be no clear centre in hierarchies and that the organisation would be
comprised with nodes creating a network. In fact, we can see that the digital-
ization and the RPAs that we have encountered in this chapter might well
create the opportunities to go in a heterarchical path. As we could see in one
of our cases, is not clear cut that the transactional activities should be centra-
lised to a centre and that the processed information will be sent out to a the
organisational periphery (Robson 1992). Instead, digitalization and RPAs will
allow for an organisational structure that resembles heterarchies: where
power and information (knowledge) can be found more than ever before on
the periphery of large multi-national companies.
187
References
Bartlett, C. A., & Ghoshal, S. (2002). Managing Across Borders: The Transnational Solution.
Harvard Business Press, Boston, MA.
Busco, C., Giovannoni, E., & Scapens, R. W. (2008). Managing the tensions in integrating
global organisations: The role of performance management systems. Management Accounting
Research, 19(2), 103-125.
Chandler, A. D. (1966). Strategy and Structure: Chapters in the History of the Industrial Enterprise.
MIT Press, Cambridge, MA.
Deloitte (2015). The Robots are coming.
Hedlund, G. (1986). The hypermodern MNC – A heterarchy? Human Resource
Management, 25(1), 9–35.
Hyvönen, T., Järvinen, J., & Pellinen, J. (2008). A virtual integration—The management
control system in a multinational enterprise. Management Accounting Research, 19(1), 45–61.
Isaksson, D., & Wennberg, K. (2016). Digitalization and collective value creation. In
Bergström, A., & Wennberg, K. Machines, Jobs and equality. Technological Change and Labor
Markets in Europe. European Liberal Forum, Brussels, Belgium.
McKinsey & Co. (2016). Digital globalization: The new era of global flows. March 2016.
McKinsey Global Institute.
Newman, M., & Westrup, C. (2005). Making ERPs work: accountants and the introduction
of ERP systems. European Journal of Information Systems, 14(3), 258–272.
Perlmutter, H. V. (1969). The tortuous evolution of the multinational corporation. Columbia
Journal of World Business, 4(1), 9–18.
Robson, K. (1992). Accounting numbers as “inscription”: Action at a distance and the
development of accounting. Accounting, Organizations, and Society, 17(7), 685–708.
Schäffer, U., & Weber, J. (2016). Digitalization will radically change controlling as we know it».
WHU Controlling & Management Review, 60(6), 6–17.
Quattrone, P., & Hopper, T. (2005). A ‘time–space odyssey’: management control systems in
two multinational organisations. Accounting, Organizations, and Society, 30(7), 735–764.
Quattrone, P. (2016). Management accounting goes digital: Will the move make it wiser?
Management Accounting Research, 31, 118–122.
188
CHAP TER 9
Uncertainty and Complexity

in Predictions From Big Data:
Why Managerial Heuristics Will
Survive Datafication
GUSTAV AL MQVIST
Introduction
Target Corporation is the second largest discount store retailer in the U.S.
The Company belongs to an industry that constantly craves for more cus-
tomer information in order to improve marketing effectiveness. To reach this
goal as efficiently as possible, merely knowing the past or present is not
enough. Data must also enable them to predict1 the future. When reports
broke that Target had managed to use the consumption history of one of its
customers, a high school girl, to infer her pregnancy (well before her father
knew) it, indeed, stirred some controversy (Sanders, 2014). However, as The
New York Times Magazine emphasised at the time, it actually came as little
surprise to marketers with an interest in predictive analytics. Amazon has
since patented anticipatory shipping, to utilise similar abilities on a large scale
within its logistics.
Digitalization enables today’s marketing professionals to know their digital
customers better than ever before. What this really means is that they possess
more extensive information about their past and present behaviour. By logical
extension, most marketers believe that these Big Data will contribute to the
realisation of commercial objectives (Erevelles, Fukawa & Swayne, 2016).
The purpose of this chapter is to go beyond the data themselves to illus-
trate mechanisms in the predictions they generate. The dual function of all
1 Prediction and forecasting will be used interchangeably throughout the chapter.
189
GUSTAV A LMQV IST
prediction is to monitor forecast quality and improve decision-making (Win-

kler & Murphy, 1968). This chapter will address both of these aspects, build-
ing upon the forecasting and judgment and decision-making ( JDM) litera-
tures, to present a non-technical investigation of the challenges inherent in
forecasts from big data.
The Golden Rule of forecasting advocates sophisticated prediction models
that incorporate all available information (Armstrong, Green & Graefe, 2015).
All else being equal, this assumes that having access to more data also lead to
better predictions. And yet, at the same time, there is evidence that ignoring
part of the information can improve forecasts (Hogarth & Makridakis, 1981;
Soyer & Hogarth, 2015) and that simple prediction models often do well
(Makridakis & Hibon, 2000). Understanding this paradox will nuance the
expectations placed upon predictions from big data. This chapter will argue
that forecasts require a good match between the uncertainty of the environ-
ment and the complexity of the prediction model.
What constitutes a good prediction? What is meant by big data? Which
dynamics determine the predictability of the future? And, are extensive infor-
mation and sophisticated models really pre-requisites for businesses to antici-
pate the behaviour of their digital customers? These and similar questions
will be answered in the following. The chapter will begin with some com-
ments on prediction, continue with a summary of the literature on big data,
then illustrate some mathematical-statistical constraints in forecasting, and
conclude by saying that simple managerial heuristics (businesses’ rules of
thumb) may yet survive – even in this age of Datafication.
Predictions From Big Data
PREDICTION
Predictions estimate future states of the world when the outcomes cannot be
known for sure. Following Hogarth, Lejarraga, and Soyer (2015), prediction
involves two populations or settings. In the first setting, data are analysed and
interpreted, thus, enabling learning (L). This knowledge is projected into the
future in the second setting with the aim of fitting an outcome: the target (T).
Predictions can only be effective to the degree that the information in L and
T match. They must overcome both uncertainty and complexity. The former is
190
U NCERTA I N T Y A N D COMPLEX I T Y I N PR EDICT IONS F ROM BIG DATA
the dispersion of the probability density function of T while the latter are the
interdependencies among the prediction model’s variables (Schoemaker,
2004). As a result, as Nils Bohr, Nobel laureate in physics once said: “Predic-
tion is very difficult, especially if it’s about the future.”
Predictors and Predictands

Take a music streaming service about to recommend a playlist to one of its
users. It cannot perfectly anticipate how it will be received. However, it can
make a qualified guess. Certain cues in the past (having just listened to Oasis),
may have been positively correlated to one criterion (next wanting to hear The
Beatles), yet negatively to another (Blur). Thus, forecasts extrapolate what is
already known into the future and hope that the original relationships still
hold. L and T become linked through predictors (independent variables) of
predictands (dependent variables).
Forecasts can be divided into two classes: deterministic and probabilistic. The
former yield a single value (“AC Milan will win the 2017/2018 Europa League”)
while the latter allocates probabilities to all potential outcomes (“there is a 15
% chance of AC Milan winning the Europa League; there is a 14 % chance of
Arsenal FC winning it,” et cetera).2 Predictands can be binary, categorical, integer,
real-valued or complex variables. For example, a prognosis that Gary Oldman
will win the Oscar for Best Actor in 2018 is a deterministic forecast of a binary
predictand (he either wins it or does not). Table 9.1 below contains some fur-
ther examples.
For predictions to guide decision-making, these distinctions matter. Deter-
ministic forecasts only project the most likely outcome. Hypothetically, the
percentage chance of the most likely outcome can still range between ~ 0 %
(think of the person holding the most tickets in the state lottery) and 100 %.
For example, weather presenters normally provide temperature forecasts as
integers and rain prognoses binary since they do not disclose absolute likeli-
hoods. And yet, the actual probabilities may differ from one day to another;
some days are inherently more uncertain than others. Therefore, accepting a
deterministic forecast at face value always involves a risk, although it is possi-
ble to shift between deterministic and probabilistic forecasts and across pre-
dictands through designated transformation rules (Stephenson, 2002).
2 The probabilities were inferred from a bookmaker’s betting odds (in September 2017) assuming a
margin of 2–7 %.
191
GUSTAV A LMQV IST
Table 9.1. Examples of Predictands
Predictand Scenario
An e-commerce company wants to infer whether those who have recently purchased product A
Binary
also will buy product B.
A TV production company is eager to know what interest the next season of its main series will
Categorical
attract: massive, average or none (quantified categories).
An advertising agency tries to anticipate how many followers a particular social media group
Integer
will have by the end of the week.
An online store hopes to estimate the amount of data traffic its site will experience during the
Real-valued
Christmas sales.
A betting site observes a rapid surge in clicks on a particular team and needs to figure out if
Complex
and how this will affect the eventual distribution of bets.
What is a Good Prediction?

When the public learns about predictions, these are either the especially good
or the particularly poor. Think of the lucky bettors who backed Leicester to
win the 2015/2016 Premier League at 1000–1 odds or the unfortunate pollsters
who were sure that Bremain and Hillary Clinton would win their respective
elections. Meanwhile, the everyday efforts to incrementally improve predic-
tion quality are seldom known.
Prediction quality is assessed through verification. This process systemati-
cally operationalises, measures, and evaluates predictions in order to enable
continuous improvements. Good forecasts are those with high accuracy (low
prediction error). That is, observations (O) and forecasts (F) that correspond.
Evaluations are sensitive to the choice of accuracy measure (Makridakis &
Hibon, 2000). The predominant metric is the Mean Absolute Error (MAE) (see
equation 1), which measures the average prediction error (Murphy, 1993):
(1)
BIG DATA
The digital age has enabled Datafication. Where there were once small but tidy
sets of information, subjected to deductive analyses by humans, there are now
enormous amounts of fragmented data, which computers inductively mine for
192
correlations (Mayer-Schönberger & Cukier, 2013). This shift is driven by big

data. The term Big Data emerged informally in the mid 1990s and would fea-
ture in the odd academic publication a few years later (Diebold, 2012).3
According to Laney’s (2001) Three V’s framework, big data compiles exten-
sive (volume), multi-faceted (variety) data instantaneously (velocity). Indeed, this
imaging is there for all to see. Search engine(s) and social media platforms
generate enormous amounts of data about their users, of numerous types and
origins, and do so in real-time. E-commerce platforms constantly monitor
their customers in order to improve segmentation and customise offerings.
Streaming services log activities to study preferences in greater detail than
ever before. All of the above data are vast, diverse, and immediate.
IBM has since coined another V – veracity – to pinpoint data ambiguity
(Claverie-Berge, 2012; Scroeck et al., 2012). Others have highlighted how the
meaning of data may change (variability), that big data only becomes meaning-
ful once information can be extracted from them (value), and that they must
be presented in an understandable way (visualisation). Accordingly, there are
actually Seven Vs of big data in total (Sivarajah et al., 2017).
The academic literature on big data is still new and mostly discusses data
problems or process challenges related to acquisition, cleansing, aggregation,
analysis or interpretation (Gandomi & Haider, 2015; Sivarajah et al., 2017).
And yet, the key to big data is arguably that they can utilise heterogeneity to
improve segmentation. As samples increase, this enables the study of previ-
ously marginalised sub-groups of the population (Fan, Han & Liu, 2014).
Big data’s predictive potential, however, has attracted the most widespread
interest. Be that Google’s (in) ability to predict flu outbreaks (Lazer, Kennedy,
King & Vespignani, 2014), an airline’s improved estimated time of arrivals
(Brynjolfsson & McAfee, 2012) or a consultancy’s successful forecasts of niche
auto sales (LaRiviere et al., 2016). Allegedly, “It’s a simple formula: Using big
data leads to better predictions, and better predictions yield better decisions”
(Brynjolfsson & McAfee, 2012, p. 65).
Unfortunately, the truth of the matter is more complicated. Firstly, due to the
novelty of big data, there is a deficit of empirical evaluations. Without system-
atic verifications of its predictive accuracy in the real world, it is difficult to
assess how good predictions big data actually enables, especially across tasks.
3 See Ward and Barker (2013) for further definitions.
193
GUSTAV A LMQV IST
Secondly, as highlighted in the introduction, it is well established that simple

prediction models may outperform complex approaches. Meanwhile, predic-
tions from big data remain advanced and employ either regression techniques or
machine learning (Gandomi & Haider, 2015).4 Thirdly, big data implies that the
number of variables and dimensions increase, which in-turn results in higher
complexity. Taken together, these concerns indicate that predictions from big
data remain most difficult. Two explanations for this will be discussed in the
following section.
THE BIAS/ VARIANCE DILEM M A

The first consideration is the bias/variance dilemma (Geman, Bienenstock, &
Doursat, 1992), which relates to prediction uncertainty; it stipulates that there
are three sources of total prediction error. These are bias, variance and a resid-
ual (ε) (see equation 2). Bias is the systematic deviation from the true score
while ε is noise (presumed exogenous). This leaves variance. Here, variance
could be seen as an important Eight V of big data and represents the predic-
tion model’s sensitivity data. Assume that L consists of several samples,
drawn from the population T, and that each sample yields a unique prediction
to fit the true function. Then variance measures the spread over these predic-
tions. An unbiased model could suffer from high variance, or vice-versa. In
practice, there is often a tradeoff between the two.
Total error = Bias + Variance + ε (2)
Regression techniques are superior at minimising bias within any dataset

(in-sample). But this is the easy part. Bias can be reduced from additional
parameters alone. As the mathematician John von Neumann jokingly said:
“With four parameters, I can fit an elephant, and with five I can make him
wiggle his trunk” (see Dyson, 2004). However, similar exercises are suscepti-
ble to noise, which can be confused for signal. Statisticians refer to this prob-
lem as overfitting.
Although predictions should aim to minimise total error, Brighton and
Gigerenzer (2015) still find a tendency to “develop, deploy, or prefer models
that are likely to achieve low bias, while simultaneously paying little or no
4 This chapter will only discuss the former approach. On machine learning, see Hasan, Shamsuddin
and Lopes (2014).
194
attention to models with low variance” (p. 1772). They use a subset of London
temperatures to show the danger of overfitting polynomials to small samples
and how this increases total error in prediction (out-of-sample).
Similar data will be applied in the following section to illustrate the other
side of model complexity – functional form – that can have the same adverse
effect. A dataset was obtained and analysed, courtesy of the Swedish meteo-
rological and Hydrological Institute’s (SMHI). Figure 9.1 contains a 360-tem-
perature observation from Stockholm throughout 2016.
As a first, a linear function along with 2nd (quadratic), 3rd, 4th, 5th and 6th
degree polynomials are fit to the observations from January through Novem-
ber (n = 329). As expected, increasing model complexity gradually explains
more and more of the variance in temperature (in sample).5 As a second, the
same functions are extrapolated to predict (out-of-sample) the December tem-
peratures (n = 31). Suddenly, the quadric function has the highest accuracy.
Neither the linear function nor the more complex polynomials match the
environment equally well. Similarly, although large datasets limit overfitting,
the right degree of model sophistication depends upon the uncertainty of the
environment (Hogarth & Karelaia, 2007). Different functional forms will find
it easier than others to fit some patterns.
20
15
10
°C 0
-5
-10
Observations (Jan-Nov) Observations (Dec) Forecast (linear)
-15
Forecast (2nd degree poly.) Forecast (3rd degree poly.) Forecast (4th degree poly.)
-20 Forecast (5th degree poly.) Forecast (6th degree poly.)
Figure 9.1. Model complexity and prediction accuracy. A linear function and 2nd, 3rd, 4th, 5th and 6th
degree polynomials fit to Stockholm temperatures from January-November 2016 (in grey; n = 329)
and then extrapolated to forecast December 2016 (in black; n = 31). The quadratic function does best.
5 The respective R2 values read .26, .68, .75, .75, .79 and .79.
195
GUSTAV A LMQV IST
THE CURSE OF DIMENSIONALIT Y

The second issue is the curse of dimensionality (Hughes, 1968), which affects
prediction complexity. Big data places high demands upon a prediction
model, as the number of dimensions per sample increases. Assume a large
sample (n) with coordinates throughout an m-dimensional space. For instance,
say the goal is to predict future purchases among a large group of digital
customers about whom extensive information is available. Along with several
underlying parameters, the prediction model stipulates independent variables
as predictors of purchases. As model complexity increases, however, more
correlational relationships involving the independent variables appear. Some
are found for a reason, others merely by chance.
Following Fan, Han and Liu (2014), the accompanying risks include spurious
correlations and incidental endogeneity, both of which violate key assumptions of
regression methods. The former refers to the fact that the more parameters a
prediction model contains, the likelier it becomes that variables incidentally
emerge as correlated within the sample. The latter means that independent
variables could be correlated with ε (thus endogenous). As a result, predictions
from big data require analytical and practical tools for variable selection and
dimension reduction. Whether more data can be leveraged into better fore-
casts will partly depend upon which function they are expected to fill within
the prediction model.
Consider the following related example from finance. The risk of a portfo-
lio depends upon the covariance of the individual assets’ returns. As of
November 2017, there have been 361 companies listed on the Nasdaq Stock-
holm stock exchange.6 Building a portfolio from these technically requires
estimating the entire covariance matrix to maximise the risk-adjusted return.
And these elements can be combined in many ways. Here, this translates into
having to estimate 65,341 covariance parameters. And even the slightest errors
among these estimations could end up having a significant effect upon the
portfolio due to noise accumulation. Again, model complexity comes at a price.
IMPLICATIONS
There are numerous examples of how forecasts are constrained by the bias/
variance dilemma and the curse of dimensionality. Macroeconomists regu-
6 http://www.nasdaqomxnordic.com/aktier/listed-companies/stockholm
196
larly use extensive data for their predictions. And evaluations undertaken by
Sweden’s central bank, the Riksbank, indicate that a relatively simple time
series7 remains as good at predicting inflation as its most refined prediction
model (Adolfson, Andersson, Lindé, Villani & Vredin, 2006; Sveriges Riks-
bank, 2017). The rudimentary Basel I framework of the 1970s still outper-
forms Basel III in predicting financial turmoil; this is also the case when
tested against recent data (Aikman et al., 2014). Moreover, in the aftermath of
the latest financial crisis, Goldman Sachs’ former CFO admitted his complete
surprise at having supposedly witnessed 25 sigma events occur several days
in a row.8 The same dynamics are also known to apply in weather forecasting:
an application of big data where prediction quality has been the focal point for
decades.
This section concludes accordingly that big data forecasts remain suscepti-
ble to prediction uncertainty and complexity. Therefore, big data alone does
not guarantee better predictions. The natural next step must be more system-
atic verifications of big data forecasts’ quality across a range of real-world
tasks. This will require paying equal attention to the prediction models and
their performances as to the original data themselves. There can be more to
the interaction between a specific prediction model and a particular environ-
ment than what first meets the eye.
Managerial Heuristics
Importantly, the alternative against which to evaluate big data predictions are
not naïve models of the coin-flipping kind. Businesses already have extensive
experience in dealing with an uncertain future and apply managerial heuris-
tics doing so (Artinger, Petersen, Gigerenzer & Weibler, 2015). Managerial
heuristics are businesses’ rules of thumb for handling specific tasks in their
environments. In fact, there is evidence that these remain very much up to the
challenge.
In 2008, Wübben and Wangenheimat set out with the ambition to quantify
how much businesses would gain from switching to more refined ways to anal-
yse their customer relationship management (CRM) databases. Ideally, this would
7 Technically a Bayesian vector autoregression (BVAR) model.

8 Even one such event happening by chance is so incredibly unlikely it practically translates into
impossibility; like winning the 1-in-2,500,000 state lottery 21–22 successive times (Dowd, Cotter,
Humphrey & Woods, 2008).
197
GUSTAV A LMQV IST
have enabled better predictions as to whether or not customers remained active

in order to anticipate repurchases. They studied independent sets of CRM data
from companies in three different industries: an apparel retailer, an airline, and
a CD retailer.
The marketing departments had used a straightforward metric – recen-
cy-of-last-purchase together with a cutoff time – whereby a customer who had
not made a purchase for a certain number of months was classified as inac-
tive. Wübben and Wangenheimat (2008) put this heuristic to the test and
compared its predictions to those made by stochastic models that statistically
estimated parameters for customers’ purchases and dropout rates. To their
surprise, they found “no clear evidence for the superiority of these models for
managerially relevant decisions in customer management compared with
simple methods that our industry partners used”; instead, they concluded,
“the heuristics the firms used worked astonishingly well” (p. 92).
Macroeconomic forecasters have also been found to use intuition and heu-
ristics in their predictions. Additionally, they interpret the task more broadly
than merely as a problem of maximising accuracy (Wennberg & Nykvist, 2007).
Some interpret these and similar findings to be counter-intuitive.
Brynjolfsson and McAfee (2012) would still insist: “that throughout the busi-
ness world today, people rely too much on experience and intuition and not
enough on data.” (pp. 66) But no method, however advanced, can maintain
its superiority in face of conflicting evidence. Because in nature and business
alike, the pursuit is not maximum sophistication; it is fit: effectiveness in an
environment. For as long as prediction remains an empirical issue, manage-
rial heuristics cannot be rejected a priori.
Discussion
In closing, the chapter will briefly discuss what the future might hold for
predictions from big data. In an eschatology of sorts, Mayer-Schönberger and
Cukier (2013) have suggested that theory and inferential statistics will be
replaced, as big data implies N ~ all. Correlation will replace causality and a
paradigmatic shift from reason to association follow. Given enough data,
computers will inductively figure out the world by themselves.
However, big data does not make theory redundant. On the contrary, had
it not been for common sense, Google’s algorithm would have had people
believe that high school basketball games predicted flu outbreaks, as their
198
seasonality happened to be similar (Lazer et al, 2014). Going forward, this

divide will surely cause more debate.
Some consider big data to be a new computational and statistical paradigm
(Fan, Han & Liu, 2014); it is fair to assume that the unique characteristics of big
data will see to it that certain methods are refined while others are replaced.
Tools for analysing high dimensionality data will likely be in demand.
To end on a pessimistic note, the future will tell whether or not big data
spirals into an unfortunate man versus machine scenario. In the meantime,
ethical issues remain a concern. Ensuring personal integrity and protecting
citizens’ interests are critical challenges for societies and democratic institu-
tions (Helbing et al., 2017).
Conclusions
The following recommendations are, hereby, provided for businesses consid-
ering using big data for forecasts:
ÜTo distinguish between the learning and target population

ÜTo define and operationalise predictors and predictands
ÜTo recognise the difference between deterministic and probabilistic
forecasts
ÜTo monitor prediction quality through a systematic verification process
ÜTo beware of the bias/variance dilemma and the curse of dimensionality
ÜTo evaluate prediction models competitively (out-of-sample)
ÜTo avoid overly complex models
Nonetheless, their ability to predict the future will be anything but certain.
199
GUSTAV A LMQV IST
References
Adolfson, M., Andersson, M. K., Lindé, J., Villani, M., & Vredin, A. (2006). Modern Forecasting
Models in Action: Improving Macroeconomic Analyses at Central Banks (No. 188). Sveriges Riksbank.
Aikman, D., Galesic, M., Gigerenzer, G., Kapadia, S., Katsikopoulos, K., Kothiyal, A., ... &
Neumann, T. (2014). Financial Stability Paper No 28: Taking uncertainty seriously-simplicity versus
complexity in financial regulation (No. 28). Bank of England.
Armstrong, J. S., Green, K. C., & Graefe, A. (2015). Golden rule of forecasting: Be
conservative. Journal of Business Research, 68(8), 1717–1731.
Artinger, F., Petersen, M., Gigerenzer, G., & Weibler, J. (2015). Heuristics as adaptive decision
strategies in management. Journal of Organizational Behavior, 36.
Brighton, H., & Gigerenzer, G. (2015). The bias bias. Journal of Business Research, 68(8),
1772–1784.
Claverie-Berge, I. (2012, March 13). Solutions Big Data IBM. [Presentation]. Retrieved from
https://www-05.ibm.com/fr/events/netezzaDM_2012/Solutions_Big_Data.pdf
Diebold, F. (2012). A Personal Perspective on the Origin (s) and Development of “Big Data”: The
Phenomenon, the Term, and the Discipline, Second Version. Penn Institute for Economic Research,
Department of Economics, University of Pennsylvania.
Dowd, K., Cotter, J., Humphrey, C., & Woods, M. (2008). How unlucky is 25-sigma?
The Journal of Portfolio Management, 34(4), 76–80.
Dyson, F. (2004). A meeting with Enrico Fermi. Nature, 427(6972), 297–298.
Erevelles, S., Fukawa, N., & Swayne, L. (2016). Big Data consumer analytics and the
transformation of marketing. Journal of Business Research, 69(2), 897–904.
Fan, J., Han, F., & Liu, H. (2014). Challenges of big data analysis. National Science Review, 1(2),
293–314.
Gandomi, A., & Haider, M. (2015). Beyond the hype: Big data concepts, methods, and
analytics. International Journal of Information Management, 35(2), 137–144.
Geman, S., Bienenstock, E., & Doursat, R. (1992). Neural networks and the bias/variance
dilemma. Neural computation, 4(1), 1–58.
Hasan, S., Shamsuddin, S. M., & Lopes, N. (2014). Machine Learning Big Data Framework
and Analytics for Big Data Problems. International Journal of Advances in Soft Computing & Its
Applications, 6(2).
Helbing, D., Frey, B. S., Gigerenzer, G., Hafen, E., Hagner, M., Hofstetter, Y., … & Zwitter, A.
(2017, February 25). Will Democracy Survive Big Data and Artificial Intelligence? [Blog
post]. Retrieved from https://www.scientificamerican.com/article/will-democracy-survive-
big-data-and-artificial-intelligence/
Hogarth, R. M., & Karelaia, N. (2007). Heuristic and linear models of judgment: Matching
rules and environments. Psychological Review, 114(3), 733–758.
Hogarth, R. M., Lejarraga, T., & Soyer, E. (2015). The two settings of kind and wicked
learning environments. Current Directions in Psychological Science, 24(5), 379-385.
200
Hogarth, R. M., & Makridakis, S. (1981). Forecasting and planning: An evaluation.

Management Science, 27(2), 115–138.
Hughes, G. (1968). On the mean accuracy of statistical pattern recognizers. IEEE Transactions
on Information Theory, 14(1), 55–63.
Laney, D. (2001, February 6). 3-D Data Management: Controlling Data Volume, Velocity and
Variety. [Research Note]. Retrieved from https://blogs.gartner.com/doug-laney/files/2012/01/
ad949-3D-Data-Management-Controlling-Data-Volume-Velocity-and-Variety.pdf
LaRiviere, J., McAfee, P., Rao, J., Narayana, V. K., & Sun, W. (2016, May 25). Where
predictive analytics is having the biggest impact. [Online article]. Retrieved from
https://hbr.org/2016/05/where-predictive-analytics-is-having-the-biggest-impact
Lazer, D., Kennedy, R., King, G., & Vespignani, A. (2014). The parable of Google Flu: Traps in
big data analysis. Science, 343(6176), 1203–1205.
Makridakis, S., & Hibon, M. (2000). The M3-Competition: results, conclusions and
implications. International journal of forecasting, 16(4), 451–476.
McAfee, A., Brynjolfsson, E., & Davenport, T. H. (2012). Big data: The management
revolution. Harvard Business Review, 90(10), 60–68.
Mayer-Schönberger, V. & Cukier, K. (2013). Big Data: A Revolution That Will Transform How We
Live, Work, and Think. London: John Murray Publishers.
Murphy, A. H. (1993). What is a good forecast? An essay on the nature of goodness in weather
forecasting. Weather and Forecasting, 8(2), 281–293.
Sanders, N. R. (2014). Big data driven supply chain management: A framework for implementing
analytics and turning information into intelligence. Pearson Education.
Schoemaker, P. J. (2004). Forecasting and scenario planning: the challenges of uncertainty and
complexity. Blackwell Handbook of Judgment and Decision Making, 274–296.
Schroeck, M., Shockley, R., Smart, J., Romero-Morales, D., & Tufano, P. (2012). Analytics:
The real-world use of big data. IBM Global Business Services, 1–20.
Sivarajah, U., Kamal, M. M., Irani, Z., & Weerakkody, V. (2017). Critical analysis of Big Data
challenges and analytical methods. Journal of Business Research, 70, 263–286.
Soyer, E., & Hogarth, R. M. (2015). The golden rule of forecasting: objections, refinements,
and enhancements. Journal of Business Research, 68(8), 1702–1704.
Stephensson, D. B. (2002, October 29). David B. Stephenson’s proposed classification scheme for
forecasts. Retrieved from http://www.cawcr.gov.au/projects/verification/forecast-types.pdf
Sveriges Riksbank (2017, May 2). Utvärdering av Riksbankens prognoser. [Report].
Retrieved from http://www.riksbank.se/Documents/Rapporter/Riksbanksstudie/2017/
rap_riksbanksstudie_170502_sve.pdf
Ward, J. S., & Barker, A. (2013, 20 September). Undefined By Data: A Survey of Big Data
Definitions. Retrieved from http://www.adambarker.org/papers/undefined_by_data_
ward_barker.pdf
Wennberg, K. & Nykvist, B. (2007). The Psychology of Economic Forecasting. Global Business
and Economics Review, 9(2), 211–226.
Winkler, R. L., & Murphy, A. H. (1968). “Good” probability assessors. Journal of Applied
Meteorology, 7(5), 751–758.
Wübben, M., & Wangenheim, F. V. (2008). Instant customer base analysis: Managerial
heuristics often “get it right”. Journal of Marketing, 72(3), 82–93.
201
CHAP TER 10
Explaining the Behaviour

of News Consumption
ADA M ÅBONDE
Brave New(s) World

The way in which people find and consume news1 and information about the
world is changing. As the news industry is facing a textbook example of tech-
nological disruption (Küng, 2015), both scholars and industry have been very
focused upon what is being published (Anand, 2016) in their quest to under-
stand and adapt to the new rules of the game. Another point of departure is
suggested in this chapter: I ask why news consumers are changing their beha-
viour. I adopt a behavioural perspective when looking to explain news con-
sumption, thus, arguing that although content is important, it is even more
critical to understand the functions of consuming certain content.
News publishing has been traditionally associated with large fixed costs
(Anand, 2016). The costs of producing and distributing content today has
greatly diminished and there has been an explosion in the number of news
publishers and the amount of content created; fragmented media consump-
tion is a result (Küng, 2015). The internet, smartphones, and social media
have all had their profound impact. Indeed, with social media, people have,
in a sense, become their own media producers.
In the digital era, money is not the primary currency anymore. Rather,
consumers’ time and focus are what matters (Goldhaber, 1997; Simon, 1971;
Wu, 2017), since this is what brings in ad revenue (Anand, 2016). With the
digitalization of media, news must now compete for consumer’s attention with
virtually all other kinds of information (“How the world was trolled,” 2017):
1 The American Press Institute defines news as “that part of communication that keeps us informed of
the changing events, issues, and characters in the world outside” (Dean, 2013).
203
A DA M Å BON DE
both from other news sources and from completely new competitors provid-
ing other kinds of content (“Funnel Vision,” 2017).
While some news publishers are succeeding in the digital world; others are
not. Various online, mobile, and social media sources are on the rise (Küng,
2015; Wahlund, Rademaker, Nilsson, & Svahn, 2013); while print subscriptions
continue to fall. Recent figures show that as many as 85% of Swedes consume
news online from time to time: a 66% increase from in 2007 ( Davidsson &
Findahl, 2016). Social media has become an important news source for many
(Müller, Schneiders, & Schäfer, 2016)especially younger, users use Facebook as
their primary source for news about political and societal issues. At the same
time, research suggests that Facebook use contributes to societal knowledge
gaps. Against this background, we investigate the antecedents of using Face-
book as a substitute for other news sources. We argue that exposure to news
posts on Facebook increases the feeling of being well-informed, regardless of
actual knowledge acquisition. This might lead users, especially those with a
low need for cognition (NfC. In Sweden, more than every other Facebook user
(53%) is using the platform as a news source (Davidsson & Thoresson, 2017).
Consuming news is the third most common activity on Facebook, preceded
only by chatting through messenger: of which 81% of Swedish Facebook users
do, and being part of groups: of which 63% are (Davidsson & Thoresson, 2017).
Recently, some news publishers have actually seen an increase in their num-
ber of digital subscribers. For example, both The New York Times and The
Washington Post are now bringing in more money through subscriptions – for
digital and print combined – than they do with advertising (“Funnel Vision,”
2017). Digital sales for news publishers in Sweden are also increasing overall
(both subscriptions and ads), yet are still at very low levels and, in most cases,
are simply not enough to compensate for the loss in print sales, with only one
exception: evening papers (Ohlsson, 2017; Wahlund et al., 2013).
What the future holds for news and journalism is not evident. Neverthe-
less, understanding the opportunities and threats of digitalization, and how
this influences human behaviour (Levitin, 2015) is important not only from a
business perspective (Anand, 2016; Küng, 2015); it is also significant for
democracy (“How the world was trolled,” 2017). While the disruption is
largely driven by digitalization and diffusion of new technologies, per sé, a
behavioural approach provides a theoretical basis for looking at individuals’
contemporary news consumption behaviours.
204
EX PLA I N I NG T H E BEH AV IOU R OF N EWS CONSUMP T ION
I provide a cursory overview of behaviourist notions of learning. I argue

that, as the environment changes, so does the function of behaviours. I fur-
ther present four examples of news and media companies that are viewed as
being successful in the digital sphere and provide a behavioural explanation
for why they are successful. I end the chapter by concluding that news pub-
lishers must move away from thinking solely about what content they are
producing, and begin to primarily focus upon the function the consumption
their products offer their consumers.
***
The remainder of the chapter is organised in three parts. First, behavioural
learning theory is reviewed. Here the role of nature and nurture in human
behaviour in general is briefly presented and how it relates to digital technol-
ogies. Secondly, I discuss the specific behaviour of consuming news. Different
functions of news consumption are considered and examples of news and
media companies that are viewed as being successful in the digital sphere are
presented in light of this. Third, the chapter concludes with a summary out-
lining the main points made and I offer some implications for practitioners.
Human Behaviour and Digitalization
NATURE, NURTURE, AND HUM AN BEHAVIOUR

One’s behaviour is commonly attributed to nature or nurture – or both.
Nature represents inherited capabilities or predispositions that have developed
over many generations through evolution: that is to say, it is embedded in the
genetic code (Dawkins, 2006). Nurture represents the effects of interacting
with the environment: that is to say, one learns from experience (Myers, 2010:
p. 291). Nature and nurture are different sides of the same coin; both represent
the influence of the environment upon one’s behaviour, yet their time horizon
differs: nature over the course of evolution of the species, nurture during the
lifetime of the individual (Skinner, 1974, via Delprato & Midgley, 1992).
The role of nature is quite easy to grasp (at least on a superficial level).
Combinations of genes promote different behaviours. Natural selection
favours combinations of genes that foster behaviours that increase the chance
of survival of the organism (Dawkins, 2006). That is to say, organisms that
behave in ways that increases their survival rate tend to live; organisms that
205
A DA M Å BON DE
engage in less successful behaviours tend to die. As a result, certain combina-

tions of genes – and the related behaviours – have a greater probability to
survive to the next generation (Dawkins, 2006).
As an example, nature could explain why humans react to sudden noises
or movements in our surroundings; the species has evolved to react to novel
stimuli, which is why our attention automatically shifts to new things that are
happening around us (Dawkins, 2006). Evolutionarily, this behaviour is likely
to have been favoured, since it has increased the chance of survival. For
example, by acknowledging sudden movements in the corner of the eye,
individuals could save themselves from dangerous animals lurking in the
bushes. Or by being extra susceptible to strong colors, one might be better
suited to avoid poisonous plants. Of course, today’s world is different. In large
parts of the world, the number of dangers from which humans need to shield
themselves has diminished – or at least they are of another kind.
The impact of nature is straight-forward; the principles of learning are less
intuitive. Humans learn either by association or by observation (Myers, 2010,
p.291-293): learning by doing or learning by watching others do, respectively.
Here, the focus will be upon learning by association: that is to say, doing.
Associative learning is further divided into classical and operant conditioning
(Bouton, 2016; Myers, 2010). Classical conditioning is when an organism
“learn[s] to associate two stimuli and thus to anticipate events” (Myers, 2010,
p.292). The archetypical example of classical conditioning is Pavlov’s dogs: a
dog learns to associate the ringing of a bell (stimuli 1) with food (stimuli 2);
then when the dog hears the bell ring, it begins to salivate (response) regard-
less whether or not there is any food present (it has learned that the bell usu-
ally implies food is coming; therefore, its body starts preparing for digesting it
[Bouton, 2016]).
Operant conditioning is when an organism “learns to associate a response
[…] and its consequence and, thus, to repeat acts that are followed by good
results […] and avoid acts followed by bad results” (Myers, 2010, p.293). When
a behaviour is followed by good results, it is reinforced, which increases the
probability of the behaviour occurring again (Bouton, 2016). On the other
hand, when a behaviour is followed by bad results, it is punished, which
decreases the probability of the behaviour occurring again (Bouton, 2016).
The prototypical example of operant conditioning is known as the Skinner
Box: a rat learns to associate the behaviour of pressing a lever (response) with
206
receiving a food pellet (consequence); and, consequently, presses the lever

over and over again (it has learned that the behaviour of pressing is followed
by a good outcome: that is to say, the behaviour is reinforced (Bouton, 2016).
In theory, learning can explain all human behaviour (indeed, the behaviour
of organisms in general). For example, operant conditioning can explain such
things as why people exercise or eat candy: either behaviour is followed by a
good consequence (positive reinforcement, in terms of an endorphin rush or a
sugar rush, respectively).
According to behavioural learning theory, all behaviours that are not inher-
ited through evolution are caused by external rather than internal factors: that
is to say, within the organism. Changes in the environment are consequently
bound to influence how an organism acts (Bouton, 2016; Delprato & Midgley,
1992; Simon, 1956) it follows quite naturally that widespread digitalization has
had profound impact on the daily lives of millions, if not billions, of people
around the world. Two important digital inventions – the smartphone and
social media – and their impact upon human behaviour are discussed in the
next section, in the light of behavioural learning theory.
THE SLOT M ACHINES IN OUR POCKETS

News consumers currently find themselves in a constant state of distraction.
Concerns have been raised, warning that the digital world is bad both for
individuals (Levitin, 2015) and for society at large (“How the world was
trolled,” 2017). As information becomes ever more abundant, the new scarce
resource has become people’s attention. Companies currently must compete
for consumers’ time and focus, perhaps more than ever before. As a result,
new technologies are often designed with the main goal of making its users
spend as much time as possible with the technology, regardless of what is the
actual value for the user of using the technology (Harris, 2016).
This discussion is not new. In the last fifty or so years, there has been much
debate around the increasing amount of information that digital technologies
bring with them (see Goldhaber, 1997; Simon, 1971; Wu, 2017). What is new,
however, is the mobility and level of engagement of modern technology. More
specifically, the little “Swiss army knife-like appliances” (Levitin, 2015) many
people carry with them at all times: that is to say, smartphones and the
enthralling social networks often accessed through these.
207
A DA M Å BON DE
One can think of the many different reasons why people turn to their
smartphones in an almost compulsive fashion – and people do. On average,
Americans touch their smartphones more than 2,600 times per day: that is,
every 33 seconds (“How the world was trolled,” 2017). For example, it has
been suggested that people are very aversive toward having nothing to do:
that is, to be left with their thoughts (Wilson et al., 2014). Smartphones offer
a simple escape from the discomfort of boredom.2
Smartphones are said to also supplant thinking (Barr, Pennycook, Stolz, &
Fugelsang, 2015; Storm, Stone, & Benjamin, 2016)access to the internet and its
associated knowledge base is at one’s fingertips. What consequences does this
have for human cognition? We frame Smartphone use as an instantiation of
the extended mind – the notion that our cognition goes beyond our brains –
and in so doing, characterize a modern form of cognitive miserliness. Specif-
ically, that people typically forego effortful analytic thinking in lieu of fast
and easy intuition suggests that individuals may allow their Smartphones to
do their thinking for them. Our account predicts that individuals who are
relatively less willing and/or able to engage effortful reasoning processes may
compensate by relying on the internet through their Smartphones. Across
three studies, we find that those who think more intuitively and less analyti-
cally when given reasoning problems were more likely to rely on their Smart-
phones (i.e.; extended mind. A behavioural explanation for this would be that
individuals learn to manage just as well by “outsourcing” their thinking to
their smartphones: rather than having to remember things, they have con-
stant and immediate access to the full body of knowledge on the web.
In both of these examples, escaping boredom and outsourcing thinking
represents behaviours that are reinforced by the environment through oper-
ant conditioning. As users learn that they can get a fresh dopamine kick just
by checking their social media feed one more time (Levitin, 2015), or that they
manage just as well by relying upon Google to retrieve information rather
than trying to remember it themselves (Storm et al., 2016), this is likely to
trigger a self-reinforcing spiral of behaviour. This ensures (or at least increases
2 Attempts have even been made to infer when people are bored based upon their mobile phone usage
patterns (Pielot, Dingler, Pedro, & Oliver, 2015). The implications are straightforward: if a content
producer or advertiser can know when a consumer is bored, they are more likely to grab the
consumer’s attention by pushing content to the personal device at these specific times than when the
consumer has their mind focused upon other things. This would presumably also apply to news, for
which attention is sought.
208
the probability) that people keep coming back again and again. Similar to the
rat in the Skinner Box, smartphone and social media users are lured to pull-
and-refresh only one more time to see whether there is some new piece of
information waiting to be consumed3. This is why smartphones have been
likened to slot machines (Harris, 2016).
With new contexts, comes new behaviour. Consequently, the function of
news consumption inevitably changes, as do the technologies behind it. I will
discuss news consumption and digitalization in the next section.
The Function of Consuming News
FROM CONTENT TO FUNCTION

News publishers have typically been concerned with the content they are
producing, as this has been assumed to attract consumers and, hence, reve-
nues (Anand, 2016). However, going back to nature and nurture, it is difficult
to explain the evolutionary value of news consumption. While behaviors
such as reacting to novel stimuli or being extra vigilant toward negative
(threatening or dangerous) information can be explained by natural selection,
the behaviour of consuming news cannot: for example, under normal circum-
stances, an individual will not be more or less likely to survive or mate as the
result of news reading. Evolution may explain how we read (for example,
focusing upon certain words, colours or features), yet not why we read (for
example, picking up the paper to begin with). Thus, the behaviour of news
consumption must be a learned one.
From a behaviourist perspective, it is possible to imagine several reasons why
an individual learns to consume news. According to the definition provided by
the American Press Institute (see the introduction; Dean, 2013), the function of
consuming news is to stay informed in general. This, however, does not have
any inherent value in itself: it is not evident that there are (directly observable)
positive or negative consequences for an individual that “stays informed”.
The value of staying informed through news can possibly be explained by
cultural practices. As the human species began organising in larger groups
(and eventually societies), there may be value for the individual to keep track
3 The fact that there are not rewards to reap every time users check for new content makes them even
more likely to become “addicted”: Different kinds of irregular reward patterns, called intermittent
reinforcement, have been shown to be even more effective in reproducing behaviour than regular: that
is to say, continuous reinforcement (Bouton, 2016).
209
A DA M Å BON DE
of what has happened: not just in the immediate surroundings, however, in

other parts of the community as well. Individuals are better able to make
good decisions by staying informed about political developments, threats of
war, emergencies, or similar things that may have an effect upon their lives.
At the aggregate, a common basis of information also enables public discus-
sion, which is vital for the proper functioning of a democracy.
An even simpler, and perhaps more accurate, explanation since it is directly
experienced by the individual, would be to have something to discuss in
social situations. Or, rather, to avoid the adverse consequence of being per-
ceived by one’s peers as being uninformed when talking about something (an
example of negative reinforcement: the individual avoids negative feelings/
expectations by consuming the news). In this case, the behaviour of consum-
ing news has a purely social function. The behaviour enables smooth interac-
tion with other members of the species: with colleagues at the coffee machine
or at home by the dinner table.
News may also be consumed simply as a means for passing the time. This
runs the spectrum of escaping boredom (that is to say, avoiding negative
consequences, such as the dreadful “having nothing to do”), to actively seek-
ing entertainment and stimulation (that is to say, pursuing positive conse-
quences, such as reading a witty comic strip).
A counter-argument to the aforementioned functional explanations could
be that behaviour depends upon factors within the individual. Say, for exam-
ple, that an individual is very interested in sports. The argument would then
be that this individual perceives joy or some other kind of satisfaction or
utility, by consuming news about sports. Many other causes of behaviour
residing within the individual, apart from interest, can come to mind as well:
for example, cognitive style, personality, norms, self-enhancement, needs or
a sense of belonging. From a behaviourist perspective, however, these are not
causes of behaviour.
From a behaviourist perspective, news consumption is reinforced due to
cultural practices. However, it has no inherent value for the individual: that
is to say, survival or mating is not contingent upon it. Consuming news is
reinforced by such things as enabling decision-making, facilitating social situa-
tions or helping the individual escape unpleasant feelings or states. Just as
checking social media feeds or looking for new notifications on the smart-
210
phone, however, the behaviour of consuming news does not promote any
other action from the individual apart from continuing to do the same thing.4
TOWARD AN UNDERSTANDING OF CONTEMPOR ARY NEWS CONSUMPTION

Four news and media companies that are viewed as being successful in the
digital sphere are The Economist, BuzzFeed, Quartz, and Vice (Küng, 2015).
These do not only differ in the content they publish; the function of consum-
ing their products also differs.
The Economist has been almost notoriously slow to adapt to the digital
world5; yet it has still managed to stay successful in the face of industry digital-
ization and disruption (Anand, 2016). While the content of The Economist
provides its consumers – often busy individuals with little time to spare – with
a perspective of the world (Anand, 2016), the function of consuming is some-
thing completely different. The Economist have kept their style and form
intact; one function of reading The Economist now is an opportunity to stay
away from the growing problem of information overload. Their content still
provides perspective, yet the function has become the reduction of information.
BuzzFeed and Quartz are both digital news publishers dedicated to under-
standing virality. These companies have outspoken strategies to target young,
smart, and bored-at-work individuals (Küng, 2015). As such, their content is
often clever and enticing, accompanied by headlines bordering on click baits6.
Although they offer interesting or clever news pieces, however, the function
of consuming their content is a few moments of distraction.
Vice also targets a young audience, oozing more sex, drugs, and rock and roll
than does BuzzFeed and Quartz. While they claim to cover important and cur-
rent topics – in principle, anything that their audience wants – they tend to focus
a lot upon adopting its tone and style to appeal to the audience (Vice’s founder
has even claimed that it takes young people to address young people, which is
why the average age of their journalists is only 25 years old [Küng, 2015]).
4 For an interesting discussion on culture and behavior, see Skinner’s (1986)(b paper on how human
behavior has grown weak).
5 Although The Economist now complements its weekly magazine with smartphone and tablet apps,
audio versions of the paper, and are active on social media.
6 Click bait is not one specific type of formatting. Rather, it refers to a number of methods intended to
attract attention and, quite obviously, make news consumers click on published content. Ultimately,
the goal with click baits is to make content travel online: that is, to make it viral. Different click bait
techniques include tweaking content and wording to vex curiosity, encourage interaction, or engage
emotionally (Chen, Conroy, & Rubin, 2015).
211
A DA M Å BON DE
Vice is known for its raw and intimate coverage “behind the scenes” on
controversial topics. One example is when reporters from Vice visited North
Korea with the Harlem Globetrotters: they played basketball against the
North Korean national team, and met with Kim Jong-un (Küng, 2015).
Another example is when a Vice reporter accompanied white nationalists
during the Charlottesville riots in August 2017, interviewing several of the
nationalist leaders, and “provid[ing] viewers with exclusive, up close and
personal access inside the unrest” (Vice News, 2017). With its Gonzo-style
journalism7 the function of Vice is more entertainment and excitement than
it is to inform people.
These news publishers, and others similar to them, have successfully man-
aged to be relevant: not primarily in what they publish, but how they publish
it. Although content is important, it comes in second place. The primary focus
must be upon what is the function of the behaviour of consuming certain news.
By shifting perspective away from content to function, it becomes possible
to understand why people act the way they do. From a content perspective, it
seems unreasonable that anyone would rather consume news on social media
than on a news publisher’s website. From a functional perspective, it makes
more sense: for example, if the function of the behaviour is to escape boredom
rather than to become informed. Many traditional news companies are poorly
fit for the digital world: not because they have poor content, but because they
do not look to the functions of the behaviour of consuming their content.
Conclusion
In the digital era, people do not pay (directly) with money, but rather with
their attention. So, any publisher trying to attract a public online will compete
for the attention of the masses (Goldhaber, 1997; Simon, 1971; Wu, 2017). One
common way to try to do this is by focusing upon what content to publish
(Anand, 2016). In the worst-case scenario, the result is increased sensational-
ism (for example, click baits) and a degradation of quality. Even if a company
succeeds in this new environment, there is always a risk that someone else
will enter the playing field and outmanoeuver them: either because they have
more money, more brains, or better algorithms with which to fight.
7 “Gonzo journalism” is an approach inspired by the practices of Hunter S. Thompson, for example
evident in his book Fear and Loathing in Las Vegas, where the writer departs from the journalistic
aspiration of objectivity and instead combines subjective experiences with fiction (Küng, 2015).
212
Another strategy then, which is perhaps controversial for a news publisher,

is to focus less upon the content that is produced. This does not mean that the
things being published are irrelevant; from a business perspective, it is just as
(if not more) important to also recognise the function of consuming the con-
tent. The behaviour of consuming news has little evolutionary value for
humans. Instead, consumers have learned the behaviour through cultural
practices, which is reinforced by good consequences. These consequences
vary depending upon what function the behaviour serves, and can include
things such as enabling good decision-making, facilitating social interaction
with other individuals, or escaping boredom. By focusing upon function
instead of content, it is possible to explain contemporary news behaviours:
such as why consumers migrate from news publisher websites to social media
platforms to consume news content.
In conclusion, news publishers should shift focus from content: that is to say,
what they are producing and distributing, to function: for example, what is the
purpose of consuming their product. Because digital technologies have
changed the function of consumer behaviour related to news in numerous
different ways.
213
A DA M Å BON DE
References
Anand, B. (2016). The Content Trap: A Strategist’s Guide to Digital Change (1st ed.). New York, NY:
Random House.
Barr, N., Pennycook, G., Stolz, J. A., & Fugelsang, J. A. (2015). The brain in your pocket:
Evidence that Smartphones are used to supplant thinking. Computers in Human Behavior, 48,
473–480. https://doi.org/10.1016/j.chb.2015.02.029
Bouton, M. E. (2016). Learning and Behavior: A Contemporary Synthesis (2nd ed.). Sunderland,
MA: Sinauer Associates, Inc.
Chen, Y., Conroy, N. J., & Rubin, V. L. (2015). Misleading Online Content: Recognizing
Clickbait as “False News.” In In Proceedings of the 2015 ACM on Workshop on Multimodal
Deception Detection. ACM. (pp. 15–19). https://doi.org/10.1145/2823465.2823467
Davidsson, P., & Findahl, O. (2016). Svenskarna och internet 2016: Undersökning om svenskarnas
internetvanor.
Davidsson, P., & Thoresson, A. (2017). Svenskarna och internet 2017: Undersökning om svenskarnas
internetvanor.
Dawkins, R. (2006). The Selfish Gene (30th anniv). New York, NY: Oxford University Press.
Dean, W. (2013). What is the purpose of journalism? Retrieved September 20, 2017, from
https://www.americanpressinstitute.org/journalism-essentials/what-is-journalism/
purpose-journalism/
Delprato, D. J., & Midgley, B. D. (1992). Some fundamentals of B. F. Skinner’s behaviorism.
American Psychologist, 47(11), 1507–1520.
Funnel Vision. (2017). The Economist, 425(43), 61–62.
Goldhaber, M. H. (1997). The attention economy and the net. First Monday, 2(4). https://doi.
org/http://www.firstmonday.org/issues/issue2_4/goldhaber/
Harris, T. (2016). How Technology Hijacks People’s Minds — from a Magician and Google’s
Design Ethicist. Retrieved September 20, 2017, from http://www.tristanharris.com/2016/05/
how-technology-hijacks-peoples-minds - from-a-magician-and-googles-design-ethicist/
How the world was trolled. (2017). The Economist, 425(44), 21–24.
Küng, L. (2015). Innovators in Digital News. Reuters Institute for the Study of Journalism,
University of Oxford & IB Tauris. https://doi.org/10.1080/21670811.2016.1153333
Levitin, D. J. (2015). Why the modern world is bad for your brain. The Guardian, 1–6.
Retrieved from http://www.theguardian.com/science/2015/jan/18/modern-world-bad-
for-brain-daniel-j-levitin-organized-mind-information-overload
Myers, D. G. (2010). Psychology (9th ed.). New York, NY: Worth Publishers.
Müller, P., Schneiders, P., & Schäfer, S. (2016). Appetizer or main dish? Explaining the use of
Facebook news posts as a substitute for other news sources. Computers in Human Behavior, 65,
431–441. https://doi.org/10.1016/j.chb.2016.09.003
Ohlsson, J. (2017). Medieutveckling: Medieekonomi 2017. Myndigheten för press, radio och tv.
214
Pielot, M., Dingler, T., Pedro, J. S., & Oliver, N. (2015). When Attention is not Scarce –
Detecting Boredom from Mobile Phone Usage. UbiComp, 825–836. https://doi.
org/10.1145/2750858.2804252
Simon, H. A. (1956). Rational choice and the structure of the environment. Psychological Review,
63(2), 129–138. https://doi.org/10.1037/h0042769
Simon, H. A. (1971). Designing organizations for an information-rich world. Computers,
Communications, and the Public Interest, 72, 37. https://doi.org/citeulike-article-id:986786
Skinner, B. F. (1974). About Behaviorism. New York, NY: Knopf.
Skinner, B. F. (1986). What is wrong with daily life in the Western world? American Psychologist,
41(5), 568–574. https://doi.org/10.1037/0003-066X.41.5.568
Storm, B. C., Stone, S. M., & Benjamin, A. S. (2016). Using the Internet to access information
inflates future use of the Internet to access other information. Memory, 8211(August), 1–7.
https://doi.org/10.1080/09658211.2016.1210171
Vice News. (2017). Charlottesville: Race and Terror. Retrieved November 15, 2017, from
https://news.vice.com/video/charlottesville-race-and-terror-vice-news-tonight-on-hbo
Wahlund, R., Rademaker, C., Nilsson, P., & Svahn, M. (2013). Mediernas roll i marknads-
kommunikation. In G. Nygren & I. Wadbring (Eds.), På väg mot medievärlden 2020. Lund,
Sweden: Studentlitteratur.
Wilson, T. D., Reinhard, D. a., Westgate, E. C., Gilbert, D. T., Ellerbeck, N., Hahn, C., …
Shaked, A. (2014). Just think: The challenges of the disengaged mind. Science, 345(6192),
75–77. https://doi.org/10.1126/science.1250830
Wu, T. (2017). The Attention Merchants: The Epic Scramble to Get Inside Our Heads. New York, NY:
Vintage.
215
CHAP TER 11
Digital Transformation
Supporting Public Service Innovation:
Business Model Challenges and
Sustainable Development Opportunities
PER ANDERSSON AND L ARS -GUNNAR M AT TSSON
Introduction
Digital transformation is a megatrend as is the widespread concern for sus-
tainable development. In this chapter, we discuss how digital transformation
of public services may promote sustainable development. We take the per-
spective of private business firms that are involved in service innovation
processes, focusing upon their business modelling challenges. To a certain
and varying extent, public services have become “privatised”, yet they are
always subject to publicly decided and enforced norms and regulations. We
assume that public services are provided with the, admittedly complex, “com-
mon good” in mind, which is in line with sustainable development criteria.
Therefore, public service innovation, enabled by digital transformation is an
important phenomenon subject to development and implementation of both
private business models and of public norms and practices. We investigate the
nature of these challenges and opportunities, with the help of four cases on
digital transformation of public services: City Lighting, Public Transporta-
tion, Healthcare, and Education.
The disposition is as follows: first, our analytical framework is presented;
secondly, the four cases follow; third, we discuss opportunities for sustainable
development based upon the cases; lastly, we identify five business-modelling
issues that we suggest should receive increased attention in both practice and
research.
217
PER A N DERSSON A N D LA R S-GU N NA R M AT TSSON
Analytical Framework
Our analytical framework is based primarily upon our recent research on
ICT enabled service innovations, business models, and dynamics of market
networks. (Andersson & Mattsson 2015) This contains three major compo-
nents, which are all related to digital transformation: business modelling,
public service innovation, and market reshaping.
BUSINESS MODELLING
In Andersson and Mattsson (2015), we stressed the importance of develop-
ment and implementation of new business models (Osterwalder et al. 2005)
during digital transformation. In an extensive analysis of the literature on
business models, Zott et al. (2011) found that a business model is a holistic
system-level perspective that includes both content and processes; it is firm-
centric, yet is boundary spanning with a focus upon value creation rather
than value capture. Related to this is how the firm will obtain revenues for its
value-creating activities and how value is calculated. Ehret et al. (2013) and
Mason and Spring (2011) propose a network approach to understand the role
of business models. This is in line with the original use of the concept in
analyses of how new ventures manage to integrate resources by means of
network relationships. Business models of networked firms must, in some
way, overlap or be complementary. A general assumption is that a business
model expresses the business logic of the firm, what value the company offers
to customers and, relating the concept to a business network perspective, the
infrastructure and the architecture of the network of partners. Applying a
network view, La Rocca and Snehota (2017) stress the limits to autonomy of
individual business firms to perform their own business model. They need to
observe and adapt to differences across counterparts.
PUBLIC SERVICE INNOVATION

An innovation is a new way to create value. When it comes to public services,
an innovation should be hopefully evaluated with reference to aspects of
“common wellbeing” and with a focus upon the beneficiaries. Innovations
always require some new combinations of resources, yet should also adapt to
existing resources: especially in a use context. Resources co-create public
service innovations, which are controlled by public and by private actors.
218
DIGI TA L T R A NSFOR M AT ION SU PPORT I NG PU BLIC SERV ICE I N NOVAT ION
A service innovation must find use in socio-economic technical contexts. This

might require added innovative resource combinations or adaptation between
existing and new resources.
Digital transformation tends to stimulate and enable service innovation.
The “service dominant logic” stream of research associated with Stephan
Vargo and Robert Lusch (for example, Vargo and Lusch, 2004 and 2014) has
stimulated research on business models and service innovation (Maglio and
Spohrer, 2013), business models and service orientation (Kindström, 2010)
and digital transformation and service innovation (Araujo and Spring, 2006).
Digital transformation is seen to have effects upon the packaging of the ser-
vice-based value offering, as well as on infrastructures and external partner-
ships. Vargo et al. (2015) use the concept of “service ecosystem” to also include
actors other than market actors.
Little attention in research on service innovation seems to be directed to
public service innovation.
M ARKET RESHAPING
Market processes are embedded in socio-economic contexts characterised by
interdependencies, interaction, and connectedness. Cooperation and compe-
tition co-exist in markets. Innovation is crucial to achieve efficiency and
effectiveness: that is to say, for the creation of value. This means that value is
co-created in the market. The overlapping of networks and intermediation in net-
works will likely change during innovation processes since new resource
combinations develop, as exemplified by converging industries, bundling of
services, technical platforms.
SUSTAINABLE DE VELOPMENT
Concern for unsustainable development is widespread: in terms of negative
ecologic, social, and economic effects of human activities. This is especially
true when it concerns global warming. The UN Agenda 2030 specifies 17
sustainable development goals and some general directions for how to reach
them. Public service innovations aimed at promoting sustainable develop-
ment should be especially important for public actors with an influence upon
public services. The World Economic Forum argues that opportunities to
develop and implement service innovations based upon digital transforma-
tion should be given a high priority (WEF, 2016).
219
To sum up, digital transformation is associated with challenges to develop

viable business models. This chapter will focus upon public service innova-
tion. Service innovations involve market reshaping, especially with shifts in
intermediation and overlapping. These three processes are interconnected
and associated with business model challenges for the involved private busi-
ness firms, thus, influencing the achievement of sustainable development.
Four Cases
The cases all describe early stages of applications of digital transformation;
they include: city lighting, healthcare, public transportation, and education.
The digital transformation challenges are collected from a longitudinal com-
parative research project that contains in-depth studies of companies in 15
different contexts engaged in digital transformation processes, as described
in Chapter 2 of this book. Andersson & Rosenqvist state that digitalization is
associated in many of the studied sectors with overlapping between sectors/
networks, which sometimes includes tension between different industrial
logics. The processes in most cases open up for considerable changes and
re-positioning of companies where business development and business model
challenges become part of the digital transformation.
The discussion here builds upon a broad set of ongoing case studies, and on
secondary sources on digitalization. We employ qualitative methods overall:
specifically, a multiple comparative case-study approach, complemented with
a rich set of secondary sources (reports and so on).
Of the four case narratives, Philips, is based purely upon secondary
sources. Getinge and Nobina were initially part of two master program proj-
ects at the Stockholm School of Economics1 that were performed in 2016.
Complementary information was collected in follow-up interviews within the
digital transformation research project (Digital Transformation in the Net-
worked Society). Sensavis, was initiated in the research program with a set of
interviews in 2016, and completed in a master thesis in 2017.2
1 Rosén, A & Lechner, J. (2016), Advanced Market Monitoring. Sensemaking and Strategic Analysis for
Medical Device Complementors to Big Data Platforms, Business case research thesis, Stockholm
School of Economics, June 2016, Fedrizzi, M. & Liedholm, A. (2016), Nobina and Combined Mobility
Services in Sweden. A scenario-based analysis, Business case research thesis, Stockholm School of
Economics, June 2016,
2 Badhani, S. & Mut, J. (2017), Business Model Innovation in Edtech. An Exploratory Study of Business
Model Innovation in a Complex Market Environment, Master thesis, Stockholm School of Economics.
220
CASE 1 PHILIPS: LIGHTING AS A SERVICE 3

Philips launched their strengthened vision for sustainability in 2012: “At
Philips, we strive to make the world healthier and more sustainable through
innovation. Our goal is to improve the lives of 3 billion people a year by
2025…We have identified three key innovation areas: 1) Improving people’s
lives; 2) Improving energy efficiency of Philips products; and 3) Closing the
materials loop….” The company has embraced a circular economy mindset
over the past five years. The transition toward a circular sustainable economy
in the Lighting sector is intertwined with both digitalization and a radical
business model shift toward service sales: introducing “lighting as a service”
and products as service providers. The business model promotes customer
access to products and systems instead of customer ownership. Value for
customers become performance based. The new concept emerged with a
Dutch architectural company.
“In 2009 Rau Architects, an architectural agency specialising in sustainable
building design, approached Philips for help to upgrade lighting at its Amster-
dam office. They only wanted to buy light, and not the expensive lighting
infrastructure (lamps, luminaires, cables, and controls) that it would eventu-
ally need to replace and dispose. They wanted the exact amount of light for
workspaces and rooms that employees needed, but nothing more. Rau Archi-
tects, Philips and an installation partner, CasSombroek, began a co-creation
process to design a tailor made, intelligent lighting system maximising natural
sunlight, adapting LED light fittings to the building and installing a motion/
daylight sensor and controller system. It proved to be an extremely interesting
experiment in – and proof-point of – how little artificial light an office actually
needs. Since 2010, Rau Architects only pays for the actual amount consumed
light (lux), not the equipment or the raw materials used in the products.” 4
By moving from a one-time sale to a ‘pay per lux’ model, Philips maintains
ownership of the materials while Rau pays for maintenance and servicing with
the option to adapt or upgrade the set-up. The installation of the LED lamp
3 This case builds upon secondary sources from Philips, see e.g.: https://www.philips.com/c-dam/
corporate/about-philips/sustainability/sustainable-planet/circular-economy/light-as-a-service/
case-study-circular-economy-lighting-NUS.pdf, http://www.lighting.philips.com/main/services and
from other sources, e.g.: “Circular economy snapshot: Philips light as a service” (https://www.issuelab.
org/resource/circular-economy-snapshot-philips-light-as-a-service.html) and https://smartcitiesworld.
net/news/news/philips-light-as-a-service-offering-1137,
4 Zero Waste Council: ”Circular Economy Snapshot: Philips Light as a Service”, p.1, see footnote 3.
221
system saved 35% energy. Digital transformation continued in the next phase
when Philips implemented smart energy meters, which provided further
insight into the energy consumption per space. The report states that this mon-
itoring and optimisation process saved another 20%. Thus, more than half the
energy was saved. At the end of the contract period, Philips lighting products
can be taken back into its production process and the raw materials re-used.
Philips argued that the new system provided additional benefits for the
development of smart cities, beyond efficiency and reduced carbon emission.
With lighting levels that can be adjusted and no compromise in light quality,
it can be used to create safer roads and streets, more productive offices, more
liveable cities and attractive public spaces, and more patient-friendly hospitals.
“… The company developed a commercially successful business model with
significant environmental and financial benefits for customers.”(ibid)
Managed lighting services extend the lifetime and performance of the
lighting products. This allows the customer to take full advantage of the
newest lighting solutions, increase energy efficiency, and reduce operational
costs. Philips pays the upfront costs of installation and is compensated
through a performance contract: the energy savings the retrofit produces.
Exploring possibilities from a second-hand market enables Philips to capture
new value from used parts and luminaires; co-creation with like-minded
companies creates a platform for innovation.
Philips also acknowledges barriers in switching to the new orientation.
Lighting is not widely recognised as being the key to energy efficiency. People
do not see the electricity costs associated with lighting. They are unaware of
new energy-efficient lighting technologies.
Philips’ CEO stated: “I don’t want to make this sound easy. In our health-
care business, for example, a lot of customers initially thought: ‘A second-hand
product? We don’t want it.’ Of course, we are refurbishing it and guaranteeing
it as new, but convincing a hospital customer: for example, is challenging and
requires a major educational program. We still have much more to do given
the size of the market, but as we work with hospitals and establish ourselves
as technology partners—and not just sellers of a ‘box’—we can more easily
convince customers of the mutual benefits of circular-economy principles.
Similarly, for municipal-lighting customers, the thinking around the tender-
ing process needs to change. These customers are used to looking at the initial
purchase price, not the total cost of ownership and the ecological impact.
222
Changing the ownership of the lights is also tricky, as it often gets into legis-
lative issues with municipal governments. There are supply challenges in
operating in this new way, as well. We need to get our products back. Street-
lights are fairly simple because the lights don’t walk away, but consumer
lamps are another story. Here we work with partners to organise for collec-
tion, but even then it’s very hard. Currently, in Europe we recover about 40
per cent of our lamps, of which 85 per cent are recycled for reuse.”
Still, the company has started the process of fundamentally redesigning its
business models and value chains. The basic idea is that, instead of selling
products, Philips envisions a future where it retains ownership: selling use as
a service so it can optimise the use of resources. Philips is defining new busi-
ness models, and refining concepts of legal ownership and use, adaptive
logistics, and financing strategies. Working with investment models, and
finding new potential, has become part of the switch. New business models
are developed and tested: utility funding schemes, public private partnership,
or creating a new energy service company. This includes private financing
alternatives such as instalment payment, bank loan, and financial lease, and
also fiscal measures. Public funding issues are brought up including subsidies,
economic stimulus measures, and carbon finance.
The LED and digital transformation is both a technology and business
model transition: going from analogue and regular lamps, stand-alone and
“dumb” products with replacement sales, to digital systems and LEDs that are
connected in “smart” systems and services based upon “projects”. The transi-
tion from analogue to digital is based upon a switch from lighting replacement
products to financing and leasing lighting as a service: “This will reap not only
the direct economic benefits of lighting, but also the benefits beyond lighting
fully in line with the transition from a linear to a circular society.”(pp. 26)
Digital transformation is a central part of the big shift, as the CEO states:
“We are putting networking capabilities in these lights, as well, essentially
making them part of an IT network. This lets the community adjust the lights
depending on the circumstances. For example, if there is low traffic density at
night, then the lights can be turned further down. But if there is a soccer match
one night, the lights can go up. And, of course, we can apply all sorts of algo-
rithms as well to give customers even more control. These kinds of innova-
tions help us move away from selling products and toward selling higher-value
solutions.
223
We believe that customers will increasingly consider natural resources in

their buying decisions and will give preference to companies that show
responsible behaviour: something we are already seeing. Designing products
and services for a circular economy can also bring savings to a company. The
first impression people always have is that it adds costs, but that’s not true. We
find that it drives breakthrough thinking and can generate superior margins.”
CASE 2 NOBINA: MOBILIT Y AS A SERVICE IN PUBLIC TR ANSPORTATION 5

Nobina is an established provider of public transportation in the Nordic coun-
tries. Nobina wishes to change its role from “bus operator” to “mobility pro-
vider”. Digital transformation is said to enable mobility solutions to increas-
ingly pressing urbanisation-related problems, such as traffic congestion and
air pollution.
The target of political authorities in Sweden is to double the share of public
transportation by 2030 in the following ways: 1) Carbon dioxide emissions
will be reduced and energy efficiency improved throughout the entire trans-
port system; 2) Road safety will be improved; 3) Traffic congestion will be
reduced; and 4) More equal opportunities for citizens will be offered by the
public traffic system.
Discussions between private and public organisations on mobility as a
service revealed that car-operating costs would be saved, and that time delay
and congestion could be reduced. Taking cars off the road also means reduc-
ing the number of traffic accidents. Removing cars from the road would also
generate savings for cities by allowing them to avoid road construction costs.
Transport operators are now able to provide mobility as a service, thanks
to the technological advancements made possible by the Internet of Things.
A widely acknowledged belief in public transportation is that the time has
now come to integrate different types of transportation means, both public
and private, which could offer people a flexible and convenient door-to-door
alternative to private car ownership.
One of Nobina’s representatives described the challenges: “In order to
double the public transport market share, there needs to be a shift in our sec-
tor: from a production culture to a service culture. Through new approaches
5 This is a condensed version of a case presented in Fedrizzi, M. & Liedholm, A. (2016), Nobina and
Combined Mobility Services in Sweden. A scenario-based analysis, business case research thesis,
Stockholm School of Economics: June 2016, and complementary interviews: 2017.
224
and smarter co-operation, we will increase the efficiency of our operations,

extend the range of transport services offered, and raise the quality.” The
main finding in an industry analysis and report was that increasing ride
sharing, bike commuting, and car sharing will result in substantial economic
savings for commuters as well as cities6. Moreover, the diffusion of smart-
phones and apps is what was expected to enable digitalization and Intelligent
Transportation. The spread of digital payment solutions facilitates the pur-
chase of online services and the transactions between individuals. Online
platforms require new digital technologies; however, a lot of the potential
comes also from the analysis of big data.
Nobina and the transportation industry are now looking more and more
towards new solutions for smart mobility. The industry is already moving
towards a greater integration of the services. Some projects are already run-
ning (e.g. SLL-bike sharing). “Samtrafiken”, the association of all the actors of
public transportation in Sweden, plans to include as members also companies
operating with non-public mobility (i.e. car sharing, Uber). However, the
public sector is not eager to bear the whole risk of the implementation of such
new concepts. The public sector will have to rely on the total or partial invest-
ments by private actors. A private actor might even manage the overall inte-
gration of the public/private transportation network including the implemen-
tation and maintenance of a technological platform.
The actual demand for the mobility as a service is still unknown. Since
narrow coverage would undermine the overall usefulness of the new concept,
the geographical reach is a fundamental service feature. Innovation propelled
by the public-sector will likely mean that public transportation providers will
be adopting these new solutions, thus, making companies such as Nobina, key
business partners with Uber, taxi firms, car-sharing enterprises, and other
ancillary yet necessary alternatives in order to provide flexibility for sectors
where bus and train traffic is very limited or non-existent. Striking a balance
between providing flexible solutions and not compromising basic public trans-
portation needs is one of the challenges Nobina and other involved actors
identify. An obstacle going further in this digital transformation concerns
which forms of bilateral agreements will be formed between public transpor-
tation and private companies or between private companies to achieve service
6 Deloitte Insights: “Smart mobility: Reducing congestion and fostering faster, greener, and cheaper
transportation options”. Part of the “Smart mobility” research report
225
integration. Interviews reveal no competitive conflicts between parties in dif-

ferent transportation sectors; they mainly view each other as complementary.
Parties may engage in bilateral integration while participating in a public-
private or private-private partnership. Forms for Public-Private Partnerships
need to be developed. Local public transport agencies in Sweden already
experiment with tendering mobility as service provision, as they complement
rather than compete with car ownership, thus, matching the political goals of
Sweden’s sustainable transportation initiatives.
Nobina identifies several challenges and crucial competencies for the devel-
opment and implementation of mobility as a service. The company must
refine its relationship management skills in order to maintain cohesion in the
emerging mobility-as-a-service networks, including alignment of the different
interests. Nobina needs to coordinate actors that are currently separately
running their own businesses. In addition, Nobina must be involved in build-
ing and managing a technical platform with which it has an interaction or, for
example, outsourcing to IT specialists by leveraging the existing partnership
with Ericsson for the development of driverless buses. Having a mobility
service solution in place will require marketing efforts even though the intro-
duction of incentive-based contracts has led the company to start focusing
upon the end customers: the travellers. Communicating the advantages of the
solution to travellers that are used to the traditional transportation concept is
also a challenge.
CASE 3 GETINGE: CRE ATING NE W VALUE OFFERINGS

AND E XPLORING A NE W M ARKET BASED UPON BIG DATA 7
The Getinge group incorporated its largest division, Maquet in 2000. The
case concerns Getinge’s critical care, which involves intensive care products
such as ventilators, anaesthesia systems, and advanced monitoring.
While many businesses and industries have already embraced digitization
and big data, the healthcare and medical technology sectors are still in the
early phases of development. Clinical care providers offer increasingly com-
plex and advanced procedures, thanks to the advancement of medical knowl-
edge; however, the majority of healthcare actors are still mainly analogue.
7 This is a condensed version of a case presented in Rosén, A & Lechner, J. (2016), Advanced Market
Monitoring. Sensemaking and Strategic Analysis for Medical Device Complementors to Big Data
Platforms, Business case research thesis, Stockholm School of Economics, June 2016, and complemen-
tary interviews 2017.
226
Recordkeeping is conducted manually and medical journals are kept in paper

files. Medical devices operate in “silos” and are neither connected to central
platforms, nor communicated with other devices. Due to ageing populations
and medical progress, governments, hospitals, and medical companies are
increasingly engaged in digital transformation driven by the growing need for
and costs of healthcare.
Through the use of large data sets and advanced analytics, companies and
researchers are developing models that allow doctors to intervene early on in
order for patients to avoid suffering and for governments to save money. In
the hospital environment, facilitating and supporting digital solutions such as
clinical decision support systems are seen as tools to avoid mistakes and pro-
mote best practices. Overall, government officials and opinion leaders within
the sector see digitization and the use of big data analytics as the future of
healthcare.
Although digitization and the subsequent development of big data solutions
will improve healthcare, there are considerable uncertainties regarding what
is needed for a company to survive and/or thrive in potentially radically dif-
ferent market conditions. The picture becomes increasingly unclear as the
healthcare industry varies considerably across geographies, involving more
stakeholders than a typical consumer market.
Maquet Critical Care, a subsidiary to the medical technology company
Getinge, had earned a reputation for the sophisticated medical technology
and cutting-edge engineering of their ventilators, advanced monitoring, and
anaesthesia machines. Now, the company is increasingly feeling the pressure
of the shift toward digital transformation as other medical device manufac-
turers are increasingly interested in the data generated by Getinge’s (formerly
Maquet’s) machines. As part of its digitalization, Getinge faces the uncer-
tainty of when big data platforms and solutions in healthcare will become
popular. There is also considerable apprehension regarding the impact this
will have upon the competitive landscape. Will there be one or several indus-
try platforms? What organisation will be the platform sponsor? And, to what
degree will the platform be open or closed for external developers and com-
plementors? Since Getinge operates in the niche of high-end intensive care
medical devices, the company has limited capabilities to develop the techno-
logical capabilities and skills to provide a platform with the necessary range.
Even if pooled together with the other product and customer groups within
227
the Getinge Group, the company itself does not strive to become a “big data
company”, as one of their representatives expressed it in one of the interviews.
There are several strategic questions concerning the future digital plat-
form(s) for Getinge: one concerns the scope, type, and the technical nature of
the platform. Another concerns the number of platforms that will emerge for
the potential inclusion of the data generated by Getinge’s machines and com-
peting and complementary machines in wardrooms. A third issue concerns
the degree of the platform’s openness to external actors. There are also ques-
tions regarding the strengths of the complementors and what platform com-
plementors can bring to the platform ecosystem.
Another question impinging on investments in big data-based hospital
services is how to deal with the institutional context. For example, the Swed-
ish hospital system and leadership is decentralised, which forces suppliers to
develop reimbursement schemes with each individual hospital. A second
challenge concerns who will pay for the treatments; the government, an
insurance company, or a combination of both often pay for healthcare ser-
vices. Their interests, which are obviously not the same as the patient’s, create
tension between cost-efficiency and optimum care. Thirdly, reimbursement
fee systems vary according to local jurisdiction. Vendors must contend with
diverse bargaining power among public and private buyers, as well as those
buyers whose administrators and clinicians often disagree on the value of a
product. Old devices are incompatible for integration with new devices and
are mostly replaced piecemeal, as they stop functioning. This makes hospi-
tal-wide device integration difficult and unlikely. This significantly slows
down big data applications.
Furthermore, competitors’ machines flank Getinge machines. Thus, vari-
ous software and platform connections exist linking all these machines in
order to seamlessly collect and record vital patient data. Therefore, Getinge
must become a complementor to one or more other big data platform. This
makes it an issue for Getinge of how many platforms they compete with, who
sponsors the platform, and who will have access to develop it.
Overall, Getinge faces a scenario where it does not have a broad perspec-
tive of the big data market, which it is about to enter. Platform uncertainties
create risk where they may lose competitive advantage in the high-end medi-
cal device manufacturing market to a platform provider. Machine generated
patient data is stored in proprietary systems that are largely unintegrated.
228
The more data that are available (weight, age, back problems, smoker, and so
on), the better treatment options the systems are likely to produce. Systems
provide clinicians with a better base upon which to make decisions and, thus,
devote more time to the patients.
A move into new big data-based service business implies a business model
shift from offering products to offering services. The shift from selling products
to services is expected to be hard as conservative customers still often want to
buy a product, not a service provided by a product they will never own. How-
ever, continuing to sell products diminishes medical device manufacturers’
capacity to acquire and profit from the combination of medical and technical
knowledge needed to develop big data solutions. Secondly, enabling data and
privacy legislation allows actors to easily combine, analyse and use data
securely, while restraining data and privacy legislation does the opposite.
Hence, there is a need for legislation to catch up. The strategic challenge for
Getinge and medtech suppliers in similar stages of digitalization is the fact that,
by attracting new users (medical scholars, insurance companies), complemen-
tors add value to a big data platform while often protecting intellectual property
rights in their applications (more competitive, protected technology). If platform
sponsors are wed strategically or by profitability to an application, complemen-
tors may sign exclusivity contracts with them. In a competitive environment
with each other, complementors can protect their products with intellectual
property rights: by moving quickly and by working closely with other manufac-
turers in the industry to ensure they provide exactly what is needed. The ques-
tion of number of platforms is largely driven by multi-homing costs for users.
Switching between publicly accessible platforms might lead to lower costs;
however, industry stakeholders want more proprietary: modular standards that
can be costlier to users. Thus, having widespread standards is also a challenge.
Standards tend to converge actors and can reduce multi-homing costs.
CASE 4 SENSAVIS: AN EDUCATIONAL TECHNICAL

START-UP COMPANY IN SE ARCH OF A VIABLE BUSINESS MODEL 8
Sensavis, established in 2008, is a small Swedish company that offers high
quality, 3-D visualisation software for the education sector. Sensavis’ stated
8 This case is a condensed version of a longer case presented in Badhani, S. & Mut, J. (2017): Business
Model Innovation in Edtech; An Exploratory Study of Business Model Innovation in a Complex Market
Environment, Master thesis, Stockholm School of Economics); and from interviews with the focal
company, Sensavis 2016-2017.
229
vision is to improve learning outcomes by making complex and abstract

phenomena easier to understand: particularly in the area of science, technol-
ogy, engineering, and mathematics,. The mission is to give educators all over
the world tools that enable them to include all students in learning processes.
With headquarters in Sweden, the company operates in 36 countries serving
over 550 schools, and reaches more than 220,000 students. Sensavis went
through five distinct business model stages from 2008 to 2017, in an attempt
to create an offering and a market position in a new emerging market for
digitalized education. Changing the existing market was a major challenge,
as the company’s CTO expressed: “We are trying to change 100 years of
tradition in schools, and that is the biggest challenge.”
While many sectors have undergone digital transformation over the past
few years, the education industry is still at an early stage (WEF: New Vision for
Education, 2016). Firms that develop educational technologies, referred to as
“edtech companies”, challenge this status quo by creating digital products and
services that present educators with an opportunity to change teaching and
learning methods (Hsu et al., 2013). By doing so, edtech companies shape the
emerging edtech industry, which has attracted significant flows of private
investment over the past five years (WEF: New Vision for Education, 2016). In
this emerging edtech industry, companies struggle to develop viable business
models (Watters 2016). Needs are not yet defined and many stakeholders
(students, teachers, school leaders, national and local authorities, producers
of educational material, academic institutions specialising in pedagogy, and
parents) have unclear or overlapping roles during the transformation process.
Edtech companies need to develop business models to adapt to a complex
value network, which includes private and public stakeholders and where
value is strongly connected to sustainable development goals: through eco-
nomic (efficient use of public money) and social (study results, equal opportu-
nities including access to new education technology, and education better
adapted to the diversified needs of pupils/students). The multiplicity of stake-
holder demands, needs, and values in public-private networks, contributes to
a high level of complexity within an already heavily bureaucratic organisa-
tional environment (Williams & Lewis, 2008). The public education sector,
which is characterised by high levels of bureaucracy, risk-aversion, and slow
technology adoption (OECD Conference: Innovating the Public Sector: from Ideas
to Impact, 2014) stands in contrast to innovation-driven edtech companies.
230
Navigating in a network of multiple stakeholders is a central part of the pro-

cess to develop a viable business model.
In the first step of change, Sensavis moved from being a tool provider to a
product company. The journey began with Sensavis as a pure B2B company
serving sales and marketing teams for their sales presentations, mainly in min-
ing and machinery corporations. The value proposition focused upon “high
tech” visualisations. Entering the education sector, Sensavis attempted to stan-
dardise its product by bundling hardware and software into “The 3-D Class-
room” and selling their visualisations as an interactive visual learning tool,
which enabled instructors to explain complex and abstract components of a
subject in a simple understandable manner. In the second step, Sensavis changed
from a product company to become a software provider. Sensing the wide range
of different needs and characteristics among their customer segments, Sensavis
got rid of the hardware and moved from a product company toward becoming
a software provider. Sensavis changed the focus of its value proposition, empha-
sising the software’s beneficial impact upon student motivation, educational
growth, and learning outcomes. Co-creation with users had the biggest impact
upon the subsequent changes in the value proposition of the firm.
In the third change step, the company went from software provider to
become a pure service company. Collaboration with their expanding value
network increased the company’s knowledge of the institutional context and
entry barriers. Sensavis sensed stark differences in institutional settings,
between users and buyers, as well as among the user community; the company
addressed them through a revised value proposition. “Teach – Create – Acti-
vate” (TCA) was born. The new value proposition focused upon improving
learning outcomes. Due to the complex market situation and the lack of a
digitization roadmap, Sensavis sensed that the real value for most of their
customer segments lay in the provision of guidance, support, and inspiration.
Seizing this insight, Sensavis shifted focus toward becoming a fully-fledged
service company, not only selling software, but rather acting as a trusted digi-
talization partner to schools, municipalities, and ministries of education.
When adapting to different markets, Sensavis was forced to reconfigure
the business model components to local institutional and settings.
Experience with private companies did not prepare Sensavis for the barri-
ers in the education sector. Within the first few months, the company sensed
that understanding the institutional context and positioning the company in
231
relation to it would be the first step toward delivering value to its customers.
Having one foot in the classroom helped Sensavis understand differences in
needs within and across its customer segments.
Depending upon the local institutional context, differences appeared across
markets, which complicated the sales process. Sensing these differences, Sen-
savis changed the sales pitch, adjusting the value proposition according to
which actor the company spoke, and aligned appropriate channels and rela-
tionships to match it.
Sensavis successively identified potential value flow opportunities, expanded
the network, and changed the nature of existing value flows. Product develop-
ment rooted in user centricity became central in value creation. Co-operation,
co-opetition, and co-creation with actors in the value network changed where
value was created not only by Sensavis, but also increasingly in collaboration
with customers, users, and other value network participants. Initially, close
contact with customers was necessary due to the complexity of the visualisa-
tions and constituted a demanding task in project management processes.
Analysis
DIGITAL TR ANSFOR M ATION DIMENSIONS

Important dimensions in digital transformation include the digitization of
information, assembling and analyses of large amounts of data (big data),
internet representation of physical objects (Internet of Things), and technical
platforms acting as intermediaries of information. These are all important
characteristics in the following cases.
Philips develops intelligent lighting systems: an IT network with equip-
ment composed of sensors, controls, smart lamps, and energy meters. Philips
aspires to create a platform for innovation with technology partners, comple-
mentors, and customers/users: for example, municipal authorities. Digital
transformation opens opportunities to redesign public transportation service
to a mobility service for Nobina. This digital transformation depends upon
Internet of Things: a diffusion of smartphones and apps, digital payment
solutions, and the potential of big data for planning and operating public
transportation, as well as the development of a technical platform. The Get-
inge case focuses upon the challenges and opportunities for a medical equip-
ment company when technical and medical innovations open up for the use
232
of big data to increase the effectiveness of medical treatment. For this to be

possible, one or more technical platforms need to be designed and operated,
upon which data are assembled, analysed, and made available. A major issue
for Getinge is if the data assembled by their equipment should be added to
one or more platforms, be it open or private, and how this data should be
used. The start-up company Sensavis participates in early phases of digital
transformation for education activities, essentially promoting visualisation as
an interactive learning tool. The company experiences the lack of a “road
map” for digital transformation of education as being problematic, evidenced
by the frequent changes in its business model. Sensavis´ development does not
relate to major digital transformation concepts, such as Internet of Things,
big data, and technical platforms during the studied period.
DIGITAL TR ANSFOR M ATION AND PUBLIC SERVICE INNOVATION

PROMOTING SUSTAINABLE DE VELOPMENT
A WEF report (World Economic Forum, 2016) states: “We believe that digital
solutions can bolster the three pillars of development on which the Sustain-
able Development Goals (UN) are built: improving people’s quality of life,
fostering equitable growth, and protecting the environment” (p.9). The report
also specifies that ”…digital innovation is reshaping industries by disrupting
existing business and operating models. But it is also having a profound
impact on society, presenting a series of opportunities and challenges for
businesses and policy-makers.” (p.4). Thus, the interconnected business mod-
els of both private and public stakeholders need to harbour various forms of
commercial, business value (for example, manifested in new, extended, revised
revenues and/or efficiency improvements), as well as societal values (where the
digital transformation might lead to improving people’s quality of life, foster
equitable growth, and protect the environment).
Aligning policies, enhancing collaboration, and engaging the whole ecosys-
tem of government, business, and civil society are some of the pre-requisites
for making these digital transformations and broad service innovations
actually happen.
In line with an observed trend in marketing literature and in business prac-
tice, the case firms shift to a more service oriented, customer centric strategy
(Vargo and Lusch 2004), related to public service innovation enabled by digital
transformation. For public services, interaction with public agents and adapt-
233
ing to public norms becomes important, which must be included in business

model considerations.
A central dimension in any business model is the value offering to the focal
customers. The value offering provided by Philips is formulated in a perfor-
mance contract that promises significant environmental and financial bene-
fits for the public customers. Digital transformation makes it possible to tailor
lightning to changing needs for light and to optimise resources. Retained
ownership of the products and equipment enables Philips to apply circular
economy principles and to be a technology partner to the customer in appli-
cation of new more effective lighting methods in the future. A seemingly
unresolved issue is how the value, which is co-created, should be calculated
and how a performance contracts shall be written. Also, when a contract
period ends, is it possible to open a new tendering process when Philips enjoys
such a dominant position?
A mobility solution for public transportation, as in the Nobina case, aims
for a more flexible, needs-based, and effective public transportation system
that integrates public and private operators. The value offering includes soci-
etal benefits in terms of time-savings, lower costs related to traffic accidents,
and reduced CO2 emissions. For Getinge, the added value of the data gener-
ated by use of its medical equipment depends upon how the data can be
assembled in big data platforms and be used for more effective healthcare
activities. The value offered by Sensavis in its latest business model is to help
actors in the education sector to design and implement digital transformation
projects in educational processes.
Sustainable development criteria involve a wider perspective on what types
of values are involved compared with a more narrow traditional business
model analysis. To balance and coordinate the business and the societal values
associated with digital transformation and associated service innovations
emerge as one common business model challenge. This becomes a matter of
defining the customers, and the various values for different beneficiaries: con-
sumers, citizens, business organisations, public organisations, the environ-
ment and society at large. In addition, defining the financial/economic val-
ue(s) will be part of the business model challenges. For private actors such as
Philips and Nobina, this might mean radical revenue model shifts. Increased
overlapping between public and private networks will blur the boundaries
between the customer, the key partner, and the roles of supplier and distribu-
234
tor. In this respect, variation between different communities and different

geographic locations adds to the challenges of developing business models
that serve to coordinate activities across several involved business firms and
localities, and to handle tensions between private and societal values. Further-
more, digital infrastructure issues also need to be handled: for example,
determining upon what technical platforms to base the service innovation
and the degree of the platform’s openness.
Public actors need to rethink policies in order to participate in the co-
creation of service innovations: for example, this relates to procurement
procedures, service operations, pricing and cost accounting models, infra-
structure, and principles for public/private collaboration.
The complexity of the customers and the customer interaction for the small
company Sensavis became more and more apparent for two reasons: first, the
small organisation had to adapt to different types of local school authority
networks when the company expanded internationally; secondly, the growing
service orientation meant there was closer interaction with the user networks
of deans, IT support functions and schools, teachers, and pupils/students.
There are major problems in successfully developing and implementing
business models for the service innovations in the cases. Nobina admits that
there are problems moving from a production to a service culture, for both
public and private actors, and coordinating actors that currently run their
operations separately. There is also reluctance among local authorities to
make the necessary investments in the digital infrastructure.
The four cases illustrate efforts aimed at business modelling rather than at
developing a fixed business model. Nobina recognises the importance of devel-
oping incentive-based contracts focusing upon value for end customers and the
need to strike a balance between flexibility and basic public transportation
needs. For Nobina, as a presently traditional public transport company, it is
obvious that, they need to integrate resources that are held by others especially
private transportation providers in order to provide public mobility services.
The WEF 2016 report states: ”Government leaders, regulators and civil
servants have a significant opportunity to support the adoption and dissemina-
tion of new technologies, using the key levers that are available to them: setting
the appropriate policies, regulations and national strategy for digital, while also
transforming government agencies into hotbeds of innovation. Other impor-
tant success factors include enhanced collaboration, both nationally and inter-
235
nationally, and the ability to engage the whole ecosystem of government,

business and civil society.”(p.22) WEF´s statement focuses upon public actors
while we present a private business perspective. We will now suggest five issues
in digital transformation, public service innovations, and sustainable develop-
ment that we argue should receive increased attention in practice and research.
Business Modelling Issues for

Further Inquiries in Practice and Research
All four cases show critical influences on digital transformation by public actors
and public/private interaction. The tendering process for municipal procurement
of lighting for Philips needs to change. The present focus upon initial cost,
ownership, and control of equipment needs to become a focus on total costs,
maintenance service, ecology criteria, and technical development. Legislative
issues involved in changing ownership of material, public-private partnership,
and public funding should also be solved. The tendering process needs to be
adapted for municipal customers: from initial purchase price to total cost and
ecological effects, public-private partnership, public funding, and legislative
issues involved in changing the ownership of material.
The digital transformation-based innovations by both Philips and Nobina
should be helpful in implementing sustainability targets set by public actors.
However, public-private partnership issues must be solved when developing
and implementing the mobility solution to public transportation. For exam-
ple: how should the tendering process adapt to the new criteria, what form of
bilateral agreements will be needed, what type of contract is required, and
who will develop and operate the technical platform? The institutional con-
text has a major influence upon digital transformation at Getinge as well.
Both private and public customers in the healthcare sector in Sweden, with its
decentralised organisation and local jurisdiction, are reluctant to buy services
instead of products; this is due to unclear reimbursement schemes, as well as
data and privacy legislation.
Different institutional settings, bureaucratic public organisations, the influ-
ence of multiple public and private stakeholders, and public funding play
important roles for the start-up, Sensavis.
In the context of sustainable development, the five areas for future atten-
tion and inquiry into business modelling adopt a broad perspective in practice
and in research on business models in the public service
236
First, digital transformation of public service increases customer complexity

and changes interaction with customers. Secondly, the process is also associ-
ated with changes in the value offering. Third, new partnerships need to be
established. Fourth, digital transformation implies the creation of new infra-
structures: for example, with the increased importance of technical platforms.
Lastly, dimensions of business models need to be developed and implemented
when it concerns the financial and economic, which includes revenue as well.
1. INCRE ASED COMPLE XIT Y OF CUSTOMERS AND

CUSTOMER INTER ACTION INFLUENCES BUSINESS MODELLING
All companies found an increase in the complexity and, in some ways, the
institutionalised rigidity of the networks of actors (deciders, buyers, users,
and so on) connected to the customer side. This concerns the structural com-
plexity and/or the complexity of the customer practices. Public customers’
procurement policies and practices were not adapted to the new value offer-
ings provided. For Getinge, the complexity of the customer side was mani-
fested in the discussion of how to approach regional authorities, hospital
management, intensive care departments, and their leading doctors, and so
on. Sensavis encountered similar challenges when it approached, to a varying
extent, deans, teachers, schools’ IT departments, local municipalities, central
school administration units, and so on in different parts of the world. The
newness of the service offerings challenged the seller´s own knowledge of the
customer’s varying needs of different units, and the customer´s view of the
new value offerings.
We find that more attention in practice and research is needed on how
awareness of and adaptation to the complexity of structure and practices
within the public sector shapes customer business modelling.
2. INCREASED END-USER ORIENTATION PROPELS BUSINESS MODELLING

AS FIR MS SUCCESSIVELY DEVELOP THEIR SERVICE-BASED VALUE OFFERINGS
The development of the value offering or ideas about a new value offering
over time was associated with a successively increased orientation toward end
users and in end-user contexts in all four cases. Philips describes the impor-
tance of understanding the various needs of the citizens in (big) cities. Nobina
focuses upon the total and detailed transportation needs of individual travel-
lers. Getinge directs its attention to big data through an increased interest in
237
doctors’ decisions and support needs in interactions with patients, as well as

the machines in intensive care contexts. Lastly, Sensavis successively shifts
the attention to the pedagogical needs of teachers in their interactions
between students with varying needs, which also includes student-to-student
interactions and teacher/community interactions. Hence, the deeper compa-
nies go into the digital transformation processes in order to create and imple-
ment new service-oriented value offerings, the more they need to understand
the details of the end-users’ use contexts.
business modelling successively adapts value offerings to increased end-user
orientation.
3. PARTNERSHIP CHANGES AND NE W PAT TERNS OF COOPER ATION

AND COMPETITION DRIVES DIGITAL TR ANSFOR M ATION AND
BUSINESS MODELLING
The three incumbent firms’ show the same general pattern: cooperation with
others increases in order to integrate resources and implement new value
offerings. The number of key partners needs to be expanded, often resulting
in increased overlapping of networks and, consequently, new patterns of
cooperation and competition. For the small start-up Sensavis, the later stages
of the company’s business modelling process indicated the start of similar
network complexity and expansion, as the company began to connect to
partner companies in the publishing industry.
business modelling is associated with expansion of the relevant network of
actors involved in the provision of the digital service.
4. BUSINESS MODELLING REL ATED TO SHIF TS

IN THE ROLES AND POSITIONS OF DIGITAL PL ATFOR MS
The first three cases describe an emerging digital infrastructure, which com-
prises the development of two or more competing technical platforms. In the
Sensavis case, larger platforms are beginning to emerge. In all cases, business
modelling is importantly influenced by the planned or existing technical
platforms upon which the planned value offering should be based. Which
actors will become owners and/or managers of platforms? How will the tech-
nical platforms as intermediaries between actors involved in the provision
238
and use of services relate to individual actors´ business modelling? A powerful

actor such as Philips would more easily manage the competition between
technical platforms than would a less powerful one, such as Getinge. How-
ever, for a small start-up such as Sensavis, the lack of dominating technical
platforms creates opportunities for the continued development of the compa-
ny’s own technical platform, before competition increases between the plat-
forms, thus, affecting competition and customer relations.
We find that more attention in practice and research is needed in order to
understand how shifts in digital infrastructure and technical platforms inter-
act with business modelling.
5. IMPORTANT SHIFTS IN BUSINESS MODELS DRIVEN BY EXPERIMENTATION

(INNOVATION) AND THE TESTING OF NE W ECONOMIC MODELS
A central challenge for most companies is to find a viable revenue model as
part of the business model. Business modelling is often associated with con-
siderable experimentation and successive adaptation of the model for finan-
cial viability. Financial viability includes the challenge to find, test, and
manage new revenue, as well as charging, pricing, and cost management
models. Digitalization with more service-based offerings challenges existing
revenue models. Sensavis changed revenue model several times, thus, shifting
from product service sales to more “pure” service and access-based contracts.
In the other cases, customers’ public procurement policies did not seem to be
adapted to this new form of interaction, relationship, and contracting with
service providers. While the value offerings required other contractual rela-
tions and longer-term interactions between the suppliers and the public cus-
tomers/users, the traditional procurement policies and procedures of the lat-
ter, indeed, imposed obstacles.
We find that more attention in practice and research is needed in order to
understand how experimentation related to service innovation requires that
business modelling include new aspects of economic viability of public services.
239
References
Andersson, P. and Mattsson, L-G. (2015). Service innovations enabled by the ‘internet of
things’. IMP Journal, Vol. 9 Iss 1: 85–106.
Araujo, L. and Spring, M. (2006). Services, products and the institutional structure of
production. Industrial Marketing Management, Vol. 35: 797–805.
Ehret, M. and J. Wirtz (2017). Unlocking value from machines: business models and the
industrial internet of things. Journal of Marketing Management, 33:1–2: 111–130.
Håkansson, H. Ford, D., Gadde, L-E., Snehota, I. and Waluszewski, A. (2009). Business in
Networks. Chichester: Wiley
Johanson, J. and Mattsson, L-G. (1992). Network positions and strategic action – An analytical
framework. In: B Axelsson and G. Easton (eds.), Industrial Networks: A New View of Reality,
London: Routledge
Kindström, D. (2010). Towards a service-based business model – key aspects for future
competitive advantage. European Management Journal, Vol. 28 No. 6: 479–490.
La Rocca, A and Snehota I. (2017). Business models in business networks-how do they
emerge? IMP Journal, Aug 2017.
Maglio, P.P. and Spohrer, J. (2013). A service science perspective on business model
innovation. Industrial Marketing Management, Vol. 42 No. 5: 665–670.
Mason, K. and Spring, M. (2011). The sites and practices of business models. Industrial
Marketing Management, Vol. 40 No. 6: 1032–1041.
Mattsson, L.G. (1998). Dynamics of overlapping networks and strategic action by the
international firm. In: Chandler, A.D., Hagström, P. and Sölvell, Ö. (Eds), The Dynamic
Firm, Oxford University Press, Oxford, 242–259.
Osterwalder, A., Pinneur, Y. and Tucci, C. (2005). Clarifying business models: origins,
present, and future of the concept. Communications of the Association for Information Systems,
Vol. 16: 1–25.
Vargo, S.L. and Lusch, R.F. (2004). Evolving to a new dominant logic for marketing. Journal of
Marketing, Vol. 68 ( January): 1–17.
Vargo, S.L., Wieland, H., and Akaka, M.A. (2015). Innovation through institutionalization:
A service ecosystems perspective. Industrial Marketing Management, 44: 63–72.
World Economic Forum (2016). Unlocking Digital Value to Society: A New Framework for Growth.
In Collaboration With Accenture. White Paper, January 2017
Zott, C., Amit, R. and Massa, L. (2011). The business model: recent developments and future
research. Journal of Management, Vol 37 Issues 4, July: 1019–1042.
240
Business Models
and Ecosystems
CHAP TER 12
The Role and Potential

of IoT in Different Ecosystems
JAN M ARKENDAHL, STEFAN LUNDBERG, AND STAFFAN MOVIN
Introduction
The concept of Internet of things (IoT) usually includes both technology and
services that is based upon connected devices and the use of the collected
data. In this chapter, we study how IoT can be introduced and used in differ-
ent industrial sectors. We analyse various cases with IoT products and ser-
vices in order to identify how different business aspects and conditions affect
the ability for actors to make commercial use of the new technology. We dis-
cuss in terms of IoT technology. The analysed cases are about introduction of
ICT solutions in general; this is just a part of the ongoing digitalization of
products and services.
The objective of our research is to study the conditions for the use of IoT in
order to identify drivers and benefits, as well as problems and challenges. We
want to identify common patterns and key challenges for introduction of IoT
as a new technology. In order to do so, we look into different industries:
industrial IoT (typically manufacturing), smart energy, smart homes, smart
cities, healthcare & social care, and sport & well-being.
Introducing IoT can clearly lead to improved efficiency; however, it is not
only about technical performance. The improved working efficiency is also
reflected in changed or new working processes, usually in combination with
new roles and business opportunities for market actors. Although Ericsson is
a technology-oriented company, it has discussed for many years now digital
society in terms of business transformation1. We claim that it is not enough
1 https://www.ericsson.com/digital-services/offerings/digital-transformation
243
JA N M A R K EN DA H L , ST EFA N LU N DBERG, A N D STA F FA N MOV I N
that the technology works as expected. We want to highlight the need to also
consider the business model aspects when new technology is introduced.
Digitalization is not only about increased efficiency; it is also about the
opportunity to offer new services or to offer them in innovative way. The value
of specific technical solutions must be seen in an overall context. A new tech-
nology may result in direct benefits within an actor´s existing business; it may
also be that the full benefit of a new solution cannot be achieved by providing
the solution itself in isolation. In order to achieve the full benefits, a solution
may need to be combined with other solutions (that is offered by other actors).
This implies that one needs to have an understanding of the customer’s entire
needs and how different actors can cooperate, as well as study ecosystems and
networks of actors (business networks) and how the actors interact.
So what do we mean with a business ecosystem? This term has been widely
used over the past years, especially for mobile services (Basole, 2009); it has
also been used to describe and analyse IoT services, businesses, and actors
involved (Westerlund, Leminen, Rajahonka, 2014), (Ghanbari, Laya, Alon-
so-Zarate, Markendahl, 2017). We will look into business ecosystems with a
multitude of co-existing “businesses”, each described by a value network. Our
results support the assumption that there are several different ecosystems and
industries in which IoT is included or can be included; we do not think that
IoT will have an ecosystem of its own.
The main question discussed in this chapter is the following: What are the
main business-related challenges that we can observe when a new technology
such as IoT is introduced?
The primary research contribution is discussing the introduction of new
technology from three separate perspectives: i) How a new solution can be
evaluated differently depending upon the overall context; ii) How the value
can be assessed; and iii) What activities are performed by different actors.
Our key point is that one needs to consider the role of new technology and
services within an overall business context and not only the technology itself.
The chapter is organised as follows: first, we discuss theory, the analysis
approach, and the data collection, which lead to the cases to be presented.
Then, we present a number of selected cases, what can be observed from
these cases, and a summary of common patterns and challenges. A discus-
sion on observed patterns, drivers and challenges then follows and, lastly, a
summary.
244
T H E ROLE A N D POT EN T I A L OF IOT I N DI F F ER EN T ECOSYST EMS
About Theory and Methodology
IOT AND BUSINESS MODEL RESE ARCH

There have been a number of research papers over the last five years discuss-
ing the business aspects of IoT products and services. Westerlund et al (2014)
looked into business models and the business ecosystem aspects for IoT appli-
cations, claiming the ecosystem structure is not clear with many multi-indus-
try solutions and that actors are still looking for their roles.
Business model definitions usually consist of a number of different dimen-
sions or components. Although they differ, we claim they cover the same
aspect, to a large extent; it is a matter how you organise the analysis and what
aspect you include in each component. For example the STOF model has four
main components or domains describing services, technology, organisation,
and finance (Bouwman, Haaker and De Vos, 2008). Here, the organisation
domain includes the value network. The business model definition used by
Chesbrough and Rosenbloom (2002) consists of six main components. One
component is firm organisation and value chain; the other is firm in the value
network. The business model canvas proposed by Osterwalder is a well-
known example where different components are used to describe the business
model (Osterwalder and Pigneur, 2010).
The concept of networked business models, including business networks
and partners, is important for our analysis. A key issue discussed in research
on networked business models is the move from single firm business models
to networked business models (Palo and Tähtinen, 2013), (Bankvall, Dubois
and Lind, 2017). For our analysis, it is relevant to capture multi-actor aspects
of value creation and how the value network can be composed. Citing the
work of Palo and Tähtinen (2013) about the network-level business model ”by
developing collective understanding of the business opportunities and shap-
ing the action to exploit them”.
RESE ARCH APPROACH

For our analysis, we need to identify what are actors in the ecosystem that
interact with each other; this includes being customer, supplier, partner or
competitor. In our analysis, we are interested in seeing which actors do busi-
ness with each other or cooperate in order to do business. We are also inter-
ested in evaluating the importance of the new (IoT) solution that we study.
245
Hence, we need to look into the value proposition; that is to say, what type of
value is offered to the customer: possibly an end-user or it could also be a
company or an organisation.
Since we look into specific services where the IoT solution plays a part, it is
natural to identify the activities involved: both the ones where a new IoT-
based solution plays an important role, as well as all other activities that make
up the overall service. Examples of overall services include the following:
waste management in a city, cleaning of office buildings, maintenance of
infrastructures such as power plants or bridges, and municipality homecare
services for the elderly.
Besides the activities and analysis, we need to look into which actors perform
different activities: “Who is doing what?” Identifying which actors have rela-
tions is linked to this: that is to say “Who is doing business with whom?” This
reasoning leads to concepts such as the ARA model discussing Actors,
Resources, and Activities (Håkansson and Snehota, 1995) and the activity sys-
tems (Zott and Amitt, 2010). The activity systems of a business model can be
described from different perspectives, content, structure, and governance. The
content refers to which activities are performed. The structure describes how
activities are linked and the governance describes who performs the activities.
DATA COLLECTION AND ANALYSIS

We have a rich set of primary data from different industrial sectors collected
at workshops and interviews: from 2016-2017. Most of the case studies and
findings are the result of a Swedish research project called “IoT Ecosystems”2.
Besides academic researchers, the project included big industry companies
(Ericsson and Sandvik), one SME, providers of energy and telecom services
(Vattenfall, Telia), Stockholm City, and an employer organisation (Almega).
The participation of these big organisations enabled us to get in contact with
a large number of people with different experiences.
WORKSHOPS AND INTERVIE WS

We invited guests to a total of 18 workshops, and organised approximately 20
separate interviews (see Table 12.1). The collected data is used in the following
two ways: i) to present selected cases, which illustrate key characteristics when
2 The work is supported by the Swedish Innovation agency, Vinnova and the Strategic Innovation
Program, IoT Sweden: research grant #2015-06151
246
IoT is introduced; and ii) to present common pattern and challenges. We have
described and identified cases and conditions from different sectors based
upon the collected data: industrial IoT, smart energy, smart homes, smart cit-
ies, healthcare and social care, and sports and well-being. During the inter-
views in the workshops, we collected information on which actors were doing
what, how actors interacted with each other, and how working processes were
or were not changed. People sharing insights about IoT-related projects and
initiatives provided the basis for identifying drivers, benefits, obstacles, and
common patterns related to the introduction of IoT products and services.
Those interviewed were open about their experience within their respective
organisations and partners. They could also share insights about cases of gen-
eral interest. We found that technology itself usually works as expected; how-
ever, there are still obstacles for commercial breakthrough. We want to identify
these obstacles and understand the reasons for them. The collected data
enabled us to identify common patterns, challenges and problems, and select
good cases to illustrate these hurdles: see next section Cases and Findings.
Table 12.1. Sources of Primary Data Collection

Industry Sector Companies and Organisations Contributing to Data Collection
Smart cities ABB, ElectriCity, Envac, Ericsson, Fortum InfraNode, Qlocx,
Riksbyggen, Scania, Skanska, Stockholm City, Veolia, Volvo
Smart homes and smart energy ABB, Ericsson, Electrolux, Ellevio, Fortum, HEBA, Intel, NCC, RISE,
Telia company, Vattenfall
Sport, health and wellbeing Biosynch, Ericsson, Interactive Institute, MSD, MTC, RaceFox,
The Swedish School of Sport and Health Sciences
Healthcare and social care Alleato, Almega, Biosynch, Cenvigo, Hemfrid, Intel, Joicecare, Phoniro,
Sensative, Stockholm city, Stockholm county council, Telia Company
and Municipalities Nacka, Norrtälje, Södertälje, and Uppsala
Industrial IoT and ABB Corporate Research, ABB Robotics, Berotec, Clayster, Combient,
manufacturing EzeSys, Ericsson, Sandvik, SICS, Telia, Volvo
How To Use The Collected Data? We present a number of cases in this chapter
that illustrate the use of IoT and/or Smart home/city products or services.
Here, it is important to note the observations represent two types or levels of
“cases”. On the one hand, we have the specific IoT solutions: for example, the
connected waste bin/ soap dispenser/ bolt/ smart lock; we call this an IoT-
based innovation. On the other hand, the collected data is used to describe
247
specific business contexts or environments where the IoT-based innovations

are used. Examples are maintenance, cleaning, and waste management or
homecare services.
When we look into our cases, we make use of the aforementioned
approaches, structure the primary data, and analyse the activities from three
different perspectives:
a) In which activities is an actor involved?
How large or important is that actor’s contribution?
b) How important is a specific activity for the overall service?
What is the value of a specific activity and how important is the actor’s
role in performing said activity?
c) What other activities are needed in order to exploit the value of a
specific activity? Which actors are involved in these other activities?
An IoT-based innovation
A business environment where the innovation is used
Figure 12.1. Illustration of IoT-based innovation playing a small or big role in a business environment
or being linked to/being a part of another service.
Cases and Findings

In this section, we present selected cases along with key observations and
identified patterns and challenges that arise when introducing new technol-
ogy. The challenges will be further discussed in the next section.
SELECTED CASES
The Connected Trashcans: An Example of Urban Furniture. Connecting trashcans
to a monitoring system has become a solution to avoid emptying half-empty
trash cans, thus, saving time and the environment. The trashcans report
when they are full. One provider of such a solar panel-based solution is Big-
248
Belly3. This kind of system is used in many towns in Sweden: for example,
Uppsala. Since the trashcan is a connected system, it has a feature where you
can offer people WiFi services. Other features may include information to
visitors about the local area.
Will this be a complete business solution focusing on efficiency for waste
management? Or can you expect that it is the beginning of a larger service,
which can include all cleaning and care in an area? Equipment, such as trash-
cans, is sometimes called urban furniture. Other examples include: lamp
posts, bus stops, ad signs, and billboards. This urban furniture can provide
the basis for a connected infrastructure that can be used for different services
without relying upon mobile operator networks. Operators can also make use
of the urban furniture instead of deploying (their) own base stations (Ghan-
bari, 2016). This possibility for re-use of equipment is highlighted on Bigbel-
ly’s webpage: “Bigbelly provides a public right-of-way platform to deliver
Smart City solutions and host communications infrastructure”.
The Connected Bolt: Going From Product to Service. The Swedish start-up com-
pany, StrainLabs’4, has developed a connected bolt that opens up for a new
type of business: in the area of smart monitoring and inspection. A sensor
embedded into a cavity in the head of the bolt monitors preload and temper-
ature. When it detects that a bolt is about to come loose, it alerts the user;
thus, pre-emptive action can be undertaken. Inspection today is a manual
activity performed in the field, posing several challenges. An offshore oil
platform has a hundred thousand bolts in hard-to-reach places that need to be
inspected over a 5-year period. A wind turbine has giant blades: high above
the ground. The amount of time and money spent on inspecting and report-
ing globally upon critical applications is staggering.
The StrainLabs solution is an example how one can offer a service based
upon a connected device. The CEO says: “We no longer sell a screw; we sell
a service to monitor the screw connection”5. This service represents a large
share of a set of activities for monitoring and inspection. Power by the Hour,
the Rolls-Royces maintenance program for airline companies is another
well-known example: where a product offer is replaced by a service. This
3 http://bigbelly.com/
4 http://strain-labs.com/bringing-internet-of-things-to-bolted-connections/
5 https://www.nyteknik.se/startup/uppkopplade-skruven-larmar-nar-den-blir-los-6579513
249
includes a complete engine and accessory replacement service on a fixed-cost

per flying-hour basis.
The Connected Soap Dispenser: A Part of a Cleaning Service. SCA is Europe’s
largest private forest holding company, producing solid-wood products, pulp,
publication papers, and renewable energy. One product is liquid soap for
everyday hand washing in schools, restaurants, nursing homes, and public
spaces. SCA also offers soap dispensers and other equipment for toilets. Soap
dispensers can be connected using the Tork EasyCube™ concept, which
focuses upon improved cleaning 6.
Tork EasyCube is a cloud-based service that collects data from connected
devices. Displayed in web applications, information directs cleaning teams to
exactly where they are needed. SCA offers connected devices, visitor count-
ers, real-time data collection and display, and an analytics tool for optimizing
operations.
The solution is tested at amusement parks and zoos, such as Skansen and
Furuvik; customer cases are presented on the SCA web site. Clear benefits
have been identified for amusement parks that can handle unpredicted large
variations in number of visitors 7. However, the value with the connected
soap dispenser is not so clear for cleaning regular offices or workplaces, where
the new technology is more used for improving the efficiency of toilet clean-
ing. This is one of the overall represented activities: cleaning, facility manage-
ment, running an office or an amusement park.
Connected Vehicles: Toward Transport as a Service. The concept of connected
cars and self-driving (or driver-assisted) cars has been discussed for some time
now. Although concepts such as the connected vehicle cloud8 were presented
some years ago, the big break-through for consumers is still to come. For
professional use, however, the situation is different.
Manufacturers of transport vehicles have looked for years into the possibil-
ities with connected vehicles. Companies such as MAN, Volvo, and Scania
have developed and have used solutions where one can track different types
of data: for example vehicle, engine, driving, and driver information. Scania
has a large number of connected trucks that are delivered to different trans-
6 http://www.torkusa.com/easycube/
7 https://www.tork.se/kundcase/furuviks-zoo
8 https://www.ericsson.com/digital-services
250
port companies9. Scania and other companies foresee a development path

from the connected vehicle to a connected fleet, possibly leading to “transport
as a service” instead of selling vehicles.
Vehicle manufacturers have different types of vehicles on the market; all have
specific in-vehicle communication systems. This leads to potential problems for
the transport companies that usually have trucks from many manufacturers.
The actors in this sector, however, have agreed upon a solution that helps trans-
port companies handle this multitude. The Fleet Management System (FMS10)
standard enables common interfaces and third parties to access vehicle data.
Regardless of which manufacturer produced a certain vehicle, if it is equipped
with an FMS interface (gateway), there is the same output for all vehicles.
Digital and Smart Locks. Digital locks or smart locks are part of a growing
market, mainly focusing upon offices: as an access system to real estate build-
ings, and in various applications within other industries. Digital locks offer
opportunities in other sectors as well, primarily in the field of homecare for the
elderly. “Keys” are installed in mobile phones, so there is no need to collect
physical keys before one visits the home of an elderly person. Attempts have
been made in several rounds over the past 25 years, and it is only now there is
an interest and an understanding of digital locks in the homecare sector.
Technology solutions are emerging, thus, developing the opportunity to
produce advanced solutions at attractive prices. This means that digital locks
should be considered as part of a larger whole since they could possibly solve
other needs too, if used in a more complex context. Insurance companies have
started accepting digital locks, so they can replace the old locks in all homes.
In this case, the lock becomes a part in a significantly larger whole, meaning
that a digital lock will not automatically be a “smart” solution. Their role, in
a certain context, must be properly analysed. For the individual private house
owner, the individual lock is a fully adequate solution; however, it may not be
so for an entire municipality.
The Connected Service Box or Delivery Room. The company Qlocx11 offers
solutions where logistics companies and users have access to common deliv-
ery “boxes” or rooms. Qlocx develops these smart delivery containers and
spaces to ensure that one receives goods or deliveries without physically
9 https://www.scania.com/group/en/scania-reaches-milestone-250000-connected-vehicles/
10 http://bus-fms-standard.com/Truck/index.htm
11 http://www.qlocx.com/
251
needing to be there. This allows for a simpler and more efficient logistical
flow in the workplace.
Qlocx has developed smart digital locks to devices adapted for package
delivery for private persons and housing companies. One can open the device
with a mobile phone and hand out digital keys to anyone. Typical use cases
are delivery companies that can deliver mail or a neighbour who wishes to
borrow a tool when you are not at home. Hence, one can receive or share with
neighbours by giving out a digital one-time key. With the Qlocx delivery box,
one can also rent or loan one’s car to someone else without physically needing
to be in place to issue the key.
Qlocx also offers smart delivery spaces for companies, which makes it
possible to receive goods or deliveries without anyone physically needing to
be there to receive them. This allows for delivery to working spaces outside
of regular working hours, as well as a simpler and more efficient logistics flow
to the workplace. The Qlocx solution can be used for service boxes, delivery
rooms, and even delivery containers.
Table 12.2: Identified Drivers and Benefits of Selected Cases
Case Drivers and Benefits

A. The connected trashcan Improved working efficiency
Re-use of infrastructure
B. The connected bolt Improved working efficiency
Possible to offer new services
C. The connected soap dispenser Improved working efficiency
D. Connected Vehicles Improved awareness
Improved uptime
E. Smart and digital locks Improved working efficiency
F. Smart delivery box or container Improved working efficiency
WHAT WE CAN LE ARN FROM THE SELECTED CASES

When looking at the value proposition that is enabled or strengthened by the
introduction of IoT products or services, a number of clear drivers and bene-
252
fits can be identified. For example, improved awareness and control of activ-
ities leads to improved working efficiency and processes. This implies better
resource utilisation, improved uptime, and reduced costs. In addition, the IoT
solution also offers the possibility to be re-used for purposes other than the
originally intended one. Finally, the IoT solution may lead to the possibility
of offering new types of services: that is to say, new types of businesses and
revenues (see Table 12.2).
We can also obtain greater insight by studying specific activities and which
actor that is doing what in the selected cases. This relates to the components
of the value chain or network: for example, business model aspects as “firm
organisation and value chain” and “firm in the value network” (Chesbrough
& Rosenbloom, 2002). Here, we can also identify whether the offer consists of
one or several services and if one service is part of another. In addition, we
identify whether or not it can be offered as a stand-alone service and if the IoT
solution enables a new product and/or service (see Table 12.3).
Table 12.3: Analysis of Main Activities From Cases
Activity Who Performs Stand-alone or Part Type of

Under Study the Activity of Another Service Novelty
Emptying Municipality Stand-alone New product (new service)
Trashcans itself
Monitor bolts Bolt provider or Stand-alone New product and service
in large systems System owner
Smart cleaning Facility manager or Part of overall cleaning New service
of toilets cleaning company
Monitor heavy Vehicle manufacturer Part of support and Improving service
vehicles & fleet vehicle maintenance
Access using Homecare provider Part of homecare Improving Service
smart locks
Delivery of Goods Delivery company Stand-alone or Improving service
part of delivery service
IDENTIFIED PAT TERNS AND KE Y CHALLENGES

We have identified a number of obstacles besides drivers and benefits, and
have observed some recurrent patterns. These are all related to how new
technology or a new solution is introduced in the market. We will now briefly
describe these key challenges.
253
Being Part of a Solution. A specific IoT solution often tends to be a small part
of the overall solution or even “a part of the whole” overall service. An IoT
solution may be too small in order to be a sustainable stand-alone business.
For a “part of a whole solution”, a lack of understanding the overall picture
(service) may limit the potential with the new solution.
Unclear Business Context. Another observed challenge is the uncertainty
about ”in what kind of business you are”. Is the business about selling an IoT
product, or does it concern providing a service based upon the IoT product,
or providing a type of overall service where the IoT product of service is just
one component among others?
Market Position. Another challenge is the need to change and obtain insight
about the role and/or market position. For manufacturing companies, this is
typically about moving from selling and maintaining products to offering
“something as a service”.
Fragmentation. We can find a multitude of similar technical solutions, each
with it its own dedicated infrastructure, although it is deployed in the same
location for a single customer. This fragmentation leads to scalability prob-
lems when a large number of solutions need to be deployed and maintained.
This fragmentation of solutions can be observed in several areas: facility
management, factories, and homecare services.
Discussion
In this section, we will discuss and delve deeper into the identified patterns and
key challenges. But first, we will discuss the impact of the overall context using the
smart lock as an illustration. Smart locks can, of course, be part of the “lock busi-
ness.” We will discuss smart locks as part of homecare or municipality services.
SMART LOCKS: AN ILLUSTR ATION OF THE ROLE OF THE OVER ALL BUSINESS CONTEXT
Homecare Service.
There is a need in the homecare sector to manage a growing group of elderly
people without compromising the quality of care with the available human
resources. This means there is a need to develop innovations that can
increase productivity without having a greater workload; it is also important
not to lose human contact with the elderly since it is so vital for the well-being
of the elderly person. Therefore, it is necessary to streamline the peripheral
services that form part of total service production.
254
There is no technology tradition in homecare, which makes it difficult to

pinpoint what could or should be replaced by technology without sacrificing
the meeting between the care and the elderly. However, various technolo-
gy-oriented projects have begun to create an understanding of what could be
improved with technology solutions through the use of early start-up compa-
nies, (Ihlström, Eriksson, and Svensson, 2009). There has also been a new
generation of homecare workers who have a different habit of technology in
their daily lives (Griffiths, et al, 2012).
The problem today mainly consists in not seeing that vertical solutions will
lead to future problems in operation and maintenance. This means that a
municipality today can focus upon a certain solution: for example, digital
lock, instead of analysing the entire home service value chain. A digital lock
is ”part of a whole” and the supplier of the lock will never be able to expand
its business to cover the entire value chain. The digital lock is then a product
or technology solution rather than a service. The product approach is to solve
the issue “how to close this passage when needed and open when it should”.
The service approach of “who must be able to get through this passage and at
what time, who should provide an access license or permissions and in what
way, how to deal with acute access permission and under what conditions,
how should the locks fit into the support organisation, how should the access
licenses communicate with an invoice or administration system”, and so on.
The product can create added value for the homecare service provider if it
is part of a context in which its skills are utilized. This is not a problem for the
lock manufacturer; for homecare service, however, it will be once new technol-
ogy solutions are implemented and exist in parallel with the user’s home. This
can force the homecare service provider into different support and service
system: each with its own conditions; sometimes, it can even be necessary to
have the same data in different systems. For example, if the lock system has a
database of users and their access rights, then a time planning system has the
same users in another database. That means that same information must be
maintained in different systems. Without the possibility to only use one single
database for all systems, this will eventually lead to error and different user
information in the different system. The users quit or change phone numbers
and so on; this is information that must be added in all the systems that use
this kind of information. Information gaps between the different systems will
occur sooner or later, which can be crucial for the safety of the providers’
255
operations. Homecare services need to change strategy; instead of implement-

ing an independent solution one at a time, they will need to look at the over-all
picture instead (Laya, 2017) and analyse which parts of the value chain can be
replaced by a more efficient sub-service (such as a digital lock system).
With a different approach, the homecare service owner and the principal
can ask different questions to the market. In other words, instead of asking for
the “best digital lock”, ask for “the best and most effective delivery of home
services to the users.” However, the latter requires that you have a clear
understanding of how to integrate sub-businesses with each other in order for
the whole to work. One becomes aware that each digital product must be
included in a context that makes it a service and that some parts of the context
may be part of another digital service (Miori and Russo, 2017). If the focus is
only upon the lock as a product solution, then it misses the importance of the
context into which the locks must fit. It is important to understand that the
upfront investment in a product is outweighed by the support and mainte-
nance costs for the management over many years.
Digital locks must be installed, configured, keys need to be distributed, and
permission is to be given that may change over time; the lock needs to be
maintained, and the keys are included in an administrative tool either man-
aged by the principal or as part of the service. In case of problems, there is a
support service that can be shared by other digital services, and so on.
The Smart Lock in a Homecare Context

Each actor who delivers a shared service to the homecare service will try to
understand how to expand its own service within the value network in order
to be more competitive. Features such as log-in/log-out and time reporting are
linked to the lock solution itself. Other discussed features are time planning for
the homecare staff and a shared dairy about the elderly patients (Markendahl
and Laya, 2014).
One example is the Swedish lock manufacturer Phoniro12, which purchased
the digital video service provider Vision and now supplies these two services
on the common platform, Phoniro Care, thus, expanding its home business
service. The company’s ambition is to become a supplier of complete solutions
for the care of the elderly. Digital cameras, camera surveillance, care tools,
and DoseSystems for medical management are available in their current
12 http://www.phoniro.se/
256
product portfolio. Phoniro has evolved from a digital environment: where

locks were their first digital solution for elderly care.
Other lock manufacturers have given up the elderly homecare service market
today because Phoniro has such a big headline amount that it is difficult for
other lock actors to compete: not necessarily because Phoniro offers the best
lock solution. What Phoniro does provide, however, is an understanding of the
homecare context; it knows that homecare has several situations that could
benefit from a digital solution. Phoniro builds an ecosystem on its platform. The
company did not come up with the camera surveillance; instead, it acquired it.
Phoniro did not build the DoseSystem; it simply has the right to include it on its
platform. By doing this, the company strengthens its offer and shows it can
handle several solutions in one platform. Thus, Phoniro has a very strong USP.
Only other players in the same segment are in competition for locks in the
home service market, such as Tunstall13. Founded in the UK, the company
began to deliver security alarms in Sweden. Both Phoniro and Tunstall
expanded within the home-service industry: the former from a now well-estab-
lished lock service; the latter from a well-established security alert service.
Figure 12.2 illustrates a simple analysis of the homecare service business
where the two routes of expansion of the two companies are depicted.
Homecare Services
Tunstall
Core Business Social alarm
Tunstall
Social alarm & Lock Expanded Business
Phoniro
Core Business Lock
Phoniro
Expanded Business
Lock & Night vision
Company X
Core Business Night vision
Company X
Figure 12.2. Examples of development paths for companies in the homecare business.
13 http://www.tunstall.com/
257
THE SM ART LOCK IN A MUNICIPALIT Y SERVICE CONTE X T

The field of elderly homecare is only part of the municipality’s activities in an
area. The municipalities also run schools and other community service facilities
in which it also uses similar services (see Figure 12.3). Using all similar services,
such as locks, the municipality turns to different users – even to the same one -
within a given area or neighbourhood. Therefore, it is preferable to make an
overall view of the need for locked areas before any lock system is purchased.
HCS
School
BR Lock Day
Nursery
Society
Service 4
Figure 12.3. Example of society services in a geographic area or district.
Another level up reveals several players in a particular geographical area

using similar services. There may be reasons to ask who is best suited to offer
the access service discussed in the aforementioned examples.
On the other hand, a company in the homecare business or a school manage-
ment may be appropriate if the municipality chooses to procure access on the
lowest level. The municipality, however, will find that it has an access provider
for each of its services: one for healthcare and another for school, and so on.
On that level, no company or organiation without access service as its core
business is probably the most suitable for schools, daycare centers, health
centers, stores, and so on.
Perhaps the best solution would be to make a deal with an access operator
that takes care of all locks and delivers access permission upon the terms set
up by the municipality? In order to make such a judgment, the whole picture
needs to be clear.
258
There are many questions that must be answered before selecting an appro-
priate sub-service provider; if they are not analysed, the municipality will pay
more than necessary for parts of a particular service to be performed in par-
allel organisations.
In summary, this discussion of the invention of ”the digital lock” illustrates
several challenges in relation to the lock business, homecare, and municipality
services. Moreover, the analysis shows how these challenges are interrelated.
Comments on the Identified Key Challenges

A SM ALL OR BIG PART OF THE OVER ALL SOLUTION
One of the challenges we identified was the need to understand if the new
technical solution represents a small or big part of the overall resolution. This
also includes determining whether the solution can solve a small or big part of
the overall problem. Most of the results with a connected device contribute to
an increased awareness, which enables the user to use its resources more effi-
ciently. This applies to the screw connection monitoring: toilet cleaning, and
homecare staff using cameras for nightly check-up of the elderly patients. One
type of value in all cases is that no staff is sent out if it is not necessary.
In the case of connected bolts and waste bins, looking into existing activities
can identify the benefits. If one is aware, one does not need to send out people
“just to check”. In the soap dispenser case, the situation is more complex. The
full value of the information from the soap dispenser is revealed when it is
combined with other information (visitor counting data) and put into an over-
all context. Hence, we can conclude that a solution that solves a limited part of
a bigger problem can be very beneficial if it is put into the overall context.
Conversely, the added value is little or not clear if the overall picture is
missing. Using the soap dispenser case again, substantial value can be seen
when it is combined with visitor data, especially during periods with a high
and varying number of visitors such as an amusement park. However, the
potential value is much lower if the solution is used as a stand-alone feature in
an environment with less frequent users: for example, an ordinary office.
IN WHAT BUSINESS ARE YOU?

For some solutions, such as with the connected bolt and the connected waste
bin, it is quite clear in what business you are: installation monitoring and
259
waste management. The answer is not so clear when it concerns the con-
nected soap dispenser and the complete EasyCube concept. In these case, it
could be any of the following types of businesses:
ÜCleaning of toilets
ÜCleaning of facilities
ÜFacility management in general
ÜSelling connected products to facility managers
ÜOffering toilet refill services (soap, towels, toilet paper)
ÜOffering refill services for a building (coffee, copy paper, and so on)
ÜSelling connected devices in general
ÜSelling systems for staff resource planning
By comparing SCA and StrainLabs, we see that SCA offers products and
solutions that enable a more efficient cleaning operation for the customer;
whereas StrainLabs makes use of the connected bolt in order to offer a com-
plete and new type of service.
ON FR AGMENTATION AND STOVEPIPE SOLUTIONS

The fragmentation with so-called stovepipe solutions has been identified for
many years. Well-known examples from smart cities, facility management,
and social care services are reported in research (Markendahl and Laya, 2013).
One actor typically provides each solution, thus, solving a specific task or
problem. The solution includes sensors, service platform, and communication
infrastructure. The main problem is the complexity to deploy and maintain
all multiple solutions. Clear evidence can be found from homecare services.
Representatives from the IT department of a municipality in Sweden say: “It
is not feasible for us to maintain this multitude of different systems. An elderly
person may have three or four different systems in their apartment and we
have to ensure that they all work”. These types of systems are typically a
social alarm, digital lock, and camera for night surveillance (see Figure 12.2).
Findings from the current research confirm that fragmentation remains a
key characteristic. In each of the different studied sectors, we have found new
or remaining cases of parallel stovepipe solutions. One example comes from
factories with multiple robots or machines in an assembly line; the manufac-
turer of a given machine is connected to and monitors each machine. No
common overall picture for the factory owner is provided since the machine
260
manufacturers do not want to share data. Another example from sport and
health is the multitude of sensors and apps that use different platforms. An
individual may need to have several sensors, each reporting data through a
dedicated app with its own login and interface.
We tried to determine at the workshops the motivation and drivers for the
fragmentation. Most actors did recognise the pattern of fragmentation. Some
of the expressed reasons include the following:
i) Fear of losing or changing the customer relations
ii) Distrust among actors to share platforms and data
iii) Hesitation to give one’s “own” data to others
iv) Lack of motivation to change one’s own business model
A common solution to the fragmentation problem is to introduce open and

shared service, platforms, and infrastructure. There are a number of projects
and initiatives about open and common platforms. In Sweden, we can find
examples from transport14, healthcare15 and smart homes16. This technical
approach is usually combined with the idea of a neutral third party managing
the shared and open platform; however, it is not clear what the business
model for the use and operation of a shared platform looks like. One key
question is the following: Can anyone be the trusted third party or should
service providers (offering end-user services making use of the platform) be
excluded from taking that role?
Another approach allowing co-existing company-specific solutions is to
agree upon a standard for data formats and interfaces. In one of the aforemen-
tioned cases, leading truck manufacturers developed the FMS standard (Fleet
Management System). Hence, transport companies with vehicles from differ-
ent truck manufacturers can more easily handle a multi-vendor fleet.
We also want to mention here the solution used by mobile parking ticket
providers. There is currently a multitude of providers and solutions for mobile
phone parking in Sweden (see Figure 12.4 on the left). The common feature is
that the ticket exists as a record in the database with active parking sessions for
all parking ticket providers. The ticket control staff has handheld devices
14 https://www.vinnova.se/p/open-transport-effectiveness-service-platform/
15 https://joinup.ec.europa.eu/community/epractice/case/sweden%E2%80%99s-health-innovation-
platform-helps-third-parties-develop-healthcare-
16 https://www.acreo.se/projects/smarta-hem
261
connected to the database with active parking session and car registration
data. By entering the registration number of a car, valid parking information
is provided (see Figure 12.4 on the right). These two examples illustrate that
strategies to handle the multitude of solutions, manufacturers, and service
providers do, indeed, exist.
Figure 12.4. Ticket machine illustrating multiple payment solutions for parking (left) and a snap shot
showing the user interface of the handheld device for the ticket control staff (right). Pictures from
(Markendahl & Laya, 2013). Photo: Jan Markendahl.
Summary
The purpose of the research described in this chapter is to identify the condi-
tions for using IoT in different industries: industrial IoT, smart energy, smart
homes, smart cities, healthcare and social care, sports and well-being. The
research provides more insight into business opportunities and obstacles
surrounding the introduction of IoT (and also digitalization) in different sec-
tors. The following are the main obstacles:
ÜSpecific IoT solutions often tend to be a small part of the overall solu-
tion; hence, it may be too small in order to be a sustainable stand-alone
business.
ÜThere may be uncertainty and/or lack of knowledge regarding of which
overall services or business the IoT solution may be part.
262
ÜThere could be fragmentation due to the diversity of parallel “closed”

solutions that lead to insufficient scalability.
ÜThere is distrust and hesitation among actors to share common/open
platforms and data.
ÜThere is the fear of changing one’s own business model.
The analysis of our cases indicates that most of the challenges occur due to
the fact that the solutions initially have been developed using a single firm
business model. Potential gains with IoT and digitalization risk not being
achieved if actors continue to stick to the single-firm business model. Hence,
co-operation and networked business models need to be considered in order
for IoT solutions to be commercially successful.
In addition to collaborating with other actors in the ecosystem, it is impor-
tant to have knowledge about specific sectors, the customers, and their ways
of working. This is especially the case when the customer is an actor in the
public sector. Although public sector actors have staff for procurement and IT
support, it may be difficult or take too many resources to integrate different
technical components into an overall solution. Moreover, it will also take
resources to educate the staff and change work processes.
The situation is different when the customer is a technology-oriented com-
pany, a manufacturing industry, a telecom, or an energy provider. These
types of actors usually have both a high level of understanding technology as
well as good control of the work processes, which are often automated with
little or no human involvement. Hence, a technology-oriented company can
more easily handle and integrate an IoT solution.
Providers of IoT technology and solutions who have the insight and com-
petence about their customers and their work processes can help their custom-
ers to integrate and maintain the solution. Hence, these actors would be more
useful as technology providers to their customers, regardless of whether or
not they provide a product or a service. In summary, we believe that knowl-
edge about specific sectors and business contexts is important in order to
adapt specific technical solutions to the overall context.
263
References
Bankvall, L., Dubois, A., & Lind, F. (2016). Conceptualizing business models in industrial
networks, Industrial Marketing Management, Volume 60, January 2017, Pages 196–203
Basole, R. C. (2009). Visualization of interfirm relations in a converging mobile ecosystem,
Journal of Information Technology, 24(2)
Bouwman, H., Haaker, T., & De Vos H. (2008). Mobile Service Innovation and Business Models,
New York: Springer
Chesbrough, H., & Rosenbloom, R.S. (2002). The role of the business model in capturing
value from innovation: evidence from Xerox Corporation’s technology spin-off companies.
Industrial and Corporate Change, 11(3), 529–555.
Ghanbari, A. (2016). Competition for Mobile Service Provisioning: Is it About Infrastructures, Services or
Both? Licentiate dissertation, KTH, Stockholm
Ghanbari, A., Laya, A., Alonso-Zarate, J., & Markendahl, J. (2017). Business development in
the Internet of Things: A matter of vertical cooperation, IEEE Communications Magazine, vol.
55, No. 2
Griffiths, F., Cave, J., Boardman, F., Ren, J., Pawlikowska, T., Ball, R., Clarke, A., Cohen, A.
(2012). Social networks – The future for health care delivery, In Social Science & Medicine, Volume
75, Issue 12, Pages 2233–2241,
H. Håkansson H., & Snehota, I. (1995) Developing Relationships in Business Networks, Routledge:
London
Ihlström Eriksson, C. & Svensson, J. (2009). Co-creation in Living Labs: Experiences From Halmstad
Living Lab. Halmstad, Sweden.
Laya, A. (2017). The Internet of Things in Health, Social Care, and Wellbeing, PhD dissertation.
Stockholm. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-212548
Markendahl, J., & Laya, A. (2013), Business Challenges for Internet of Things: Findings From
E-Home Care, Smart Access Control, Smart Cities & Homes, 29th IMP Conf., Atlanta.
Markendahl, J., & Laya, A. (2014), Transformation of Home Care Services, Working
Processes and Business Models Due to Introduction of Mobile Technology, 30th IMP
Conf., Bordeaux
Miori, V., & Russo, D., (2017), Improving life quality for the elderly through the Social Internet
of Things (SIoT), 2017 Global Internet of Things Summit (GIoTS), Geneva, pp. 1–6.
Osterwalder, A., & Pigneur, Y., (2010), Business Model Generation: A Handbook for Visionaries,
Game Changers, and Challengers. New Jersey, NY: Wiley
Palo, T., & Tähtinen, J., (2013), Networked business model development for emerging
technology-based services. Industrial Marketing Management, 42(5), 773–782.
Westerlund, M., Leminen, S., & Rajahonka, M. (2014). Designing Business Models for the
Internet of Things, Technology Innovation Management Review, 4(7), 5–14.
Zott, C., & Amit, R. (2010). Business model design: An activity system perspective. Long Range
Planning, 43(2–3), 216–226
264
CHAP TER 13
Digitalization, Collective Intelligence,

and Entrepreneurship in the Care Sector
ERIK L AKOM A A
Abstract: Parallel to the formal private or public (health) care organisations

in Europe, a number of community-driven care projects have emerged. They
may supplement the formal organisations by reducing costs or provide care
to groups that, for some reason, do not have access to the formal sector. Draw-
ing upon the Ostromian theory of commons and on previous theory and
research on open software development (which share some of the character-
istics of “open care”), I use historical cases of community-driven care to
examine the prospects for such projects to help remedy the cost crisis in the
care sector. I explore under which institutional settings “open care” is likely
to emerge and when open care projects have potential to scale. It is found that
open care is more likely to emerge and prosper when it builds upon existing
organisational structures: where the participants do not need to create new
hierarchies or governance structures, and where they share common values.
Introduction
Research in health economics shows that the health sector is characterised by
high costs, a high degree of regulation, and a lack of entrepreneurship. There
is also a limited dissemination of process innovations in the care sector (Cutler,
2011)1.
Despite the fact that digitalization has fundamentally affected industries
and, in many cases, led to dramatically reduced costs, it has primarily led to
better technical equipment in the field of care: not to reduced costs or to new
forms of organisation.
1 In this context, care should be viewed in the broad sense of the term, and include child care, elder care,
health care, assistance to persons with disabilities (mental and physical), and the treatment of addiction.
265
ER I K LA KOM A A
Cutler discusses why innovation has not reduced inefficiency and waste in
healthcare as in other sectors. Examples include the slowness to adopt effi-
ciency savings and the fact that doctors waste time on routine administrative
tasks that could be provided by less-trained personnel or by information
technology.
Cutler (2011) further argues that improved production processes of the type
that has been observed in such sectors as retail, logistics, and manufacturing
are far slower in spreading to healthcare due to the lack of organisational
innovation:
“Medical care is complex, and it is natural that there will be inefficiencies in

complex settings. Indeed, in any industry where human action is important, there
are bound to be mistakes. The failure of medical care is not so much that mistakes
are made, but rather that the system has not evolved mechanisms to minimize
those mistakes. For many years, Toyota was famous for its attention to error
reduction; Wal-Mart is equally known for its supply-chain management.” 2
The explanations can be found in the current organisation and financing

model. Healthcare is almost always, and other forms of care often, funded by
third parties. This means that the business must be controlled and audited. A
public entity will rarely be able to spend money on something that is beyond
its control. The same holds true for a listed company.
A privately held company or foundation can act more freely. However,
these are bound by regulations that cover the healthcare sector in most coun-
tries. This binding provides a low level of experimentation and, in principle,
new solutions must be approved before they can be used. Meanwhile, there
are limited incentives for innovation, as the financing models imply that the
businesses receive compensation for actual costs.
Different rules within different jurisdictions also make it less possible to
scale. Although people are very similar between different countries, the
organisation of care is often nationally regulated. This organisation is due to
compliance and documentation, among other things, that constitute a signifi-
cant part of the business. This fact may explain why we do not find globally
integrated care companies. Those that do exist are often conglomerates with
separate national parts.
2 Cutler (2011 p 2).
266
DIGITALI ZATION, COLLECTIVE IN TELLIGENCE, AND EN TREPRENEURSHIP IN THE CARE SECTOR
On the other hand, new solutions have arisen outside the realm of orga-
nised care. I call these “open care”.3
Open care has similarities with open software development, as it is bottom-up
and (generally) not for profit. One type of open care projects is where people are
organised to offer care, yet do not do so within the framework of the public or
private formal institutions. This often occurs on a voluntary basis. Another type
involves projects where patients themselves participate in care, exchange expe-
riences, and pool their knowledge. A good example is Alcoholics Anonymous
in drug addiction. One can summarise open care as “community and/or collec-
tive intelligence-based care projects”.
Although it is easy to observe the connection to open source (as in open
source software), it should be noted that open care is not necessarily IT based.
IT might play a significant role in some open care projects (for example, as a
means of communication and to create critical mass); however, open care does
not need to have any connection at all to IT. Open care projects might have
existed hundreds or thousands of years before the invention of the first com-
puter. That said, most “open” projects will probably be driven by digitalization.
This chapter will focus upon open care projects that, in some way, build
upon the potential of digitalization. I will use cases collected as part of an
ongoing European research project on open care. Approximately 30 cases
from European countries currently exist. 4 Cases include both ongoing and
historically discontinued projects. My research will describe when open care
arises and make an attempt to answer the question of when open care can help
meet the challenges of care and the conditions under which open care emerges.
This chapter is organized as follows: first, I review the theoretical back-
ground on the concept of the commons and the connection to open care, fol-
lowed by a discussion of the reasons for the increasing costs of care. Then, an
empirical part will present some findings from existing open care projects.
Finally, it will conclude with a discussion on when open care projects are
likely to emerge.
3 As a concept, open care does not exist in previous research (other than as a synonym for “outpatient
care”).
4 The cases collected include community-organised clinics in Greece, care for immigrants in France,
parallel imports of pharmaceuticals in Romania, and various online forums where patients can discuss
their (sometimes rare) illnesses and, in some cases, provide feedback to caregivers and researchers
upon the side effects of medications, for example.
267
ER I K LA KOM A A
Theoretical Background
The theoretical foundation for developing this concept draws inspiration
from research on the commons (Hess & Ostrom, 2005; Ostrom, 2007; Ostrom,
Burger, Field, Norgaard & Policansky, 1999; Poteete, Janssen & Ostrom, 2010)
and open source software development. Institutional economics explains that
community self-organisation is a third method for organising activities, apart
from the traditional market and government division that, in many situations,
works well (Ostrom, 2007). Collective intelligence and open care in general
are classic examples of such commons. In many cases, self-organised commu-
nities work better than hierarchical systems; however, they do have their own
challenges. This body of research also points to practical guides on how
self-organised communities can better overcome collective action problems.
Self-organisation and basic economic models predict that conflicts of inter-
est cause voluntary collective action to fail, even when such cooperation is to
everyone’s mutual benefit. Mancur Olson concluded:
“Unless the number of individuals in a group is quite small, or unless there is

coercion or some other special device to make individuals act in their common
interest, rational, self-interested individuals will not act to achieve their common
or group interests” (Olson, 1971).
This occurs when rational, self-interested individuals have a stronger incen-

tive to free ride than to contribute to collective benefits. The collective action
problem can be theoretically shown in n-player prisoner dilemma games,
where cooperation fails despite mutual gains
The “zero contribution thesis” in public good production, however, is not
the full story. While cooperation is challenging, I also empirically observe
that many examples of successful voluntary organisation are common. The
work of Ostrom and her team showed that self-organised communities could
solve collective action problems using cooperative norms. They examined
real-life common pool resources, such as fisheries and grazing land. They
found that, over time, communities organically developed collaborative insti-
tutions to overcome collective action problems. The rules for managing
common pool resources could be monitored, and the community could
impose sanctions. The studies found that, in a setting with repeated interac-
tion and communication, social norms can replace an externally imposed set
of rules: sometimes even outperforming them (Ostrom, 2014).
268
Ostrom focused upon the commons, which is any resource to which mem-
bers of some group share access. Individuals can extract resources from the
common pool for private use, but at the risk of degrading the commons
through excess use: the “tragedy of the commons”(Hardin, 1968). One way to
solve this collective action problem is privatising the resource into parcels of
private property, while another is assigning management to a central author-
ity. Ostrom showed that groups could also cooperate and act as their own
stewards: in practice, transforming the resource into common property.
Successful cooperation is far from guaranteed and often fails. The potential
for successful self-organisation, however, is wider than the simple self-inter-
ested theory would predict. Individuals often follow norms of reciprocity and
are willing to restrict their own use of the common resource as long as most
others do the same.
In addition to trust and reciprocity, successful commons governance
requires an active community and evolving rules that are well understood
(Ostrom, Stern, & Dietz, 2003). The longer-term survival of these institutions
also requires so-called design principles. These include boundary rules,
restrictions on the use of resources, monitoring, graduated sanctions on
offences, conflict resolution, and the ability of the participants to elect leaders
and modify rules. Cooperation works because the participants monitor each
other and can sanction or exclude cheaters. Over time, social norms, inter-
nalising the preference to follow the rules, often evolve. This phenomenon
allows for high levels of cooperation, without the need for close monitoring or
costly sanctioning.
Organisational cooperation requires individuals to keep their promises to
each other. Simple theoretical models often predict that credible commitment
in negotiations is impossible without the coercive power of an external author-
ity, such as the state. Ostrom et al. (Ostrom, Walker, & Gardner, 1992) argued
that other mechanisms could also effectively enable credible commitments:
“Empirical evidence suggests, however, that individuals facing social dilemmas

in many cases develop credible ex ante commitments without relying on exter-
nal authorities”.
This was possible through repeated interaction, communication, and the abi-
lity to sanction those who acted opportunistically and broke their promises. In
this setting, the threat of sanctions could create sufficient incentive to cooper-
269
ER I K LA KOM A A
ate and often outperform other arrangements. The authors concluded that
self-governance is possible and that
“[When] individuals are given an opportunity to restructure their own situation,

they frequently – but not always – use this opportunity to make credible commit-
ments and achieve higher joint outcomes without an external enforcer.”
Defining healthcare as a commons would stretch the definition too broadly,

making it useless. There are, however, specific elements of healthcare provision
that can be viewed as common pool resources. One important example is the
provision of complex healthcare requiring the collaboration of different actors.
The Increasing Cost in Healthcare

Over the past few decades, health expenditure has outgrown the overall
economy in developed countries driven by factors such as ageing, higher
incomes, and the adoption of new technologies.
Relative to the rest of the economy, health spending was historically fairly
stable, yet began to grow rapidly around the 1950s in both the United States
and in Western Europe (Getzen, 2014). Between 1960 and 2010, health spend-
ing as a share of GDP grew from approximately 5 to 17 per cent in the United
States and from 3 to 10 per cent of GDP in Western Europe (Rebba, 2014).
Interestingly, the rate of growth of healthcare expenditure is similar in the
United States and Western Europe, albeit it remains at a lower level in Europe
(Getzen, 2014). The increase in expenditure slowed sharply in recent years,
though this is likely to be a temporary effect of the economic crisis.
The high cost increase in healthcare also affects equality by making health-
care unaffordable for low-income individuals. Even in Europe, healthcare
tends to have a significant component of private out-of-pocket spending.
Lower-income groups are, therefore, more likely to perceive a lack of access
to health services even in countries that have universal healthcare. High
expenditures have placed great pressure on public finances and created an
impetus for reform aimed at increasing productivity in healthcare in order to
maintain the long-term viability of the welfare state. The costs of healthcare
delivery vary significantly by provider and region beyond what can be
explained by quality and input costs. This suggests that many providers
produce at sub-optimal levels of productivity
270
The causes of the high cost and low effectiveness in healthcare have been
intensely debated in recent years, with no definitive answer. While this issue
is not fully understood, it is often argued that the particular characteristics of
healthcare cause unique organisation, which reduces the incentives for process
innovation, thus, creating a bias toward high cost increases (Weisbrod, 1991).
Weisbrod (1991) writes:
“To understand the markets in which healthcare is provided and financed, it is

useful to consider ways in which healthcare differs from most other commodi-
ties. First, it sometimes involves the preservation of life or, at least, major effects
on the quality of life. Second, it is a technically complex commodity that abounds
with informational asymmetries, adverse to consumers.”
One important explanation appears to be that the ethics of healthcare tend to

incentivise technological change that is focused upon increasing health qual-
ity and saving the patient regardless of the cost, rather than on lowering costs
(Weisbrod 1991). Technological improvement can either focus upon improv-
ing the quality for a given cost or decrease the cost for a given quality. There
is a strong bias in healthcare toward the former: not for technological, but
rather for institutional reasons.
Firms that invest in innovations know that providers, regardless of the cost,
would almost never deny a new treatment or drug that improves the chances
of survival. Focusing upon drugs or treatments that lower costs may not be as
profitable. Technology often decreases costs, but has had a tendency to make
health costs increase instead.
Open Care and Digitalization

In this section, I will use cases collected as part of the European Open Care
research project to outline what can be considered open care. The cases in
this study are limited to those using information technology.5
The first identified category of open care is online communities for patient
interaction.
Patient information sharing sites can play a key role as knowledge brokers
in the healthcare sector. This fact is particularly true if patient organisations
can be encouraged to participate in the platforms. As collectives with varied
members and activities, patient organisations have a unique capability of
5 For a thorough discussion on the project and the concept of open care, see (Sanandaji & Lakomaa, 2016).
271
ER I K LA KOM A A
easing the flow of information across networks (Nicholas & Broadbent, 2015).
One such online platform is PatientsLikeMe, which was founded in the
United States in 2004.6 The social media platform, which has a global out-
reach, also engages many European patients and patient organisations. The
health information-sharing site encourages users to input data about their
symptoms, environmental triggers, medication, and so on over time. The
result is the creation of ongoing medical records. Users are encouraged to
communicate with others who have similar health statuses and exchange
knowledge. PatientsLikeMe also process aggregated and de-identified data,
which forms the basis of future health advancements. In addition to providing
useful information to those who experience health issues, the mass data gath-
ered at PatientsLikeMe is also useful to increase the understanding of dis-
eases. Numerous scientific publications rely upon the data gathered by the
patient communication platform.7 In the long run, the mass data obtained
from this and similar platforms can play an important role in fostering collec-
tive intelligence in healthcare (Tempini, 2015).
The second category is multi-function health communication platforms of
which several may be considered open care. The Hungarian PraxisPlatform
is a platform that, in addition to facilitating communication between patients,
serves as a way for healthcare professionals to communicate with patients.8
The latter role is achieved through sending therapy-related information to
patients in order to increase their adherence to and compliance with medici-
nal therapy and medical device use. Through the online platform, pharmacy
care services to large patient populations can also be conducted. PraxisPlat-
form is an example of how a single platform can fill two different roles: first,
e-healthcare, through which the traditional healthcare system can efficiently
reach out to patients at typically low costs and, second, as a social patient
communication platform.
The combination of facilitating patient-to-patient communication and
healthcare sector-to-patient communication (as well as patient-to-healthcare
sector feedback) might create synergistic effects for patients in addition to
6 There are similar sites in other countries, e.g., Carenity, which is now established in several European
countries (Castejón, Chekroun, García, Gay, & Rebollo, 2013).
7 An example is the paper by Naujoks et al. (Naujoks et al., 2016), in which patient-reported data from the
PatientsLikeMe community are used to explore how migraines impact the day-to-day life of patients.
8 PraxisPlatform website, https://www.praxisplatform.hu/.
272
healthcare professionals. For example, these might occur since patients can
receive complementary information through the same platform and the bur-
den upon health professionals to reach out with online information can be
reduced if patients receive from other patients some of the information they
are seeking. Through these forums, patients can also help each other better
understand the information given to them by health professionals.
HealthUnlocked is another example of a social-patient communication
platform developed in Europe; its aims is to become the social network for
health. HealthUnlocked is a peer-to-peer support network through which
individuals with health issues can communicate safely online, with guidance
from credible institutions and organisations. Founded in 2009, the platform is
multi-functional since it also encourages patient advocacy organisations to
become engaged as well. Through HealthUnlocked, these organisations can
communicate with their members about health-related matters, as well as
allow members to foster patient-to-patient health sharing.
Communication platforms are also encouraging and simplifying open com-
munication between care providers. Hospitals and health clinics tend to be
organised in a hierarchical manner, in which communication between differ-
ent units and even between different doctors in the same hospital is often
limited. Information sharing to patients is even more limited within the tradi-
tional hierarchy of healthcare provision. Information-sharing applications
during recent years have disrupted this system by encouraging more open
communication. An example is Klara. This communication platform was
launched in 2014 and simplifies information sharing from doctors to patients.
The cloud-based web and mobile apps offered by Klara have since spread to
hundreds of health systems across the United States, including solo providers
and large medical groups. Klara has gradually moved toward simplifying
communication between healthcare workers and healthcare systems. The
company is currently attracting capital to finance future improvements. The
aim is to allow patients to exercise greater influence over the healthcare that is
provided to them, as well as allow operational efficiency in health provision
by simplifying information sharing. The example of Klara shows that open
information sharing among patients, between patients and health providers,
and among health providers can occur through the same basic platforms.9
9 PR Newswire (2016). “Healthcare Messaging Platform Klara raises $3 Million from Lerer Hippeau
and Project A to become the Central Nervous System of Healthcare”, 2016-09-14.]
273
ER I K LA KOM A A
A similar platform in Europe is ENJECT. ENJECT is a 4-year coordination

project funded by COST: a European funding organisation for research and
innovation networks.10 ENJECT focuses upon promoting new models of
healthcare delivery, thus, incorporating wireless, digital, and mobile technol-
ogies. The stakeholders in the process of health delivery are connected in
information-sharing networks. The aim is to promote the concept of “con-
nected health”, in which the patient can gather, link to, and interpret informa-
tion from different sources. Providers, patients and researchers can also use
aggregate data in order to improve decision-making. This information shar-
ing from providers to patients is relevant in the scope of open care, since the
information can be fed into social patient communication platforms and peer-
to-peer support networks.
A third category is that in which digitalization realises “long tail effects”
(Anderson, 2006). RareConnect is an international platform for rare disease
communities. The platform, which has been developed by Eurordis and
NORD, acts in cooperation with patient organisations. The organisations
contact the site managers to gain permission to set up community pages.
These pages have learning resources in the form of moderated forums and
spaces for patients to share their stories with one another. Individual patients
can also connect with health professionals in the forums. Another platform is
HealthTalk, developed in partnership between the charity DIPEx and the
Health Experiences Research Group at Oxford University. The platform
collects text and video narratives from patients who communicate their expe-
rience of having a certain disease. The experiences of individual patients are
presented on a timeline through the early stages, diagnosis and treatment,
hence, reflecting the entire patient experience. The patients included for each
disease are chosen to represent a range of disease stages, ages, genders, socio-
economic status, and so on. (Nicholas & Broadbent, 2015).
When Does Open Care Emerge?

Open care projects are unlikely to emerge where the traditional care institu-
tions work well, but rather where they are inadequate. This phenomenon may
apply either where care systems failed due to an economic crisis or in areas
where no such systems have been established: for example, care for recent
10 ENJECT website, http://enject.eu/about/.
274
immigrants or for minorities or people in remote geographic areas. They may

also emerge where scale effects exist but the number of patients in a specific
jurisdiction is too small to cater to the demands of the users or patients. The
ability of a patient with a rare disease to share knowledge with others might
be of little value if the number of patients in a country is four and the know-
ledge exchange is organised within the national healthcare system. The
value, however, could be enormous if the patient is able to interact globally
with tens of thousands of other patents.
Several projects that have been able to scale successfully are those in which
the users themselves are the prime beneficiaries. The aforementioned Patient-
sLikeMe and other social media platforms are devoted to the acquisition of
knowledge. This also applies to non-digital open care projects, such as Alco-
holics Anonymous: where their own participation is both important for the
alcoholic participant and for the other participants in the gathering. Therefore,
projects based upon self-help and exploiting economies of scale and the lack of
regulatory border obstacles may have greater potential than other projects.
Projects based upon the acquisition of big data may also be successful even
if they are likely to rely upon external incentives for participation, insofar as
data collection does not give immediate benefits to the participants. The
interest in creating such incentives may be substantial, as alternative oppor-
tunities for obtaining these data are often missing within traditional health
services.
To help solve the healthcare cost crisis, open care projects must be able to
scale. The insights from Ostrom offer some help in hypothesising when this is
likely. Most open care projects are organised as non-profits. As they do not
have a bottom line and the efficiency of the projects is difficult to evaluate, thus,
allowing room for opportunism. A strong common culture might be the rem-
edy. Historically, many projects that can be defined as open care have been
organised by religious organisations where a common set of values already
exist – a person who is involved knows what is a good outcome. This phenom-
enon also increases the costs for opportunistic behaviour from outsiders. The
cost for a person to follow religious rules and rituals is low if the person is a
believer in the faith; however, it can be costly for an outsider. Religious organ-
isations may also provide an organisational structure, thus, alleviating the need
to create one for a new project.
275
ER I K LA KOM A A
A common understanding of what is considered a good outcome also exists

in other communities. What constitutes superior performance in the open
software culture is generally agreed upon: for example, the writing of struc-
tured and commented code. Different functions are also more or less likely to
be developed in the absence of financial incentives (Von Krogh, Spaeth &
Lakhani, 2003). Research on software development, therefore, could be used
to also understand open care projects.
Conclusions
Open care is a novel concept that can be useful in understanding the forma-
tion of care projects outside formal (health) care institutions in a world where
health care costs is increasing and, in some cases, limited access to care is a
growing problem. In this chapter, I have given some examples of how
IT-based open care projects – both in the collective intelligence and the com-
munity provision type – may help solve some healthcare challenges.
As historical examples show, digitalization often, yet not always, facilitates
open care projects. Through the use of the cases and previous literature on
the commons and institutional entrepreneurship, it is possible to hypothesise
where open care projects may emerge: mainly where the traditional public or
private healthcare system is inadequate or has failed. This phenomenon
applies both to public and private systems.
That open care is easier to organise if the participants have shared values
because they can more easily agree upon what is a good result (Capiluppi &
Michlmayr, 2007) may also be assumed. For instance, this phenomenon has
been identified in open source software development: where there is a consen-
sus on what is considered to be good code and what the rewards are. The
same can be observed within collective intelligence projects, such as Wikipe-
dia: where the internal incentive and reward systems are based upon a com-
mon culture.
Open care may also foster innovation by means of lowering the cost of
experimentation. Formal care institutions are often risk averse due to the high
costs of failures; however, open projects – party as a result of their smaller
scale and the low stakes – might be more prone to experimentation. This
means, a few successes that could scale could then outweigh the cost.
Thus, open care may relieve the pressure upon the formal healthcare sys-
tem in two ways: first, by facilitating entrepreneurship and the dissemination
276
of process innovations. If some “open” solutions are able to scale, then they
might lower the cost for specific treatments or types of care. This fact is espe-
cially true when similar projects – due to, for example, high monitoring costs
if organised within formal care institutions – have a limited potential to scale.
Through the structuring of incentives, many of the projects described in this
chapter will avoid the problems described in the “commons” literature.
Secondly, even when they lack the potential to scale, open care projects can
help by providing benefits to groups that have limited access to formal care
institutions or where the participation of the patients is in itself therapeutic. In
both cases, open solutions might increase access to care without incurring a
cost to the formal care institutions. Open solutions might also increase the
quality of care by creating incentives for experimentation and innovation, also
without increasing the costs to the formal, private, and public care providers.
277
ER I K LA KOM A A
References
Anderson, C. (2006). The long tail: Why The Future of Business is Selling Less of More: Hachette
Books.
Capiluppi, A., & Michlmayr, M. (2007). From the Cathedral to the Bazaar: An Empirical Study of the
Lifecycle of Volunteer Community Projects. Paper presented at the IFIP International Conference
on Open Source Systems, 2007.
Castejón, N., Chekroun, M., García, J. M., Gay, C., & Rebollo, P. (2013). Patient Networks as
a Data Source for Patient Reported Outcomes Research. Carenity Experience. Value in
Health, 16(7), A608.
Cutler, D. M. (2011). Where are the Health Care Entrepreneurs? The Failure of
Organizational Innovation, Health care Innovation Policy and the Economy, Volume 11 (1–28):
University of Chicago Press.
Getzen, T. (2014). Measuring and forecasting global health expenditures. in Scheffler (ed.)
World Scientific Handbook of Global Health Economics and Public Policy, vol 3. World Scientific.
Hardin, G. (1968). The Tragedy of the Commons. Science, 162 (3859), 1243–1248.
Hess, C., & Ostrom, E. (2005). Understanding Knowledge as a Commons: From Theory to Practice.:
MIT Press
Naujoks, C., Olson, M., Simsek, D., de Vulpillieres, F. d. R., Vo, P., Kendall, K., & Cerrato, D.
(2016). How do Migraines Impact Patient Day-to-Day Life? an Exploratory Analysis of
Patient Reported Data from the Patientslikeme Community. Value in Health, 19(7),
A438–A439.
Nicholas, L., & Broadbent, S. (2015). Collective intelligence in patient organisations. : Nesta
Olson, M. (1971). The Logic of Collective Action : Public Goods and the Theory of Groups: Harvard
University Press.
Ostrom, E. (2007). Institutional Rational Choice: An Assessment of the Institutional Analysis
and Development Framework. In P. Sabatier (Ed.), Theories of the Policy Process (pp. 21–64):
Westview Press.
Ostrom, E. (2014). Collective Action and the Evolution of Social Norms. Journal of Natural
Resources Policy Research, 6(4), 235–252.
Ostrom, E., Burger, J., Field, C. B., Norgaard, R. B., & Policansky, D. (1999). Revisiting the
Commons: Local Lessons, Global Challenges. Science, 284(5412), 278–282.
Ostrom, E., Stern, P. C., & Dietz, T. (2003). Water Rights in the Commons. Water Resources
Impact, 5(2), 9–12.
Ostrom, E., Walker, J., & Gardner, R. (1992). Covenants with and without a Sword:
Self-governance is Possible. American Political Science Review, 86(2), 404–417.
Poteete, A. R., Janssen, M. A., & Ostrom, E. (2010). Working Together: Collective Action, the
Commons, and Multiple Methods in Practice: Princeton University Press.
Rebba, V. (2014). The Long-term Sustainability of European Health Care Systems. Dipartimento di
Scienze Economiche” Marco Fanno”.
278
Sanandaji, T., & Lakomaa, E. (2016 ). Care, Commons and Entrepreneurship,. SSE Working Papers
in Economic History 2016:2. Stockholm School of Economics.
Tempini, N. (2015). Governing PatientsLikeMe: Information Production and Research
through an Open, Distributed, and Data-based Social Media Network. The Information
Society, 31(2), 193–211.
Weisbrod, B. A. (1991). The Health Care Quadrilemma: an Essay on Technological Change,
Insurance, Quality of Care, and Cost Containment. Journal of Economic Literature, 29(2),
523–552.
Von Krogh, G., Spaeth, S., & Lakhani, K. R. (2003). Community, Joining, and Specialization
in Open Source Ssoftware Innovation: a Case Study. Research Policy, 32(7), 1217–1241.
279
CHAP TER 14
AgTech and the City: The Case

of Vertical Farming and Shaping a
Market for Urban-Produced Food
M ARIA J. BUSTA M ANTE
Introduction
The world population is expected to reach nine billion by 2050, and the UN
Food and Agriculture Organization predicts that food production must
increase by 70 per cent to feed the growing population (Beecham Research,
2016). With pressure to not only produce more food, but more sustainable
food, agriculture is undergoing a digital transformation as it seeks to use
technological solutions to increase yields while reducing food loss and nega-
tive environmental effects. Vertical farming, which grows products in verti-
cally-stacked layers, is an emerging trend showcasing how technological
innovation can lead to new solutions for food production. This is part of the
urban agriculture movement that strives to produce food closer to metropol-
itan areas where an increasing proportion of people live. Vertical farms utilise
new growing techniques, such as hydroponics: growing plants in water, and
aeroponics: providing water and nutrients via a spray mist. This, along with
technologies such as LEDs, sensors, and software enable farmers to con-
stantly monitor and adjust nutrients, water, and temperature to maximise
efficiency and optimise the plant’s nutritional content. AeroFarms, based in
the United States, boasts using over 30,000 data points to monitor not only
the environment around the plants; it also controls the colour, texture, nutri-
tion, and flavour (http://aerofarms.com/). The technological efficiency cou-
pled with the elimination of weather, and other potentially harmful external-
ities enables the farm to be 130 times more productive from a crop-yield per-
spective to the equivalent field farm (Ryan, 2017). This digital transformation
281
M A R I A J. BUSTA M A N T E
of food production does not disrupt the current agricultural network; instead,
it has the potential to develop a whole new one since the rural setting is
exchanged for an urban one. Actors, including urban planners, city govern-
ments, and real estate developers now have a role to play.
In order for the technological innovation to be successful, it must find a
market. In other words, although technical solutions to produce safer and less
harmful agricultural goods in cities do, indeed, exist, it is not assumed that
the consumer and the larger agriculture system or other relevant industries
will support it. Vargo et al suggest: “Market innovation does not occur auto-
matically when actors (e.g. firms), or groups of actors (e.g. innovation net-
works) introduce new ideas or products, but only when new practices (i.e.,
solutions) become institutionalized” (Vargo, Wieland, & Akaka, 2015). This
implies that it is too narrow to consider innovation strictly from a technology
perspective. There is also a need to expand the range of innovation to include
the relationships, processes, and collaborative initiatives that ultimately lead
to market innovation (Vargo, et al, 2015).
Founders of vertical farms are redefining how we think about food produc-
tion. In order for it to become institutionalised, they will need to influence the
broader ecosystem of stakeholders. The purpose of this chapter is to explore
the different dimensions of market innovation that complement the techno-
logical innovation of vertical farming by seeking to answer the following
question:
What activities are the founders of vertical farms performing in order to influence market
innovation?
This chapter will first outline an analytical framework that will be used for
discussion, and then present two cases of vertical farms in the Swedish mar-
ket. The chapter will then discuss the activities that these market entrants are
using to affect market innovation using the model. By outlining the activities,
we can then begin to see which relationships and collaborations are vital for
these farms: that is to say, the ecosystems that vertical farms are establishing
in order to build a market for their products. The chapter concludes with a
summary based upon the findings and future market considerations as we
rethink food networks and production methods in a digital era.
282
AGT ECH A N D T H E CI T Y
Concepts and Analytical Framework

The key concept and model introduced here will establish a framework that
will be used to discuss the cases.
SERVICE ECOSYSTEMS APPROACH

Ongoing negotiations within a firm – that is to say, the vertical farm – as well
as within the broader ecosystem, which includes a number of stakeholder
institutions, shape both technological and market innovation. By approaching
innovation in this way, we recognise that there is more to the success of an
innovation than just the binary analysis of the firm/innovator and customer/
adopter. A number of actors and institutions will, indeed, play a role.
The service-ecosystems view provides an integrative approach to market-
ing that examines both goods and services. In this context, innovation is said
to be “driven by collaborative efforts to find or develop new ways to create
value” (Vargo, et al 2015). This interpretation consolidates the views of tech-
nology and market innovation and defines them as follows:
“Technological innovation is the co-creation of new value propositions, or collective,

combinatorial evolution that leads to the generation of new, potentially useful
knowledge.”
“Market innovation is driven by and drives the development of new technologies,

but also requires the acceptance of a value proposition as well as the continued
exchange, integration and application of a particular technology among multiple
actors, over time (i.e. institutionalization).” (Vargo, et al 2015)
To consider vertical farming through a service ecosystems lens, we must first

acknowledge that farming has become more than just a “goods” business.
Food has transformed from a survival product to one that encompasses a
whole range of values: what it means to be organic, locally grown, all natural,
and so on. While food remains a product, vertical farming is introducing a
new value proposition. In order to institutionalise the idea that urban-
produced food is an important part of our future food system, actors will need
to engage in “institutional work and co-create institutions through multiple
iterations” until shared value is developed (Vargo et al, 2015).
283
M ARKET PR ACTICE MODEL

Markets tend to be defined by relationships that are already established,
which means actors must work to change “mental barriers” that inhibit cur-
rent actors’ ability to see the market differently (Storbacka & Nenonen, 2011).
As a result, market actors who wish to redefine or reshape a market must
undertake a number of activities. In the case of vertical farms, the founders
seek to shape a market for urban-produced food. To help outline the activities
of the vertical farms, I will use the market practice model developed by
Kjellberg and Helgesson (2007). The framework outlines three linked activi-
ties the authors call exchange, normalising, and representational practices that are
executed continuously. Exchange practices include the activities related to
economic exchange such as presenting products, negotiating prices, advertis-
ing, and so on. Storbacka and Nenonen extend this definition to include the
idea that exchange practices are also the activities that shape an accepted and
shared value proposition (2011). Representational practices include the activ-
ities that look to depict markets and/or how they work. Lastly, normalising
practices are defined as activities that seek to establish guidelines on how a
market should be shaped according to some group of actors. This can include
norms and rules: such as technological standards, codes of conduct, and so on
(Kjellberg and Helgesson, 2007).
The model will be used to outline the activities vertical farms are perform-
ing to overcome the mental barriers of current actors and develop a shared
value of urban-produced food aided by digital technology.
Overview of Vertical Farming

Food has been produced “inside” in greenhouses for quite some time. Vertical
farming builds upon the concept of greenhouses and incorporates the urban
setting due to technology providing the opportunity to produce food in an
untraditional environment: that is to say, without sun or soil inside different
types of structures. While greenhouses have made strides in technological
advancement and have been effective in producing products in bulk, they con-
tinue to work within their established networks along the value chain. The
urban location of most vertical farms today is the catalyst for the need to evalu-
ate the network and ecosystem for food production and distribution in the city.
Urban agriculture has been around for years: in the form of rooftop and
community gardens. The concept of vertical farming, however, is still rela-
284
tively new. Vertical farming is defined as “the practice of growing plants in

vertically stacked layers, vertically inclined surfaces and/or integrated other
structures. The modern idea of vertical farming uses Controlled Environ-
ment Agriculture (CEA) technology, where all environmental factors can be
controlled” (Association for Vertical Farming, n.d.). In 2010 when Despom-
mier wrote his seminal work on vertical farming, there were none in existence
(Despommier, 2011, pg. 4). Today, however, vertical farms of various size and
set-up can be found in countries all over the world: including the United
States, Japan, the UK, Singapore, and Sweden. They can be found in repur-
posed warehouses, newly constructed buildings, and shipping containers.
Some strive to sell to wholesale markets; others target consumers, while oth-
ers exist solely for the purpose of educating and training. Water efficient
growing techniques, such as hydroponics and aeroponics, coupled with
technologies such as LEDs, sensors, and software enable farmers to con-
stantly monitor and adjust nutrients, water, and so on in order to maximise
efficiency. The main output tends to be leafy greens and herbs, though efforts
to expand into other crops and fish are underway.
While vertical farming is new to Sweden, a few entrepreneurs have seen
an opportunity for the technology and have entered the market. Based upon
semi-structured interviews with founders and secondary research, the follo-
wing sections will outline two cases of vertical farms being developed in
Sweden today.
The Case of Plantagon

I said that this is almost impossible because there is absolutely no market for vertical farming.
There are no clients. (Hans Hassle, Co-founder, Plantagon)
Headquartered in Stockholm, Plantagon positions itself as a “global innova-

tion pioneer”. Much of Plantagon’s work has focused upon building relation-
ships with stakeholders and communicating the possibilities of vertical
farming through summits and other presentation opportunities. Meanwhile,
it has grown its number of patents and initiated a standardisation process for
urban food production. Plantagon develops and operates urban farms; its
business model is based upon management and performance fees. Interna-
tional awards and recognition have followed Plantagon’s ambitions and
activities. While plans and agreements for vertical farms exist in Sweden,
285
Singapore and China, the company has faced challenges regarding zoning
and financing; no fully functional vertical farm has come to fruition to date.
BACKGROUND
The seeds for Plantagon were cultivated in 2002 as a project led by Hans
Hassle. Other partners included Sweco, Åke Olsson (innovator) and the
Onondaga Nation, a Native American entity in the United States. The goal
was to develop a sustainable business for high-technology food production.
Each entity brought a different perspective to the table. Sweco was able to put
Olsson’s concept of a vertical farm into the context of their sustainable city
project, thus, growing food with limited land. The Onondaga people saw an
opportunity to protect food sources in the face of climate change. After long
discussions and negotiations, Plantagon was officially incorporated as a com-
pany in 2008, and was set up as a hybrid company: a non-profit organisation
(Plantagon International Association) and a for-profit company (Plantagon
International AB). The goal of the model is to find a balance between the
commercial and ideological driving forces of a company.
An in-depth visibility study followed: focus was placed upon building a
brand and patent portfolio rather than a prototype that would be expensive
and obsolete in just a few years. Building a strong brand would enable the
company to apply it to other technologies and pursue high-impact partner-
ships. The company approached the Swedish clean tech industry and con-
tracted large companies to do research and development since they already
had the teams, knowledge, and ideas. In this way, the organisation was able
to uncover challenges quickly: such as climate control in a vertical building
or thinking through water, logistics, and distribution channels. Energy would
become a big challenge. Plans were almost abandoned in the middle of the
visibility study in 2008–2009, due to the energy costs needed to make the
concept work. However, a turning point came when Tekniska verken in
Linköping contacted the company. As one of the biggest energy companies in
Sweden, it had been working with renewable ideas and sustainable waste
management for years and wanted a project that would showcase its work to
the public. Tekniska verken was interested in Plantagon’s vertical greenhouse
concept and wanted to connect the greenhouse with existing infrastructure.
286
“They offered us the spillage heat for free. So then we took down the energy
consumption…and then it made sense economically.”
(Excerpt from interview with Plantagon)
With the agreement, Plantagon had a viable solution to produce food vertically
in a controlled environment. The pitch was adjusted from a stand-alone vertical
greenhouse to an integrated façade system, which combines the vertical green-
house concept with a normal real estate project. The goal was to show the
potential for the vertical greenhouse to add value to real estate projects.
Rendering of Plantagon’s World Food Building in Linköping, Sweden.

Photo credit: fyyr/Henrik Vesterberg.
In February 2012, Plantagon held the ground-breaking ceremony for the

Plantagon World Food Building in Linköping, Sweden. This has been the
company’s premier project, embodying the integrated façade system with a
greenhouse on one side and an office building on the other: where people and
plants will mutually benefit from the building’s airflow. Once completed, it is
287
estimated that the building will produce 500 tons of vegetables per year in a
60-metre tall building (http://www.plantagon.com/exhibition/). The concept
incorporates automation, vertical production, and Plantagon’s patented uPot,
which optimises the indoor environment for plant growth. In addition, the
project includes a number of partnerships in order to develop integrated solu-
tions for energy, waste, water, and carbon dioxide.
Plantagon AB has driven many of the above activities: the for-profit com-
pany set up as part of the hybrid structure. The organisation drives all com-
mercial activity: including the development, and sales of Plantagon’s technol-
ogy, as well as the discussions with potential partners in the public and pri-
vate sector.
Non-profit corporation, Plantagon International Association, focuses upon
opinion-shaping activities, such as seminars, education, and lobbying to bring
to the forefront the issues and possibilities of urban agriculture. Vegetables
produced indoors cannot be classified as organic under current regulations
due to the lack of soil use, even though the food is produced without chemi-
cals or pesticides. This became a market and branding problem; therefore,
Plantagon realised there needed to be a standard for urban-grown food to
ensure safe production and to build consumer understanding.
“… So we arranged a seminar in Brussels, which was the first on urban agricul-

ture... [To discuss] what is really meant with sustainable food production in cities.”
These global activities with city officials, governments, and international

bodies help to define and establish what is vertical farming, as well as its
potential for both business and society. The non-profit’s current membership
stands between 300 and 400 people/organisations.
THE PRESENT
While Plantagon continues to work with local partners to build the World
Food Building, it has begun to focus upon other types of projects as well. One
of the big challenges for the company has been to convince real estate compa-
nies of the economic viability of incorporating food production in buildings.
The conversation has been more about the business of food production rather
than a specific technology.
288
“… We’re building a new part of infrastructure in cities…it’s much bigger than a

simple business plan. Of course we have business plans and excel sheets and all of
that…but [investors and partners] need to share this extreme long-term perspective.”
Plantagon realised it needed to show that indoor urban food production could
be profitable. The company plans to launch its first full-scale energy smart
farm in 2017 in a completely controlled underground environment in Stock-
holm; this will be done in partnership with Fabege, the owner of the building
where the farm is being built. The produce will be sold in neighbourhood
markets and in the building; this will be the first time the new urban food
label is used. The automation and vertical technology upon which Plantagon
has built its patents, however, is not the best option to produce in the under-
ground environment. Yet, the company strategy to build its brand as an
expert in the conceptualisation and design of indoor food production, rather
than tying itself to a specific technology, has enabled it to work with partners
to find the best solutions. Hence, it is working with a company it believes has
the best technology for the underground environment.
THE FUTURE
Plans for the future include adding at least nine more production facilities in
office buildings in Stockholm and ten more in residential buildings through-
out the country. The company has also noted that it has transitioned from
research and development into agritechture, or a “combination of agriculture,
technology, and architecture” to safely and locally produce food in cities
(http://www.plantagon.com/about/business-concept/agritechture/).
With the new planned projects, Plantagon does not see real estate as much
of a challenge moving forward. Once production begins to hit the market, it
will be a matter of who will buy the products. There will also be new owners,
which the company is currently negotiating.
The Case of Grönska

Every part of the value chain has new things we have to, maybe not entirely reinvent, but
rethink and perhaps do differently. (Natalie de Brun Skantz, Co-founder, Grönska)
289
Grönska is a vertical farm start-up located in Hammarbyhöjden, just outside

downtown Stockholm. The company is building out its growing system in the
basement of an existing office space. The team currently consists of a grower/
hydroponic systems expert, an architect, and a business and marketing pro-
fessional. Grönska is the first to grow plants vertically and sell them on the
market in Sweden. The company currently owns the whole value chain: from
technology development through to distribution. As the team learns about
the market by doing and iterating, Grönska sees working on a small scale as
a key advantage: not just with how they produce plants; benefits lie with
pricing strategies, business models, and distribution as well. Grönska’s goal is
to make the city self-sustainable.
BACKGROUND
The seeds of Grönska were planted in 2014 when friends Petter Olsson and
Robin Lee rented space in Hammarbyhöjden and began experimenting with
growing plants hydroponically. They welded and built their own LED light
systems to optimise what was currently available on the market. Grönska was
officially registered as a company in 2016 when Natalie de Brun Skantz joined
the team. That same year, they began selling their first products at Paradiset:
the largest supermarket provider of natural and organic products in Sweden.
“Getting out to market has allowed us to have dialogue with customers, to try
packaging, to try logistics…. And also to understand how groceries work…
Paradiset is very generous with information, so it’s more of a cooperation than
us just being a supplier.” (Excerpt from interview with Grönska)
The relationship between the two companies has been instrumental for
Grönska. They work together on pricing in an attempt to find the right level
for Paradiset, Grönska, and the end consumer.
While grants and scholarships from entities such as Almi, Stockholm Busi-
ness Region, and Stockholm Venture Cup have helped bring in some money,
the company is largely bootstrapped by the owners. Other revenue comes
from opportunistic activities: speaking engagements, consulting on architec-
ture, and real estate projects interested in incorporating food production in
their designs.
290
THE PRESENT
Grönska’s main product is basil, which is grown in pots at a capacity of about
600 plants per month under its proprietary LED lighting system. Beyond the
lighting system, Grönska uses a sensor and app to monitor and adjust envi-
ronmental parameters in the farm: such as carbon dioxide, temperature, and
humidity levels. Other parts of the process – including harvesting – are done
manually. The company currently sells about one tenth of the yield and uses
the rest for testing to optimise growth and nutrition. Growing plants in pots
was a conscious decision, since the Swedish consumer is accustomed to buy-
ing herbs in this way. In addition to Paradiset, the company also sells to
Centan: a restaurant focused upon using locally-produced products. Through
in-store promotions and demonstrations at events such as Smaka på Stock-
holm, Grönska has also invested time in talking to end consumers.
“Feedback from customers has been super positive, but there is initial confusion
about why we can’t say we’re organic even though we don’t use pesticides.”
(Excerpt from interview with Grönska)
Educating customers on how the product is grown and why it does not carry
the organic label can be time consuming. Grönska believes there is a need to
distinguish between urban-produced food and other available products; it has
supported efforts from Plantagon to establish an urban food label that would
clarify these differences for consumers.
The company is targeting wholesale, retail, and restaurants, with other
customers in the pipeline. There is opportunity selling to wholesalers and
restaurants that desire food produced in Sweden, which is often difficult to
find given the harsh winter climate. Sweden traditionally has approximately
three harvests per year: Grönska, however, is able to have 12: amounting to
one per month. In addition, its products have a consistent look regardless of
whether it is summer or winter, which is not the case for greenhouse or other
traditionally-produced goods. Prices are negotiated with the retailers and are
set at “a bit of a premium, but still competitive in price.” Grönska admits that
it is still experimenting and negotiating for the best price level. The challenge
has been how to balance growing customer interest with growing production
capacity.
291
Herbs growing in Grönska’s farm: just outside downtown Stockholm, Sweden.

Photo credit: Grönska.
Experimentation and iterations can be found throughout the whole process.

The substrate or material that anchors the plant for growing hydroponically
in the production phase is a big point of discussion. Grönska started with soil,
then tested a number of materials: from paper pulp to fabrics to coconut coir.
In the end, they have returned to using soil for the time being. The reason for
this is that it comes down to not only what is organic and cost-effective; it also
depends on what is available nearby. For example, coconut coir would need
to be imported from the Caribbean, thus, negating any positive impact of
local food production.
With regard to the business model, Grönska is thinking about different
options and considering whether or not customers are willing to book orders
beforehand, so the company can schedule production accordingly, thus,
reducing waste or whether it will need to find a secondary market for unsold
products. Grönska is also currently handling distribution; as production
grows, the experimentation of models in this part of the value chain will also
take place.
292
Grönska is currently building a larger capacity production system that will

enable it to produce up to 5,000 plants per month. Half of these will be sold;
the other half will be reserved for testing. Automation will also soon be tested
in some aspects of the production. Due to costs, however, full automation will
not be implemented until a third planned phase of expansion. Grönska deems
the current construction of a larger scale proof of concept to be instrumental
in developing its cost structures and production model. Plans are underway
to have the next capacity phase up and running by the end of 2017. Parallel to
these activities, the company is participating in events and lobbying the gov-
ernment about the importance of food production to help build self-sustaining
cities. Grönska also hopes to form partnerships with important actors in the
agriculture sector, such as the Swedish Board of Agriculture ( Jordbruksver-
ket) and others.
“We have spoken with LRF [the Federation of Swedish Farmers]. We think we
should be incorporated in the group. There’s no difference between us and a
field farmer.” (Excerpt from interview with Grönska)
THE FUTURE
As the farm grows, the plan is to introduce automation in planting and har-
vesting as well. In the long term, Grönska hopes to target crops that have a
large import rate to Sweden, cutting down on greenhouse gas emissions and
enhancing the flavour of products available to the Swedish consumer. The
company also plans to move into other types of packaging, as well as having
nine larger growing facilities in Sweden and then expand to other countries
in the Nordics and Europe. Grönska is also testing a more distributed busi-
ness model in addition to its in-house production that would place its technol-
ogy in offices or stores, thus, enabling on-site production. While it believes
the market is not there yet, it does have a test system installed at Hyper
Island: a school in Stockholm.
A number of challenges remain. Distribution is one of them. Currently, the
majority of vegetables in Sweden, even if grown near Stockholm, go through
Helsingborg. Ideally, Grönska wants to distribute directly. Due to the poten-
tial large volumes, however, this might not be possible. The aim will be to
balance the benefits of building new networks or joining existing ones.
293
Discussion
The two cases of vertical farms in Sweden show that the companies are tak-
ing different approaches in establishing their respective businesses and build-
ing a network to support their product. Plantagon is focused upon designing
large-scale projects with state-of-the-art technology; it has worked on a local
and international level with a focus upon high-level organisations and con-
tacts to grow the acceptance of vertical farming prior to bringing a product to
market. Grönska is a start-up that has concentrated upon building relation-
ships with sellers and the end consumer in Sweden in order to bring product
to market on a small-scale; it iterates and makes improvements along the way
in operations and technology. The conceptual model of markets (Kjellberg
and Helgesson, 2007) will be used in this section to outline and discuss the
activities that Plantagon and Grönska are performing to encourage other
actors to accept the value proposition of urban-produced food; these include
real estate partners, governments, consumers, among others.
Representational practices are those activities that look to reduce the ambiguity
around a market. The concept of vertical farming is new to potential custom-
ers, sellers, and partners. Therefore, both companies have worked to shape an
understanding of it by engaging with a number of different actors: govern-
ment officials, energy companies, real estate companies, supermarkets,
restaurants, distributors, and end consumers. This is achieved through meet-
ings, events, and technology demonstrations. The difference here though is
the geographic target. Grönska is more focused upon presentation in Sweden
and on influencing the local ecosystem. Plantagon has focused more upon
building an international network: for example, organising urban agriculture
summits in Brussels and Washington.
Normalising practices help to establish guidelines and rules for how the mar-
ket should work; it also includes activities related to strategic planning. One
challenge that both Plantagon and Grönska have encountered is how to
position their products since they do not fit under current rules and regula-
tions for what is considered organic. Plantagon has been leading a standardi-
sation process that has reached the international ISO level in order to estab-
lish legitimacy and uniformity when presenting urban-produced food with
new technology. Grönska supports these efforts; however, it has not been a
focus for the company. Plantagon has also put in place structures to assist
with international strategic efforts, as well as an official board of directors
294
with backgrounds in areas such as advertising, banking, non-profit, public

sector, and sustainability. Grönska has also worked with a group of advisors
on an ad hoc basis, which also represent various sectors; these include envi-
ronmental research, construction, and the public sector.
Perhaps the biggest difference between the two cases is in the exchange
practices. With no product on the market, Plantagon has focused upon setting
up large partnerships with various stakeholders: most critically, with real
estate and energy companies. Emphasis has been placed upon negotiations of
contract terms (for example, rent, subsidised energy costs, and so on) as well
as the design of the production set-up, which would enable the co-creation of
value for the parties involved. Grönska has focused upon the exchange prac-
tices related to selling product to the end consumer. Social media, packaging,
press releases, and demonstration products are all examples of activities it has
adopted to attract potential sellers and customers. Securing Paradiset as a
seller early on has been instrumental in establishing prices.
These activities by both firms are intertwined, and help shape the overall
market for urban produced food. What becomes apparent is that they are
initially focusing upon influencing two distinct markets: the real estate mar-
ket (that is to say, vertical farming structures) and the market for the end
product (for example, salad greens, herbs, and so on). Plantagon has largely
focused upon building a real estate market for urban farming. In doing so,
exchange practices related to product price is less the focus; instead, attention
has been diverted to building relationships with architects, the real estate
sector, and other actors in the city whose support will be instrumental: such
as city and national governments. Plantagon has developed renderings and
gathered data on building costs, which speak directly to its real estate and
architecture audience. Only after the company has been able to secure part-
nerships that will enable it to test and prove that urban food production is
economically feasible, has it turned to the seller and end-consumer market.
The initial focus upon brand and technology development also affected the
network Plantagon has built; it has focused upon an international network in
addition to its Swedish one.
Grönska has concentrated primarily on the customer market in Sweden:
both through building economic exchange mechanisms with sellers and par-
ticipating in activities that enable it to educate the end consumer about vertical
farming. Grönska participates in events where it can showcase its product and
295
have direct contact with the end consumer since getting product to market is
its intention. While the company also interacts and speaks with architects and
other real estate actors, it is not done as a focused strategy. This also affects
how the company positions itself, which is as a farm and as a local Swedish
producer of food. In this way, Grönska espouses the values of locally-produced
safe food.
Vertical farming does not fit squarely in the agriculture or real estate indus-
tries. Therefore, vertical farmers must educate and communicate not only
with the sellers and end users of their products; they must also reach the larger
institutions in society that contribute to the value creation and acceptance of
urban-produced food. This links back to the idea of institutionalisation
(Vargo, et al. 2015). Plantagon and Grönska are introducing a new concept of
farming; for true market innovation to occur, however, there will be an ongo-
ing need for the firms to build relationships (for example Grönska and Paradi-
set), processes (such as more efficient production and distribution methods)
and collaborative initiatives (as with Plantagon and Tekniska Verken).
Summary & Conclusion

The development of technology that enables food production in urban areas
has resulted in the emergence of firms that believe not only is food production
in cities feasible; it is also a necessity to feed a growing urban population. This
chapter’s premise was technological innovation must be complemented by
market innovation in order for digital transformation to occur and have an
impact. This research explored through two cases the current activities upon
which vertical farms are focused in order to build a market for urban-pro-
duced food: Plantagon and Grönska. A conceptual model of markets was
used to help outline the market shaping activities, as well as the network of
actors that are vital in institutionalising the idea of urban-produced food.
Vertical farming is still in its early stages of development in Sweden and yet,
it is already establishing the need for different industries to work together.
Plantagon and Grönska are building a network of actors to support their
efforts. Relationships in a variety of industries are needed to enable market
innovation: including real estate, government, clean technology, food distrib-
utors, food markets, and end consumers.
The cases revealed how many activities of the two farms are similar, yet
they differ on the scale of the endeavours and the timing at which the compa-
296
nies chose to focus upon them. The reason for this is the two firms are actu-
ally shaping two distinct markets: a real estate market and a seller/consumer
market. This has affected which actors and networks they have wanted to
influence as well as establish relationships. As part of an emerging market in
Sweden, however, both companies’ efforts appear to be beneficial in building
the larger ecosystem that will be ultimately needed for a sustainable market
of urban-produced food. For true market innovation to occur, new agricul-
tural standards and norms must be integrated into the current food produc-
tion system. In other words, vertical farming must become institutionalised.
The level of success in which vertical farms are able to co-create the value
with other institutions in the ecosystem will have a large effect upon what
does and does not eventually work. This also highlights how the values and
social forces of current markets have the ability to constrain or encourage the
integration of new technologies and form new markets.
FUTURE ECOSYSTEM CONSIDER ATIONS FOR URBAN-PRODUCED FOOD

As more technology innovation occurs, which kind of market innovation will
need to follow?
In order to scale the output of the farm, there needs to be a scaling of the
necessary expertise to run the farms. Managers, engineers, data scientists,
plant biologists, technology and software developers, and farm workers will
also be required. Laws may need to be modified to enable structures in urban
areas to be used for food production. Government and educational institu-
tions will need to support and invest in educational programs that train the
needed workforce to produce food in a digital environment. Demand and
supply mechanisms must also be balanced to ensure that goals are met, such
as the reduction of food waste and reducing transportation needs; it is not one
specific practice. The constant and iterative work of the vertical farms, with
a number of stakeholders across different industries, will lead to the digital
transformation of food production in cities.
297
References
Beecham Research. 2016. Enabling the Smart Agriculture Revolution: The Future of Farming
Through the IoT Perspective. London: Beecham Research Ltd. [Online] Available at:
http://www.beechamresearch.com/downloads.aspx [Accessed 20 June 2017].
Despommier, D. 2011. The Vertical Farm: Feeding the World in the 21st Century. (2nd ed.)
New York: Picador.
Association for Vertical Farming (n.d.). Glossary for Vertical Farming. [Online] Available
at: https://vertical-farming.net/vertical-farming/glossary-for-vertical-farming/ [Accessed
12 June 2017].
Kjellberg, H. and Helgesson, C-F. 2007. On the nature of markets and their practices.
Marketing Theory, 7(2): 137–162.
Storbacka, K. and Nenonen, S. 2011. Markets as configurations. European Journal of Marketing,
45(2): 241–258.
Vargo, S.L., Wieland, H., Akaka, M.A. 2015. Innovation through institutionalization:
A service ecosystems perspective. Industrial Marketing Management, 44: 63–72.
Ryan, K. 2017. The Future of Farming May Not Involve Dirt or Sun. Inc.: 14 Jun 2017.
[Online] Available at: https://www.inc.com/kevin-j-ryan/aerofarms-disruptive-25-2017.html
[Accessed on 27 July 2017].
298
Future Outlook
CHAP TER 15
Future Outlook on Digitalization

ROBIN TEIGL AND, CL AIRE INGR A M BOGUSZ, AND ANNA FELL ÄNDER
Introduction
Since the year 2000, digitalization’s influence on products, services, processes,
and business models is the primary reason that just over half of the names of
the companies on the Fortune 500 list have disappeared.1 Technology-based
companies are increasingly taking centre stage, thus, replacing traditional
asset heavy companies with their asset light operations. For example, in terms
of market capitalisation Microsoft was the only technology company among
the top five publicly traded firms in 2000: GE, Citibank, Walmart, Exxon
and Microsoft. By 2016, however, all five were technology companies: Apple,
Alphabet, Microsoft, Amazon, and Facebook.2 Turning to the S&P 500, the
percentage of tangible assets in these companies’ valuation from 1975 to 2015
fell from 83% of the total value to 13% with the value of intangible assets, ris-
ing from 17% to 87%.3
In addition to technology companies taking centre stage in terms of market
valuation, digitalization is also influencing value creation activities on differ-
ent levels. Digitalization is dissolving the boundaries of the firm on the more
basic process level, as some firms move the nexus of their value-creating activ-
ities, such as branding and innovation, to informal networks outside the firm.
Secondly, some firms are moving away from traditional value-chain, pipe-
line-based business models on the business model level, to multi-sided platform
business models that enable transactions among strangers in what has been
labelled the Platform Economy. We are seeing digitalization’s potential to trans-
form entire industries on an even higher level – that of industry. Furthermore,
1 https://www.weforum.org/agenda/2016/01/digital-disruption-has-only-just-begun/
2 http://www.visualcapitalist.com/chart-largest-companies-market-cap-15-years/
3 http://www.oceantomo.com/blog/2015/03-05-ocean-tomo-2015-intangible-asset-market-value/
301
ROBI N T EIGLA N D, CLA I R E I NGR A M BOGUSZ, A N D A N NA F ELLÄ N DER
there are indications “under the surface” that there are even more significant
forces at work: forces that challenge our basic assumptions of an industrialised
society and work, as we know it today; and that these forces may be taking us
rapidly from the third into the fourth industrial revolution, as discussed at the
World Economic Forum in January 2016.4
The purpose of this chapter is to provide a broad overview of digitalization
to date as well as provide some thoughts on how digitalization might influ-
ence the future of value creation in society. This research is intended for a
broad audience: from practitioners to policymakers, as well as those inter-
ested in learning about digitalization’s influence upon society. We focus first
upon crowdsourcing activities: one of the initial areas significantly influenced
by digitalization; we then turn to the Platform Economy, and the related
Sharing Economy: discussing how these new business models are challenging
our traditional strategy tenets. We then raise our level of discussion and turn
to industry evolution and digitalization’s influence upon the transformation
of industries, with a focus upon one industry in particular: the financial ser-
vices industry and the FinTech phenomenon, as well the recent trend of fus-
ing the physical with the digital. We then branch out to some emerging tech-
nologies that are currently receiving the most attention: IoT data analytics,
artificial intelligence (AI), blockchains, 3D printing, and virtual and augmen-
ted reality. We discuss how digitalization may impact the future of the labour
force. Lastly, we end the chapter with some recommendations for both man-
agers in general, as well as policymakers in Sweden.
Digitalization of Processes:
Dissolving Firm Boundaries through Crowdsourcing
One of the first hallmarks of the current digitalization era has been the emer-
gence of crowdsourcing that describes how the collective resources of a large
group of people can be used to help solve problems. While crowdsourcing has
existed throughout history, the internet along with digitalization has greatly
facilitated the means with which crowdsourcing can emerge: ranging from a
group of strangers self-organising on the internet (for example, open source
software) to user-generated content supported by hierarchical organisations
(such as a firm). Taking this into consideration, crowdsourcing can be defined
4 http://marketrealist.com/2016/01/fourth-industrial-revolution-need-know/
302
F U T U R E OU T LOOK ON DIGI TA LI ZAT ION
as “a collaboration model enabled by people-centric web technologies to solve

individual, organisational, and societal problems using a dynamically formed
crowd of interested people who respond to an open call for participation”
(Pederssen et al., 2013). As such, crowdsourcing has also led to the develop-
ment of the term Collaborative Economy.
User-generated content lies at the heart of organisations such as Wikipedia
and Facebook, in which users create content without any expectations for
reimbursement; six of the ten most popular content sites in the world are
primarily user-generated (Brynjolfsson & McAfee, 2014). A recent study of US
millennials (those born between 1977 and 1995) revealed that they spend 18
hours a day consuming media, and 30% of this time is with user-generated
content as compared to 33% on all traditional media combined (TV, print,
radio).5 Furthermore, the same study found that millennials trust UGC 40%
more and find it more memorable than traditional media.
As a result of these changes, businesses have realised over the past ten
years that they no longer can rely upon traditional media for their marketing
and branding efforts and are, thus, enabling customers to play a much more
active role in their business decisions: such as marketing and product devel-
opment. They have increasingly shifted from static websites to interactive
platforms using a variety of social media applications that enable user-gener-
ated content (UGC). Many are calling this a paradigm shift since these plat-
forms have led to a move from “one-to-many” (kinds of) communication
between the business and the customer to “many-to-many” interactive dia-
logues among the business and its customers. Burberry’s Art of the Trench
website is one of the first examples of this. Launched in 2009, it enabled users
to upload and comment upon pictures of people wearing Burberry products.
Within one year, Burberry’s ecommerce sales rose by 50%.6
In addition, numerous firms have developed user innovation communities
specifically to help in the development and testing of new services and products.
For example, companies such as Dell and Starbucks host electronic social envi-
ronments in which globally distributed customers can share their knowledge
and expertise with the community and the organisation through their com-
ments upon existing products and services, while proposing new innovations.7
5 http://corp.crowdtap.com/socialinfluence
6 http://blog.hubspot.com/marketing/examples-of-user-generated-content
7 http://blog.hubspot.com/marketing/examples-of-user-generated-content
303
Another well-known example is Threadless: a company founded in 2000 that

enables anyone to design t-shirts and a variety of other small products that are
then produced and sold through the Threadless platform if selected by the
crowd. Under this business model, customers now conduct many of the tradi-
tional in-house value-creating activities: such as idea generation, product design,
marketing, and sales. Due to their customer voting system that ensures all
t-shirts produced are sold, Threadless has significantly reduced its unsold inven-
tory compared to traditional clothing manufacturers.8 As a result, the compa-
ny’s profit margin was estimated to be 30%, which is remarkably high for a
company in a commodity industry.9
User innovation communities can lead to beneficial results, such as faster
innovation times; more successful new product launches, and increased cus-
tomer satisfaction. However, they also present considerable challenges to
managers. For example, a study of Dell IdeaStorm identified four key chal-
lenges in its first 18 months managing the community: understanding the
ideas posted, identifying the best ideas, balancing the needs of transparency
with the community against disclosure to competitors, and sustaining the
community (Di Gangi et al., 2010).
While this model of “flipping the firm” is receiving considerable attention
from researchers and managers alike, one of the most widely cited companies
is Quirky, which declared bankruptcy due to its inability to sustain profitabil-
ity. As a result, several people question whether the crowd is capable of select-
ing which products the market wants.
Digitalization of Business Models:

Strategy Tenets Challenged
Moving beyond crowdsourcing and user-generated content, the development
of multi-sided or matchmaking platforms that enable peer-to-peer transactions
has spread from industry to industry: that is to say, individual-to-individual
rather than firm-to-individual transactions. Transactions can be mediated
either by firms, such as Uber, that run their own platforms or by groups of
private individuals who use existing social networking sites, such as Face-
book; they can even be mediated by those who design their own apps using
open source software to self-organise their peer-to-peer activity in the cloud
8 http://www.fastcompany.com/1714561/company-community-threadless-puts-everyone-charge
9 http://www.inc.com/magazine/20080601/the-customer-is-the-company_pagen_2.html
304
with the help of blockchain-enabled smart contracts. These platforms may

locally connect acquaintances on a face-to-face basis or they may digitally
connect strangers from all over the world: for example, Airbnb and Upwork.
In particular, firms that have multi-sided and technology-based platforms at
the core of their respective business model have redefined the global business
landscape throughout recent decades. This has been increasingly the case
within the past ten years.10 Companies such as Amazon, Alibaba, Apple,
eBay, Facebook, Salesforce, Upwork, and Google have all emerged basing
their business models and operations upon platforms that provide the infra-
structure and rules for interactions between users, and they differ substan-
tially from traditional offerings in terms of their value chain.
One interesting observation when comparing a traditional company with
a platform company delivering similar products or services within an indus-
try reveals a remarkable reduction in the total number of employees per firm.
For example, Marriott, which was founded in 1927, had around 200,000
employees and a market capitalisation of USD 32 billion in 2017 while Airbnb,
which was founded in 2008 and now offers more rooms than Marriott, had
only 5,000 employees and a market capitalisation of USD 31 billion. A second
example is Walmart, which was founded in 1962: with approximately 2.3
million employees, it has a market capitalisation of USD 200 billion compared
to Alibaba, which was founded in 1999, and has around 36,000 employees
and a market capitalisation of USD 240 billion.
In the traditional strategy literature, value is accumulated from left to right
as value-adding activities occur in sequential processes; inputs are delivered
to a transformation process, which then become outputs. These activities are
generally conducted in-house by the firm’s employees with costs being
incurred throughout the value chain and revenues generated only on the
downstream side. Since the firm owns many of the resources, employs indi-
viduals with the necessary skills, and provides the management, organisa-
tion, and physical infrastructure necessary for the process, it incurs consider-
able fixed costs. A firm strives to achieve supply-side economies of scale in
order to then create competitive advantage, through increasing production
volume and sales to reduce the unit cost, thereby, creating entry barriers.
However, the firm reaches diminishing returns at some point, as acquiring
10 https://hbr.org/2006/10/strategies-for-two-sided-markets
305
new customers becomes more difficult and more expensive since fewer people
find the value proposition of the firm appealing (Eisenmann et al., 2016).
These basic assumptions in strategy are now being challenged with multi-
sided platforms (Eisenmann et al., 2016). Since the platform serves two groups
of users, not only are costs incurred on both the left and the right side; reve-
nues are also generated on the left and the right side of a value activity system.
Moreover, with platform-based businesses, the firm merely plays an interme-
diary role matching those that own resources with those looking for them.
Firms are relatively asset light and do not own the resources being transacted,
nor employ the individuals to perform the resource transaction process; they
merely provide the infrastructure and the rules of transaction for the two.
What is new is that successful platform-based businesses experience increases
returns to scale due to network effects. Network effects are based upon Met-
calfe’s law, which states: “The value of a telecommunications network is
proportional to the square of the number of connected users of the system”.11
Platforms strive to gain users to build critical mass such that it can more
efficiently match supply with demand and raise the likelihood that users on
both sides of the platform have their demands fully satisfied. As more people
use the platform, the value of the platform to each user rises. To create a
competitive advantage, firms strive to leverage network effects in order to
create demand-side economies of scale (Parker et al., 2016). In contrast to
traditional manufacturing and service firms, platform-based firms’ experi-
ence improved margins, as the number of users grow since users will pay
more for access to a bigger network (Eisenmann et al., 2016).
As the platform-based business model continues to penetrate industries, one
aspect that is becoming increasingly prevalent is the “winner takes all” effect.
More specifically, platform leaders are able to drive out weaker rivals and
create a barrier to entry due to their ability to leverage higher margins through
investing more in R&D and, thus, lowering their prices. This leads to mature
two-sided network industries usually being dominated by a handful of large
platforms with some extreme situations in which a single company emerges as
the winner, taking almost all of the market (Eisenmann et al., 2016).
These new market dynamics are not unique to B2C markets. Indeed, B2B
industries are also seeing a rise in the number of firms focusing upon building
11 https://en.wikipedia.org/wiki/Metcalfe%27s_law
306
platforms to support their traditional business models, as well as striving to

achieve the “winner takes all” effect. For example, with the goal of becoming a
top 10 software company by 2020, GE has invested more than USD 1 billion in
creating a market around the industrial internet through the development of its
Predix platform, based upon open source software and the blockchain.12 GE
entered a partnership with Apple in October 2017, and unveiled a new software
development kit (SDK) to enable developers to create their own industrial IoT
apps using predictive data and analytics based upon the Predix platform.13
In connection with the rise of platforms, the use of the notion “Sharing
Economy” has also become prevalent and is defined to include the renting,
bartering, loaning, gifting, and swapping of assets that are typically under-
utilised: either because they are lying unused or because they have not yet
been monetised (Felländer et al., 2015). Sharing Economy companies are
based upon two-sided networks with the premise that a consumer’s under-
used or spare fixed assets can be shared: a business model that is predicted to
grow from $26 billion globally in 2015 to $335 billion by 2025.14 Some Sharing
Economy companies have achieved substantial valuations in a very short
period due to the promise of network effects and increasing returns to scale,
as users pay more for access to a bigger network and margins improve as the
user bases grow: for example, AirBnB with $31 bln15. However, a debate is
ongoing as to whether this is truly sharing or merely a new form of capital-
ism: this phenomenon known as “crowd-based capitalism”, as some research-
ers, such as Professor Arun Sundararajan, have labelled it.
Digitalization and Industry Evolution

“Creative destruction” is a term coined by economist Joseph Schumpeter in
1942 (Schumpeter, 1942); it refers to the process whereby the creation of a new
industry or method of doing things destroys the industry or process that
preceded it. Creative destruction is obviously evident in some industries: for
example, the replacement of records with tapes, CDs, DVDs, and then by
12 http://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/how-b2b-digital-
leaders-drive-five-times-more-revenue-growth-than-their-peers
13 https://www.apple.com/newsroom/2017/10/apple-and-ge-partner-to-bring-predix-industrial-apps-to-
iphone-and-ipad/
14 http://www.pwc.co.uk/issues/megatrends/collisions/sharingeconomy/the-sharing-economy-sizing-the-
revenue-opportunity.jhtml
15 https://www.cnbc.com/2017/03/09/airbnb-closes-1-billion-round-31-billion-valuation-profitable.html
307
streaming. However, it is not clear how this process will play out across indus-
tries as digitalization progresses. Under the Sharing Economy umbrella,
although incumbents in the hotel and taxi industries have, indeed, been
forced to innovate in order to survive due to new entrants and platform busi-
ness models, neither the industries nor their processes have been “destroyed”
by the advent of sharing. Rather, traditional goods and services have been
augmented through an interconnected array of digital services, such as social
networks, location services, online payments, and rating systems. Such novel
elements bring greater benefit to consumers, which often leads to the destruc-
tion of old practices, yet this will not necessarily lead to the destruction of old
products and services.
Research investigating industry transformation by Anita McGahan (2000)
at the University of Toronto indicates that industries evolve along four trajec-
tories: radical, progressive, creative, and intermediating. These trajectories
are determined by the degree to which the industry’s underlying core activi-
ties and core assets are threatened, as they become less relevant in the market-
place due to new and alternative solutions. Core activities are the recurring
value-creating activities that attract and retain suppliers and buyers in the
industry, while core assets are the durable tangible and intangible resources
that enable these efficient core activities.
This research also suggests that the traditional model of industry life cycle
is only relevant for industries experiencing progressive or creative change and
not for radical or intermediating change. An alternate model was suggested
instead, in which an emerging industry that start-ups developed offers alter-
native value-creation solutions and grows in volume to supersede the sales’
volume by industry incumbents with traditional activities. Competitive
advantage in the industry then becomes increasingly based upon the ability
to provide products and services that are evaluated based upon the new crite-
ria as new industry standards are created (McGahan, 2000).
A study of the transformation of US industries from 1980 to 1999 by McGa-
han revealed that the 43% of industries were characterised by progressive
change, 32% by intermediating change, 19% by radical change, and 6% by
creative change. As digitalization pervades industry after industry, it will be
worth revisiting these figures to see how industries have transformed on a
global scale.
308
FINTECH AND THE TR ANSFOR M ATION

OF THE FINANCIAL SERVICES INDUSTRY?
One industry that is rapidly transforming is the financial services industry:
due to the emergence of the FinTech sector enabled by the mobile internet,
cloud computing, the smart phone, changing consumer behaviour, and a
mistrust in the established banks along with declining internet technology
start-up costs. While there are many definitions of FinTech companies, one of
the most frequently used is “those that offer technologies for banking and
corporate finance, capital markets, financial data analytics, payments, and
personal financial management” (Skan et al., 2014). The FinTech sector
attracted investments of USD 36 billion globally in 2016.16
A look into the transformation of this industry reveals that FinTech start-
ups are specialising in individual services and products traditionally provided
by the larger banks as opposed to competing with established banks in multi-
ple products. As the entrances of these FinTech companies press margins and
take market share from industry incumbents, a McKinsey study predicted in
September 2015 that the profitability of the established private banks may fall
up to 60%.
The book, The Rise and Development of FinTech: Accounts of Disruption from
Sweden and Beyond, published in 2018 by Routledge Publishers, offers an
in-depth discussion of the transformation of this industry. One area discussed
is crowdfunding, which falls under the label of the sharing economy previ-
ously discussed in this chapter. Crowdfunding, or the accumulation of small
investments in individual projects from a large number of individuals (the
“crowd”) via or with the help of the Internet and social networks (De Buysere
et al., 2012), is enabling new forms of investment for private individuals, entre-
preneurs, SMEs, and even large organisations. There are four established
categories of crowdfunding: donation-based or philanthropic, reward-based,
equity-based, and lending or debt-based (often referred to as P2P lending),
with some claiming that the particular form of real-estate crowdfunding
should be recognised as a category. P2P lending currently dominates the
other crowdfunding forms, in terms of money transacted, and while it was
predicted to account for approximately USD 25 billion of the total USD 34
16 https://letstalkpayments.com/global-fintech-funding-36-bn-2016/
309
billion globally raised through crowdfunding in 2015, reports claim that P2P
lending topped USD 150 billion in 2015 in China alone.17
One of the greatest challenges in taking an idea to the marketplace is the
ability to raise funds. Several researchers are now suggesting that crowdfund-
ing is democratising the process of innovation through providing access to
necessary capital to those individuals outside of traditional investor networks,
such as through gender discrimination, or located in remote areas. For exam-
ple, the company Laser Unicorns raised more than USD 600,000 from almost
18,000 individuals across the globe for the production of its movie, Kung Fury,
despite being located in Umeå, a town in the north of Sweden.
Other areas developing rapidly within FinTech include the robotisation of
personal wealth management as well as the penetration of cryptocurrencies
and blockchain technologies for new payment and reconciliation solutions.
The 2008 white paper “Bitcoin: A Peer-to-Peer Electronic Cash System” was
published on the internet under the pseudonym of Satoshi Nakamoto.18 A few
months later, in early 2009, the project was launched on an open-source proj-
ect repository. These initial developments have spawned more than 900
cryptocurrencies today although only around a handful have a market capi-
talisation of more than USD 10 mln.19 A cryptocurrency is a digital, decentra-
lised, peer-to-peer currency that uses cryptography to validate and secure
transactions. The largest and most well known cryptocurrency is Bitcoin:
with a market cap of almost USD 100 billion during October 2017. Since Bit-
coin does not have a central clearing house, there is no central authority in
charge of the money supply, nor are there any financial institutions involved
in the transactions, Bitcoin differs from traditional fiat currencies: that is to
say, nation-state currencies, such as the USD and SEK. The members of the
Bitcoin network perform these tasks themselves: by verifying and validating
every transaction that occurs between network members in order to avoid the
risk of double spending. Today there are more daily transactions with Bitcoin
than there are through Paypal.
At the core of Bitcoin is the blockchain protocol, which in essence is a
shared public ledger of all verified transactions. During 2015, a number of
large multinationals, such as Goldman Sachs, JP Morgan, IBM, and Samsung,
17 http://www.crowdfundinsider.com/2016/01/79612-report-china-p2p-lending-topped-150-billion-in-2015/
18 https://bitcoin.org/bitcoin.pdf
19 http://www.cryptocoincharts.info/coins/info
310
began to pay increasing attention to the bitcoin blockchain and other emerging
blockchain protocols, such as ethereum, hyperledger, and R3Corda. This was
due to the potential of a number of uses for the blockchain and their ability to
incorporate IoT, smart contracts, and machine-to-machine micropayments
within financial services and supply chains. These are just some of the exam-
ples of the many FinTech companies founded in recent years; we will be able
to see in time whether these start-ups are capable of gaining enough momen-
tum to transform the industry.
STR ATEGIES FUSING THE DIGITAL WITH THE PHYSICAL

Another area of industry transformation relates to an increasing number of
companies implementing strategies that fuse the digital with the physical. In
order to create a sustainable competitive advantage with a digital strategy, an
increasing body of research is arguing that companies that are able to fuse
their physical and digital operations, such that customers can easily move
between the two, will be more successful at creating a sustainable competitive
advantage.20 While many argue that online operations will displace physical
operations in numerous industries, research is finding that the two worlds are
complementary, and the combination of the two is the real transformation
that is occurring. For example, as part of Burberry’s fusion strategy to attract
millennials, the company developed digitally immersive physical experiences,
such as in-store digital screens that turned into catwalks: a strategy that tri-
pled the firm’s stock from 2006 to 2014. 21
Not only are physical firms fusing with digital; digital firms, such as the
two early pioneers of online trading: E*Trade and TD Ameritrade, are also
integrating physical offerings with their digital ones. More recently, Amazon
has made major efforts into the physical retail space in the USA: including the
purchase of Whole Foods Grocery Chain for USD 13.7 billion, a partnership
with Kohls – the retail department store chain to provide return locations, the
opening of bookstores, the establishment of lockers and pick-up spots in retail
stores and on college campuses, and a Treasure Truck in six cities that sell
certain items at a discount.22 Amazon’s efforts have revealed that having a
20 http://www.gartner.com/imagesrv/books/digital-edge/TheDigitalEdge.pdf, http://www.bain.com/
publications/articles/leading-a-digical-transformation.aspx
21 https://hbr.org/2014/09/digital-physical-mashups
22 https://www.cnbc.com/2017/09/19/amazon-is-firing-on-all-cylinders-to-grow-its-retail-presence.html
311
physical presence fuels online sales. For example, Amazon sold about
$500,000 worth of Whole Foods-branded products online in the first week
after making them available following the purchase of the chain.23
Other examples include Local Motors, which focuses upon motorised vehi-
cles, and First Build, focusing on household appliances. Both of these compa-
nies have created micro-factories or physical workspaces that enable the
“crowd” to participate in the development and production of products, thus,
bridging the divide between virtual community and physical participation.
Investigating the level of transformation within 20 industries, the afore-
mentioned research found that industries, such as the airlines, automobile and
insurance industries, will be those that have the greatest level of innovation
in fusing the physical with the digital within the next five years.24 Develop-
ments in areas such as the Internet of Things, 3D printing, and virtual and
augmented reality will likely hasten this transformation in many industries.
Emerging Technologies
Some of the greatest areas of uncertainty about the future relate to the devel-
opment of emerging technologies: such as the IoT data analytics, artificial
intelligence (AI), blockchains, 3D printing, and virtual and augmented real-
ity. We provide a brief overview of each of these below.
IOT DATA ANALY TICS

Both technology evangelists and consumer goods manufacturers have hailed
the rise of the Internet of Things (IoT), wherein everyday devices and prod-
ucts contain tiny computers and will communicate directly with machines
and with one another. Things can be programmed not only to alert the user
to certain events; they will also communicate with each other and even to
coordinate and make micro-payments directly among themselves without
human involvement. The name “Internet of Things” is perhaps somewhat
inaccurate. A more accurate label might be “The Internet of Small Computers
on Things”, which is another way of saying that the things themselves cannot
be endowed with the capacity to connect to the Internet. Rather, manufactur-
ers have started to install small computer processers on devices and these
computers both connect to the internet and control the device. This idea is
23 https://www.cnbc.com/2017/09/19/amazon-is-firing-on-all-cylinders-to-grow-its-retail-presence.html
24 https://hbr.org/2014/09/digital-physical-mashups
312
not new; however, advances in both computer hardware and internet connec-
tivity have occurred over the last decade, including increasingly smaller
computer parts and an increasing number of processes being hosted in the
cloud. Thus, the Internet of Things has become a realistic possibility.
To date, health and fitness has seen the most activity with established
brands; Nike, Adidas, and Apple have taken advantage of the emerging tech-
nologies in wearables. However, this is one area that is expected to further
spawn growth and transformation across all industries. From 2003 to 2012,
the number of connected devices grew from 500 million to 12.5 billion; this is
expected to grow to 50 billion by 2020. Furthermore, the machine-to-machine
(M2M) market, in which machines communicate and perform functions with
no human intervention, is expected to reach €40 billion by the end of 2017.
One driver of this rapid growth is the falling costs of sensors; a sensor that
cost €50 in 2009 fell to €15 in 2013. Another is the development of cloud
computing that enables data analytics. As businesses and their users are pro-
ducing an increasing amount of data, some cloud providers are providing the
hardware and algorithms to mine these data for in-house operational and
customer insights, such as the ability to predict and influence customers
online and offline (McAfee, 2012). The need for computing power required
for big data analytics continues to rapidly grow as the amount of data created
each day doubles every 40 months or so (McAfee, 2012). However, the cloud
enables any business, regardless of the size or financial resources, to achieve
the IT productivity of the largest enterprises with extensive IT budgets.
Thus, there are opportunities in every sector, as IoT will be employed to
lower costs and improve productivity and safety.25
ARTIFICIAL INTELLIGENCE
Artificial Intelligence (AI) refers to a “broad set of methods, algorithms and
technologies that make software ‘smart’ in a way that may seem human-like to
an outside observer”.26 Artificial intelligence is often confused or used simul-
taneously with other terms: such as machine learning, deep learning, neural
networks, and cognitive computing. One way of describing the differences
25 https://www.bcgperspectives.com/Images/The%20Mobile%20Internet%20Economy%20in%20
Europe%20Dec%202014_tcm80-178364.pdf
26 https://www.computerworld.com/article/3040563/enterprise-applications/5-things-you-need-to-know-
about-ai-cognitive-neural-and-deep-oh-my.html
313
among these is that they are similar to Russian Dolls in which deep learning
is a subset of neural networks, which is a subset of machine learning, which is
a subset of artificial intelligence.27 Arthur Samuel at IBM first defined machine
learning in 1959 as a “field of study that gives computers the ability to learn
without being explicitly programmed”.28 Thus, machine learning distinguishes
itself from more general AI as it has the ability to learn or modify itself without
human intervention when it is exposed to more data. Neural networks are a
one kind of machine learning technique that is based upon the way in which
neurons work in the brain. Deep learning, or deep neural networks, is a rela-
tively new field within machine learning that has become popular of late due
to the availability of large amounts of data combined with major advances in
processing power. Deep learning is based upon a “family of algorithms that
implement deep networks with unsupervised learning”29; it includes a large
system of neurons arranged in several hidden layers. Finally, the term cogni-
tive computing is a bit controversial since it is unclear as to whether it is a true
category of AI or just a buzzword; however, the primary idea is that the appli-
cation of AI focuses upon “reasoning and understanding at a higher level,
often in a manner that is analogous to human cognition -- or at least inspired
by human cognition,” with the purpose of making high-level decisions in com-
plex situations.30
AI is not new, yet the technology has been rapidly maturing due to greater
access to more data through maturing technologies, such as IoT and
improved access to processing power: for example, through the cloud. The
vast amount of data in decentralised networks is making the algorithms
smarter. Thus, AI using large datasets can create enormous gains for firms:
such as customer loyalty and trust, lower costs, improved quality, and
increased agility. Personalised offers and recommendations can be scaled and
large efficiencies can be created across the value chain. We are entering a new
era where every organisation must be able to handle data to create customer
and operational value.
27 https://deeplearning4j.org/ai-machinelearning-deeplearning
28 https://www.ibm.com/developerworks/community/blogs/jfp/entry/What_Is_Machine_Learning?lang-
=en
29 https://www.ibm.com/developerworks/library/cc-beginner-guide-machine-learning-ai-cognitive/index.
html
30 https://www.computerworld.com/article/3040563/enterprise-applications/5-things-you-need-to-know-
about-ai-cognitive-neural-and-deep-oh-my.html
314
However, come critical issues that must be addressed come with the oppor-
tunities of using algorithms on data. Without the proper oversight, data and
AI usage can have unintended consequences: such as machine bias and dis-
crimination, misuse of data, and sensitive data breaches. For example, ethical
considerations lie in the bias of the programmer when programming how an
autonomous driving vehicle should act in a crash situation. Furthermore,
many now argue that data are the new gold, and there is an increasing com-
petitive divide between leaders and laggards using AI on large data sets, thus,
leading to monopoly situations occurring faster than ever before.
A study conducted by MIT31 in partnership with BCG provides insights
into AI maturity levels among organisations and what they need to do in
order to develop an AI strategy. The results reveal low risk awareness and
knowledge gaps in organisations. Many believe in the future of AI, yet few
have a concrete strategy for how to get there or to govern it. Executives under-
estimate the security, and the organisational and technological capability
gaps. The study also shows that very few organisations have strong data
governance practices.
BLOCKCHAINS AND SM ART CONTR ACTS

One of the most promising areas of emerging technologies is in the use of
blockchain-based technologies to automate processes. The best known use of
the blockchain is probably as the underlying infrastructure that supports the
cryptocurrency Bitcoin. However, the technology has also been widely used
to build other automated networks. Another well-known network is the
Ethereum ecosystem, which provides a platform for the creation of new
automated systems, and a marketplace for the release of Decentralised Apps
or DApps. These DApps require the network currency Ether to operate.
Blockchain technology generally works by transmitting a transaction to a
network of computers that verifies when the transaction fulfils all the require-
ments for transaction execution: for instance, that the initiator of the transac-
tion owns the asset being transferred, or that the sender has fulfilled the
requirements for the assets to be released from escrow. The computers do the
verification; then they receive tokens in exchange for verifying the transaction,
31 MIT Sloan Management Review, September 2017. The study found that more than 3,000 executives,
managers, and analysts across industries. Complemented with 30 in-depth interviews with technology
experts and executives.
315
encrypting it, and releasing into a network-maintained ledger. This process is

known as mining; and in most blockchain systems, the tokens have come to
have monetary value as currencies in their own right, as has happened with
Bitcoin. Once the transaction has been transmitted and verified, the miners
receive their reward and transmit the individual transaction to the network; it
is typically not possible to reverse or alter the transaction. This is because no
single computer in the network controls the ledger of transactions; changes can
only be made when more than 50%32 of those in the network agree to the
change.
Blockchain automation holds a number of potential applications, beyond
just the creation of new currencies. These include self-executing smart con-
tracts, and organisations with automated vote verification and collection. To
automate land registration,33 Swedish Läntmateriet (The Swedish National
Land Survey) has already experimented with smart contracts, and Axel John-
son to automate supply chains.34 Blockchains are also being experimented
with for use in standardised international transaction systems,35 and for
online trading of securities and other commodities.36
As a way of obtaining entrepreneurial finance, creators of new block-
chain-based firms are even using smart contracts and existing cryptocurren-
cies. ICOs or initial coin offerings do so when these new firms sell tokens in
their new system in exchange for an existing cryptocurrency: typically, Ether.
The largest ICO to date, Filecoin, raised $205.8 million for tokens that allow
individuals to share unused space on their computers—and sell it for Filecoin
tokens.37
3D PRINTING
3D printing or additional manufacturing can be disruptive as it will transform
industries through shifting competitive advantage and changing organisa-
tional structures, as it accelerates product-development cycles, thus, enabling
32 Some researchers argue that the threshold is really 30%.

33 https://www.lantmateriet.se/contentassets/6874bc3048ab42d6955e0f5dd9a84dcf/blockkedjan-
framtidens-huskop.pdf
34 https://digital.di.se/artikel/axel-johnson-gruppen-vill-ta-blockkedjan-till-fruktdisken
35 Both for the transfer of funds and for smart contracts, see http://www.r3cev.com/blog/2016/4/4/
introducing-r3-corda-a-distributed-ledger-designed-for-financial-services
36 https://www.coindesk.com/coinbase-integrates-gdax-exchange-with-algorithmic-trading-platform/
37 https://www.coindesk.com/257-million-filecoin-breaks-time-record-ico-funding/
316
new manufacturing strategies, shifts profits, and requires new capabilities.38

3D printing technology has existed since the 1980s, yet the range of consumer
uses for 3D printing today is still quite limited to small plastic and metal
objects such as toys, jewellery, and small household and electronics items. For
example, Mattel has plans to sell 3D printer kits including the printer and app
with design software for children for approximately USD 300.39 However, 3D
printing within industrial and other production areas is already printing parts
in a variety of materials such as glass, titanium, and human cartilage; rapid
prototyping and cheaper product molds are currently commonplace, thereby,
removing economies of scale in many industries. As the technology continues
to mature, companies now explore how 3D printing can transform product
design. Not only can objects include moving parts; they can also be stronger
or lighter than their traditional counterparts due to novel designs often based
upon nature. For example, GE’s 3D printed fuel injection nozzles for airplane
engines are 25% lighter, saving approximately USD 1.4 million for each LEAP
plane; its new single-prop Cessna plane engine is 85% 3D printed, thus, reduc-
ing the number of parts from 855 to 12.
The improvements above are incremental, however, compared to how 3D
printing may radically transform manufacturing since it enables an idea to be
instantaneously transformed into physical reality anywhere in the world.
Today you can scan your foot with Volumental’s technology in New Balance
stores and receive your running shoes with perfectly fitting midsoles a few
hours later. Local Motors prints cars in its USA micro-factories instead of
manufacturing them in low-cost countries and shipping them over the ocean.
Ivaldi will digitally send its designs to ports across the world and 3D print
spare parts for the maritime industry. Amazon plans to 3D print objects in
delivery trucks while on the way to your door.
According to a recent the International Data Corporation, global spending
on 3D printing is predicted to grow from nearly $11 billion in 2015 to $26.7
billion in 2019 at a 27% compound annual growth rate (CAGR).40 The McK-
insey Global Institute has also estimated that 3D printing could have an
annual economic impact of up to USD 550 billion by 2025. As 3D printing
advances, traditional sources of competitive advantage, such as economies of
38 http://www.mckinsey.com/business-functions/operations/our-insights/3-d-printing-takes-shape
39 http://www.thingimaker.com/
40 http://www.idc.com/getdoc.jsp?containerId=prUS40960716
317
scale and low cost of labour, will decline in importance. As a result, manufac-
turing may be moved back to rich industrialised countries and increasingly
performed by SMEs and entrepreneurs41, encouraging disruption not only in
the manufacturing industries; it will also be evident in logistics and transpor-
tation industries. For example, Local Motors has plans to manufacture its 3D
printed cars in micro-factories in the US, cars that will be adapted to local
driving conditions, energy sources, and regulations.
Additionally, manufacturing may move to local or online communities of
individuals designing, printing, and selling a wide variety of objects. Sites
such as Thingiverse, Blender, and Shapeways currently facilitate such activi-
ties while 3D Hubs platform claims that its more than 27,000 printers in 150
countries across the globe provide one billion people with access to a 3D
printer within 10 miles of their home.
VIRTUAL AND AUGMENTED RE ALIT Y

2017 saw a revival in virtual and augmented reality with an increasing num-
ber of solutions coming onto the market and money being invested in these
technologies. Within the past few years, Facebook bought Oculus Rift for
USD 2 billion and Microsoft purchased Minecraft for USD 2.5 billion, and
Magic Leap has raised USD 1.4 billion, with USD 800 million in a series C
round: the largest series C round for an internet company at the time to date.
With such money being invested in these technologies, it might come as no
surprise that the market for VR/AR is expected to explode within the next
few years to more than USD 600 billion in 2025 with some predicting that
VR/AR technologies will replace the smartphone.
Although it is developing, virtual reality already enables various forms of
immersive training as well as the ability for multi-national employees spread
across the globe to simultaneously collaborate on project work or for children
from distant schools to build Minecraft models that they then print out on
their 3D printers. For example, Nvidia has plans to develop its software in
order to enable immersive collaboration for manufacturing design. Many
argue that augmented reality may have more potential than VR since it has
extensive uses in e-commerce and mobile commerce, as well as for use in
41 http://www.economist.com/node/21552901
318
industrial areas: such as manufacturing, equipment operations, and material

and inventory handling.42
Influence of Digitalization on the Labour Force

With its emerging technologies and new business models, digitalization has
already begun to influence the labour force. For example, the number of
freelancers in the world is rapidly increasing. One recent study in 2016
reported that the number of freelancers in the US had reached 35% of the
workforce and that this number continues to rise.43 Advances in artificial
intelligence will more than likely enable temporary freelance platform-based
organisations in ways we have difficulty imagining today. For example, a
research group at Stanford University has developed Flash Organisations: a
concept for AI-enabled temporary organisations. The group is currently
working on its Foundry software with the purpose of enabling crowd work-
forces to flexibly assemble and reassemble themselves into collectives “that
rival modern organisations in their prevalence, impact, and achievements”44.
With the help of AI, the software builds a temporary organisation by recruit-
ing freelancers from an online labour market – such as Upwork. Then, as
project demands change over time, it continuously updates the organisation
with new freelancers.
We now discuss two related areas: the immersive internet and automation,
in addition to taking a look at the implications of a freelance workforce.
IM MERSIVE INTERNET
Virtual and augmented reality solutions along with virtual world solutions,
such as OpenSimulator: the open source virtual world platform software, are
encouraging the development of the “Immersive Internet”: internet’s next gen-
eration that enables individuals to immerse themselves into an internet with 3D
sight and sound, and an increasing ability to transmit the sense of touch,
thereby, offering an immersive working and networking space. Regardless of
physical location, entrepreneurs may collaborate on value-creating activities in
these environments, with one another as well as with other individuals from
42 http://ftalphaville.ft.com/2015/12/01/2146247/welcome-to-your-simulacrum-future-a-674bn-opportunity/
43 http://www.marketwired.com/press-release/new-study-finds-freelance-economy-grew-55-million-
americans-this-year-35-total-us-workforce-2164446.htm
44 http://hci.stanford.edu/publications/2017/flashorgs/flash-orgs-chi-2017.pdf
319
both large and small firms, academia, hobbyists, and the public sector. This has
already been clearly demonstrated by the OpenSimulator project: an open
source virtual world platform project, in which the expression “meeting face-to-
face” has lost its physical meaning, indicating instead a virtual meeting of ava-
tars in a 3D online space.
Anyone with access to the internet will be able to learn just about anything
and engage in a world of economic opportunities through the next generation
of the internet; therefore, the number of freelancers across the globe will only
continue to rise. Moreover, individuals who have previously been hindered
from entering the workforce due to physical disabilities or peripheral loca-
tions will be able to learn and work through this immersive environment.
One significant question this raises is how this “mobility” of labour and the
“mobility” of physical goods due to 3D printing will impact the competitive-
ness of regions and nations.
We are finding indications of “open entrepreneurship” through the immer-
sive internet, or the process of entrepreneurs openly engaging in social capi-
tal-building activities through the free contribution to the public of intellectual
property and other resources, with the purpose of pursuing individual busi-
ness-related interests while contributing to the pursuit of collective goals.
While these entrepreneurs may give away for free their intellectual property,
knowledge, time, and other resources, they do so in the pursuit of creating a
social structure, which enables them to overcome the inherent difficulties in
attracting the necessary human, financial, and other resources due to the
uncertainties of their new venture and the liabilities of newness and small
size. Through the immersive internet, entrepreneurs may more easily over-
come these liabilities: factors that have traditionally disadvantaged small
firms compared with large organisations, which have generally led to their
failure. Furthermore, we are beginning to see signs that the immersive inter-
net is leading to, or implicated in, a migration from an economic model char-
acterised by centralised hierarchical firms controlling in-house resources to a
model of decentralised social production by communities of globally distrib-
uted firms and workers. Such fundamental changes clearly bring into ques-
tion how and to what degree well-established multi-national organisations and
their brands will continue to dominate economic activity.
As 3D printing becomes integrated with the immersive internet, as well as
becomes commonplace for household and industrial objects, we will not only
320
be able to design and experience objects in these virtual spaces; we will also be
able to simultaneously produce these objects in our garage or workplace, thus,
enabling an era of social manufacturing. Not only will the borders between
industries be blurred; the entire value creation system may be revolutionised
due to the convergence of the immersive internet with material, production,
and other related technologies: from the sourcing of inputs and production in
the supply chain through to distribution and end-user consumption.
AUTOM ATION
In a 2014 study by Carl Benedikt Frey & Michael Osborne from Oxford (2014)
on the future of employment, the authors predicted that 47% of total US
employment is at risk due to computerisation. Today, robots already cut our
lawns, vacuum our houses, write our sports articles, and make or lose our
money in the stock market, while drones deliver small packages in some
countries as well. Japan, the country that wants to be the world’s leader in
robots, now has two hotel – Henn-na Hotels, owned by the low-cost travel
agency H.I.S. Co. in Nagasaki and in Tokyo – that are run almost entirely by
robots. Robots help you check in, carry your luggage, provide concierge ser-
vices, haul your trash away, and even provide entertainment as fish in the
lobby’s tank.45 The only human tasks left are changing the sheets and taking
care of said robots.
However, there are a number of tasks that computers will have difficulty in
performing, and Carl Benedikt Frey & Michael Osborne developed in their
2014 Oxford study a list of nine task dimensions with computerisation bottle-
necks, as they are primarily non-routine tasks. These tasks fall under three
categories: perception and manipulation (finger and manual dexterity,
cramped workspace, and awkward positions), creative intelligence (originality
and fine arts), and social intelligence (social perceptiveness, negotiation, per-
suasion, and assisting and caring for others).
As a result, not all jobs will disappear. Indeed, it may be more beneficial to
consider which tasks, as opposed to jobs, will AI and/or robotisation control.
We can expect that some non-routine jobs will be augmented by computers
performing routine tasks while new jobs will appear, especially within the
high-paying sector. For example, five high-paying jobs that did not exist ten
45 http://www.japantimes.co.jp/news/2016/11/18/business/nagasakis-robot-staffed-henn-na-hotel-gets-
guinness-nod/#.WDWBV7VwpME
321
years ago include data scientist, mobile applications developer, information

security analyst, digital strategist, and green building and retrofit architect.46
However, we will see a middle-tier job squeeze as software applications and
robots continue to replace routine tasks. This will force middle-skilled labour
to compete for jobs in the lesser-skilled sector, which has the effect of reducing
wages (Felländer et al., 2015). Furthermore, the falling cost of automation and
robotisation will put downward pressure upon wages as the cost of substitu-
tion for traditional labour continues to fall. Such structural factors will hold
back wage increases and employment in routine jobs: even in the current
environment with a cyclical upswing in the US labour market.
Thus, as in previous technological paradigm shifts, it is predicted that there
will be technological unemployment as technological advances outpace the
rate at which we can find new uses for the displaced labour. Although this
prediction implies certain negative effects in the short run, the long-run effect
should be positive because history reveals that people tend to move higher up
in the value chain over time as lost jobs are replaced by new ones. With the
right training and circumstances, we can expect that self-employed individu-
als or freelancers can replace, to a certain extent, jobs that are lost due to
digitalization and robotisation. The question is with which pace firms, entre-
preneurs, and freelancers will be able to create opportunities and jobs in both
highly-skilled and lesser-skilled sectors.
A FREEL ANCE WORKFORCE?

Another area that deserves consideration is to what degree society is prepared
for a labour force that is composed of an increasing percentage of freelancers
and entrepreneurs. Labour regulations to date have focused upon protecting
the rights of workers vis-à-vis their full-time employers; however, such regu-
lations about the rights of freelancers and self-employed entrepreneurs are
often silent. Today’s digital platforms do not guarantee an individual’s
well-being in the same way that a “traditional” employer does, and individu-
als who provide their services through online platform transactions generally
bear a considerable amount of risk. For example, platforms such as Uber and
Airbnb, do not define themselves as employers in the transportation and hotel
sectors, respectively. Rather, they argue they are merely digital platforms that
46 http://www.payscale.com/career-news/2015/09/5-high-paying-jobs-that-didnt-exist-10-years-ago
322
match drivers with clients, and property owners with tourists. Hence, these
companies claim they are not responsible for the social benefits and insurance
of the drivers and property owners.
Furthermore, individuals often have no control over when and from whom
they receive work assignments. In addition, as the number of global freelance
labour platforms increases, freelancers may be affiliated with more than one
platform, thus, complicating the situation even more. For example, an individ-
ual might work full or part-time in Stockholm for Uber while simultaneously
developing software for an organisation in China or Australia through the
Upwork platform.
In most countries, because freelancers such as Uber drivers and other local
sourcing platforms are not considered to be legal employees, they are not able
to organise to obtain the collective bargaining privileges and protections that
those belonging to most labour unions have. Thus, although some people
argue that the Platform Economy offers flexibility and supplemental income
not available from traditional jobs, others argue that it signals a return to the
piecemeal labour system that exploited workers.
Moving forward, traditional unions will not capture the entire labour
force. Thus, labour market regulations must be adapted: not only to ensure
the traditional safety net for individuals, and to provide regulatory and tax
incentives to incentivise self-employment as well. One such organisation, the
Freelancers Union in the US, has emerged to serve the needs of this labour
force. However, as aforementioned, the line between employee and freelancer
is not always clear, and the Platform Economy is increasingly generating
more ambiguous situations. For example, what happens if individuals barter
or are paid in kind, rather than in money? Can residents of a community
volunteer for a housing association in exchange for reduced rent? If so, what
do they need to declare for tax purposes, given that the association may not
employ them? For that matter, does the association employ them?
Recommendations for the Future

What will the world look like in in the future? On the one-hand, it is most
likely that the penetration of the mobile Internet along with an internet-savvy
workforce will continue. Global mobile internet penetration is predicted to
reach 71% by 2019 while usage per device is to triple and will most likely
323
exceed the use of desktop and mobile browsers combined.47 By 2020, on the
other hand, the world will have its first generation of individuals who will
have grown up in an entirely digital world, and 50% of the global workforce
will be millennials.48
When it comes to firms, a study by McKinsey & Company suggests that
the average rate of digital penetration across all industries is only approxi-
mately 37% and, as penetration increases, industries will continue to see
downward pressure upon revenue and profit growth.49 Increased price com-
petition will result as the number of digital platforms grows and other tech-
nological advances decrease entry barriers and contribute to transparency.
New entrants, so-called born globals and micro-multinationals, are accessing the
global market right from their inception, which means that previously pro-
tected local providers will find themselves facing global competition
(Felländer et al., 2015). Traditional middlemen, which generated higher trans-
action costs and higher fixed costs of holding capital or labour, will be elimi-
nated; the more perfect matching of supply and demand through digital
platforms will further decrease transaction costs. Thus, although demand
may grow in many economies, companies will find it increasingly difficult to
increase their margins.
As a result, the pace of change is predicted to only increase. Research has
shown that the average lifespan of a company on the S&P 500 index has
declined from 61 years in 1958 to 25 years in 1980 to 18 years in 2012, and is
forecast to further shrink to 14 years by 2026.50 At the current churn rate, 75%
of the S&P 500 firms are predicted to be replaced by 2027.51 This “creative
destruction”, wherein some jobs, firms, and industries are transformed or
destroyed only to make way for others, has been depicted by some as thrilling
and promising, while others are more concerned. There is a lot that remains
to be seen; however, as we move into this era, which may be the beginning of
the fourth industrial revolution, characterised by the mass adoption of expo-
47 https://www.internetsociety.org//globalinternetreport/2015/assets/download/IS_web.pdf.
48 https://www.pwc.com/m1/en/services/consulting/documents/millennials-at-work.pdf
49 http://www.mckinsey.com/business-functions/digital-mckinsey/our-insights/the-case-for-digital-rein-
vention#0
50 https://www.innosight.com/insight/creative-destruction-whips-through-corporate-america-an-inno-
sight-executive-briefing-on-corporate-strategy/
51 https://www.innosight.com/insight/creative-destruction-whips-through-corporate-america-an-inno-
sight-executive-briefing-on-corporate-strategy/
324
nential technologies,52 many argue that even more significant disruptive

changes in economy, business, society, and individually are yet to come53.
We now present some more general recommendations for managers on
how to tackle the challenges of digitalization. In addition, we discuss some
policy considerations for Sweden and its policymakers on measures to ensure
that Sweden can benefit from digitalization in the future.
RECOM MENDATIONS FOR M ANAGERS

One study by the MIT Sloan Center for Information Systems Research found
that the winners in this new era will be those companies that fully embrace
digitalization. Those companies that had 50% or more of their revenues gen-
erated within digital ecosystems and understood their end customer better
than the average competitor exhibited 32% higher revenue growth and a 27%
higher profit margins within their industry than the average (Weill &
Woerner, 2015).
Digitalization is expected to even increase its pace; thus, it is impossible to
predict what industries will look like in ten to twenty years time as the major-
ity of what will happen in the future is beyond our comprehension – we do
not even know what we do not know about the future. One of the reasons that
companies are unable to survive in the long run is that they get blindsided by
changes in their industry: that is to say, they were unable to scan the periph-
ery of their industry for weak signals that later led to significant changes in
their industry, thus, rendering them obsolete.
However, scenario analysis is one strategic analysis tool that is becoming
increasingly useful in enabling companies to prepare for the future, especially
in response to the rapid pace of digitalization; it should be noted that scenario
analysis is not a forecasting tool that projects historical trend data into the
future, nor is it a vision for how a company would like the future to be. Rather,
it is a tool to enable managers to produce scenarios of how the future of around
ten to fifteen years out could potentially be dependent upon how a number of
uncertainties develop. Since scenarios promote creative thinking and challenge
conventional wisdom, they enable managers to make better decisions if they
are developed correctly. Companies that have been successful using scenario
52 https://www.weforum.org/agenda/2016/01/digital-disruption-has-only-just-begun/
53 http://www3.weforum.org/docs/Media/KSC_4IR.pdf
325
analysis are Shell, GE, IBM, and UPS, thus, enabling them to retain their
position as market leaders even as their industries transform.
While there are many scenario analysis tools, one of the most commonly
used is a 2x2 scenario matrix that enables managers to consider the question:
“What if this were to happen…?”.54 The scenario matrix is created by choos-
ing two mutually exclusive, critical uncertainties for the axes, which may
significantly change the medium to long-term future in relation to a key
strategic issue for the company. For example, a key strategic issue is whether
to open a new factory in another country, to develop a completely new prod-
uct line, or to invest in a new technology or competitor. One way to determine
this is to think about the question: “What one question would you ask a real
psychic?”
The two critical uncertainties can be uncovered by performing a PESTEL
(political, economic, social, technological, environmental, and legal) analysis
in which these trends are analysed based upon the degree to which they are
uncertain and influential for the key issue. For example, one critical uncer-
tainty may be related to governmental regulations ranging from strict to
relaxed governmental regulations, while another might be related to the
adoption of a new technology from high to low.
Four scenarios are then developed: one for each of the matrix boxes (figure
15.1). These scenarios are in narrative form and describe potential futures in
which the company needs to consider how they will perform. Conclusions
can then be drawn related to the strategic issue, thus, providing the company
with input into how it can build robust strategies and enable it to be agile and
sustain its competitive advantage regardless of which scenario unfolds. Lead-
ing indicators are then determined to scan for early warning signals for the
various scenarios.
Scenario analysis can be used in a more large manner in which all employ-
ees and stakeholders are involved in a lengthy process that could last several
months, which enables the company to develop its vision; it may even be used
in a small-scale manner among a group of a handful of individuals in their
decision processes.
54 A Note on Scenario Planning, Harvard Business School.
326
#est 4–5 most Critical Uncertainties in pairs in scenario framework

– Must be very low correlation between pair of uncertainties
#
Uncert A Must b
e
e completely
nt scen
arios
Scenario 1 Scenario 2
Uncert B Uncert B
Scenario 3 Scenario 4
Uncert A
Figure 15.1. The 2x2 scenario matrix
POLICY CONSIDER ATIONS FOR SWEDEN

While digital disruption has only just begun, the characteristics of the Swed-
ish economy suggest that Sweden is in a relatively good position to benefit
from digitalization. These characteristics include the following: a highly
skilled workforce in information technology, natural sciences and engineer-
ing, user interface and design, communication services, high levels of internet
and smart phone penetration, a global outlook, a focus upon entrepreneurship
and thriving entrepreneurial ecosystems in many medium to large cities, a
leader in the area of FinTech, a stable and advanced welfare state, and Swe-
den’s ideological concern for sustainability (Berkes et al, 2000).
Sweden, however, is still likely to feel the adverse effects of increased digital-
ization and automation. Recent studies predict that 36–60 per cent of the cur-
rent jobs in Sweden – primarily routine jobs – will be lost during the next 20
years as algorithms and robots are employed to perform an increasing number
of tasks (Fölster, 2014). This is partly because its largest sectors of employment
are currently transportation, construction, and metal manufacturing. How-
ever, although 10 per cent of Swedish jobs disappeared between 2006 and 2011
due to automation, the employment rate has remained relatively unchanged.
The reason for the stable employment rate is threefold: 1) digitalization and
327
increased revenues have increased the demand for labour, such as computer
specialists and engineers; 2) higher disposable incomes of highly-skilled indi-
viduals have increased the demand for local service sector jobs; and 3) labour
market reforms, particularly those targeted toward youth labour, have had a
positive impact on employment.
Regarding wages: lower labour mobility and lower wage flexibility cause
more wage rigidity in Sweden compared to the US. Nevertheless, the afore-
mentioned underlying forces do lead to the same wage pressures; commen-
tators and academics are concerned about the fast pace of digitalization and
automation and their potential impact on the labour market, especially since
wages in the Swedish manufacturing industry are relatively high.
As digitalization proceeds, the questions arising are of deep importance for
governments and public sector bodies at all levels. The regulatory framework
must be adapted to significantly address the protection of consumers and
employees while simultaneously enabling the creation of new jobs. How
should taxation and legal systems be designed: to promote innovation as
opposed to hinder it? How should resources be invested in virtual trans-na-
tional clusters of economic activity, such that this investment benefits local
taxpayers and citizens? These tasks will require a balancing act; some policy
suggestions are noted below. Many of these were developed in connection with
a report we published on the Sharing Economy in 2015 (Felländer et al., 2015).
Facilitate a flexible labour market. The labour market may need to be more
flexible, thus, creating incentives both for mobility and for self-employment.
However, such flexibility is difficult to balance with a safety net for vulnera-
ble individuals. The Swedish welfare model, which is characterised by risk
sharing among individuals and security for the individual, has benefited the
economy. In our view, it is even more crucial to aim for social cohesion in the
future. Nevertheless, the model must be adapted to the newly emergent
labour market.
Provide a workforce with skills that meet the future needs of the Swedish economy.
Sweden boasts a strong skill base and a traditionally high share of graduates
in natural sciences and engineering, which bodes well for the future of digital-
ization in the country. However, there are increasing concerns about the
quality of education and the ability to provide a workforce with skills that
meet the needs of Sweden’s future economy. Swedish students rank in the
328
middle of the OECD countries in science education, and the pay-offs of higher
education in Sweden lag behind those of many peer countries (Ketels, 2009).
Attract foreign skill. Sweden does not rank high in attracting foreign skill,
which Ketels (2009) argues is increasingly necessary in the global economy.
Diversity tends to spur creativity that, in turn, leads to a dynamic and inno-
vative climate characterised by higher productivity. In addition, Sweden’s
regulatory scheme and administrative practices are viewed as bureaucratic,
and Sweden has one of the highest levels of taxation in the world, especially
for individuals.
Reassess the tax base. Taxation remains a significant and unresolved issue.
The cost of labour will increase in the future as digitalization spurs a knowl-
edge-intensive service sector. Thus, increasing taxes on labour might not be
the best approach. Other tax bases must be explored instead. Some interna-
tional discussions on this issue include the consideration of both wealth and
property taxes. Furthermore, as we move more toward a platform economy,
who is responsible for reporting the sale and pay the sales or income tax?
What about exchanges in which money does not change hands: for example,
a farming cooperative where individuals receive produce from the farm in
exchange for labour or an Uber-like exchange where the number of hours
spent driving others can be exchanged for rides from other participants?
Similarly, when new currencies are created, should they be treated as goods,
services or currencies for the purposes of VAT? Such debates have already
begun to emerge with regard to Bitcoin: a cryptocurrency.55
Conclusion
The purpose of this chapter, which was intended for a broad audience, has
been to provide a broad overview of digitalization and its many influences
upon firms and society. There are many areas to explore – and only more will
emerge as digitalization continues to penetrate industries and new technolo-
gies emerge and converge with others. Digitalization and technological devel-
opments combined with political and societal changes continue to change the
world, as we know it; therefore, it is important to view these changes in the
“longue durée”. We may seem to be moving forward, but many also profess
that our technological and societal advances have led us to the edge of the cliff
55 http://www.coindesk.com/europe-inches-towards-decision-bitcoin-vat/
329
of the anthropocene era: the period during which human activity has globally
impacted the planet. Paul Dukes writes that we entered this era in 1763, when
the available data indicate the beginning of a growth in the atmospheric con-
centrations of several greenhouse gases (Dukes, 2011). Digitalization, how-
ever, offers promise in reducing the pressure of growth upon pollution and
large cities across the globe. A McKinsey study recently reported that this
trend of large city growth has been broken in countries such as the US and
India, as smaller cities and rural areas are gaining popularity across age
groups. Yet, how it is unclear how we further develop to secure the future of
the planet for generations to come; building roads tends to increase traffic,
rather than relieve it. Recent research in entrepreneurship has recognised the
basic human paradox of seeking to simultaneously fulfil both individual and
collective interests. The question remains: How will future developments
impact this dialectical nature of human beings: Will it encourage us to act
more collectively or perhaps more in our own self-interest?
330
References
Berkes, F., Folke, C., & Colding, J. (Eds.). (2000). Linking Social and Ecological Systems: Management
Practices and Social Mechanisms for Building Resilience. Cambridge University Press.
Brynjolfsson, E., & McAfee, A. (2014). The digitization of just about everything. In The Second
Machine Age: Work, Progress, And Prosperity in a Time of Brilliant Technologies, pp. 57–70.
De Buysere, K., Gajda, O., Kleverlaan, R., Marom, D., & Klaes, M. (2012). A framework for
European crowdfunding. European Crowdfunding Network.
Di Gangi, P. M., Wasko, M. M., & Hooker, R. E. (2010). Getting customers’ ideas to work for
you: Learning from Dell, How to succeed with online user innovation communities. MIS
Quarterly Executive, 9(4).
Dukes, P. (2011). Minutes to Midnight: History and the Anthropocene Era from 1763. Anthem Press.
Eisenmann, T., Parker, G., & Van Alstyne, M. W. (2006). Strategies for two-sided markets.
Harvard Business Review, 84(10), 92.
Felländer, A., Ingram, C., & Teigland, R. (2015). Sharing Economy: Embracing Change with
Caution. Näringspolitiskt Forum Rapport (No. 11).
Fölster, S. (2015). De Nya Jobben i Automatiseringens Tidevarv. Stiftelsen för Strategisk Forskning.
Ketels, C. (2009). Clusters, Cluster Policy, and Swedish Competitiveness in the Global Economy.
Stockholm: Globalisation Council.
McAfee, A., Brynjolfsson, E., & Davenport, T. H. (2012). Big data: the management
revolution. Harvard Business Review, 90(10), 60–68.
McGahan, A. (2000). How industries change, Harvard Business Review, October.
Parker, G. G., Van Alstyne, M. W., & Choudary, S. P. (2016). Platform Revolution: How
Networked Markets are Transforming the Economy--And How to Make Them Work for You.
WW Norton & Company.
Pedersen, J., Kocsis, D., Tripathi, A., Tarrell, A., Weerakoon, A., Tahmasbi, N., & de Vreede,
G. J. (2013). Conceptual foundations of crowdsourcing: A review of IS research. In System
Sciences (HICSS), 2013 46th Hawaii International Conference on (pp. 579–588). IEEE.
Schumpeter, J. (1934). Capitalism, Socialism, and Democracy, New York: Harper & Row.
Skan, J., Lumb, R., Masood, S., & Conway, S. K. (2014). The boom in global fintech
investment: A new growth opportunity for London, Accenture, p 15.
Weill, P., & Woerner, S. L. (2015). Thriving in an increasingly digital ecosystem. MIT Sloan
Management Review, 56(4), 27.
331
About the Authors
Gustav Almqvist is a PhD Candidate in the Center for Media and Economic
Psychology in the Institute for Research (SIR) and the Department of Market-
ing and Strategy at the Stockholm School of Economics (SSE). He holds a BSc.
degree in Business and Economics from SEE and a MSc. in Psychology from
Stockholm University. His research investigates judgment and decision-making
under risk and uncertainty. [email protected].
Per Andersson is a Professor at the Institution for Marketing and Strategy at the
Stockholm School of Economics. He received both an MSc and a PhD from
the Stockholm School of Economics. His research projects focus upon indus-
try change and dynamics connected to changes in information and commu-
nication technologies (ICT). His recent research involves service innovation
processes in media and ICT markets, and how companies in various indus-
tries manage digital transformation processes. [email protected]
Angelica Blom is a PhD Candidate in the Center for Retailing at the Stockholm
School of Economics. She holds a MSc. from SSE. Her research interests are
in omnichannel retailing and customer journeys. Her research has been pub-
lished in Journal of Business Research and Journal of Retailing and Con-
sumer Services. [email protected]
Karl Joachim Breunig is a Professor of Strategic Management and the Head

of the Research Group on Digital Innovation and Strategic Competence in
Organizations (DISCO) at Oslo Business School, Oslo and Akershus Uni-
versity College. He received his PhD from BI Norwegian Business School,
and holds a MSc from London School of Economics. His research focuses
upon knowledge-based value creation, in particular related to strategy, inno-
vation, management control, and internationalisation of knowledge work.
[email protected]
Maria J. Bustamante is a PhD Candidate in the Department of Marketing and

Strategy at the Stockholm School of Economics. Her research interests include
the digital transformation of agriculture and the opportunities and challenges
farms and firms will face as they incorporate new technology into their pro-
cesses. Maria previously spent 10 years in the private sector with responsibilities
333
A BOU T T H E AU T HORS
in marketing, business development, and strategy. Maria holds a M.Sc. in Gen-

eral Management from the Stockholm School of Economics and a B.A. from
The Pennsylvania State University. [email protected]
Martin Carlsson-Wall is Associate Professor in the Department of Accounting

at the Stockholm School of Economics. He received a PhD from the Stock-
holm School of Economics in 2011. His research focuses upon performance
management and how it relates to innovation, creativity, and digitalization.
Martin is currently involved in a European project exploring the role of the
finance function in digital transformation. [email protected]
Robert Demir is a Lecturer in Strategic Management at Lancaster University

Management School and Research Fellow at Ratio. He received his PhD from
the Stockholm University School of Business. His research concerns open
strategy, management innovation, and strategy as practice focusing upon the
emergence, adaptation, and renewal of management and strategy practices.
His research has appeared in British Journal of Management, Long Range Plan-
ning, and International Business Review. He has facilitated executive networks in
Sweden, China, Hong Kong, and India. [email protected]
Anna Felländer is a Digital Economist. She is a visiting fellow at the Swedish

House of Finance at the Stockholm School of Economics and is affiliated with
Sweden’s Royal Institute of Technology. She is a Senior Advisor to the Boston
Consulting Group and an advisor to the Minister of Digitalisation. She is also
a board member and expert advisor to digital start-ups focusing upon Arti-
ficial Intelligence and has recently joined the board for transportstyrelsen.
[email protected]
Henrik Glimstedt is Associate Professor of Business Administration at the

Stockholm School of Economics. He obtained a PhD in History from the
University of Gothenburg. His research focus is on the evolution of global
innovation and competition in wireless communication technologies, and
related high-tech areas. [email protected].
Mikael Hernant is an Assistant Professor at the Center for Retailing, Stockholm
School of Economics/Skövde University. He obtained a PhD from the Stock-
holm School of Economics. Mikael has 20 years of practical retail experience,
having worked at various positions for the ICA Retail Company. His research
addresses the profitability performance of retail stores. [email protected]
334
A BOU T T H E AU T HOR S
Claire Ingram Bogusz is a Researcher at the Stockholm School of Economics.

She obtained her PhD from the Stockholm School of Economics. Her
research interests lie in how code-based tools alter how individuals and organ-
isations organise, especially pursuant to work practices and entrepreneur-
ship. [email protected]
Erik Lakomaa holds a PhD in Economic History from Stockholm School of

Economics. He is the Director of the Institute for Economic and Business
History Research. His research has primarily focused upon how organiza-
tions have historically adapted to external change: for example, new technol-
ogy or new regulations. This project has received funding from the European
Union’s Horizon 2020 Research and Innovation programme under grant
agreement No. 688670. The author is grateful for data and valuable com-
ments by Tino Sanandaji, Alberto Cottica and the members of the Edgery-
ders community. [email protected].
Fredrik Lange is an Assistant Professor in the Center for Retailing at the

Stockholm School of Economics. Fredrik is also the Program Director for the
BSc. Program in Retail Management at the Stockholm School of Economics.
His current research interests relate to consumer behaviour, marketing com-
munications, and retailing. His research has been published in Journal of
Advertising Research, Psychology & Marketing, and Journal of Business Research.
[email protected]
Stefan Lundberg graduated from the Royal institute of Technology (KTH) as a

civil engineer in 1977 in the area of urban planning. He received his PhD
degree from Sweden’s Royal Institute of Technology in 2007 and is now an
Associate Professor at KTH. He has been working in the industry both at
Ericsson and at various housing companies, as well as for the government.
His area of interest is how to plan for the inclusion of the elderly in the city
and how to support ageing-in-place. [email protected]
Lars-Gunnar Mattsson is a Professor Emeritus at the Stockholm School of Eco-

nomics. He is affiliated with the Center for Business Markets and the Mistra
Center for Sustainable Markets at SSE. He obtained his PhD at the Stockholm
School of Economics. His long-term research interests focus upon market
dynamics in a business network perspective. His present research concerns
335
two phenomena: sustainable markets and ICT-based service innovations.

[email protected]
Jan Markendahl is an Associate Professor in Wireless Infrastructure Deploy-

ment and Economics at Sweden’s Royal Institute of Technology. After more
than 20 years in the industry, Jan joined KTH in 2003 as research program
manager and received a PhD degree in 2011 in Techno-Economic Analysis of
Wireless Networks and Services. Jan obtained his Docent degree in 2014 and
has been supervisor for seven PhD students. He has managed techno-economic
research projects and tasks in the EU and national projects. [email protected]
Erik Modig is Assistant Professor at the Stockholm School of Economics. His

research centres upon market communication effectiveness, as well as the
underlying explanation in the mind of the consumer. His research has been
published in various advertising and marketing journals. He has published four
books and his most recent one Bang for the Buck (2017) was awarded the Market-
ing Book of the Year in Sweden. Erik is a popular teacher and a demanded
speaker at organisations and conferences. [email protected]
Staffan Movin is the Managing Director for the Foundation Marketing Tech-
nology Center (MTC). As a non-profit organization, MTC’s mission is to
promote links between practitioners and academics in the field of marketing,
organization, and business management. Staffan has been project leader for
R&D projects addressing business models, flexible entrepreneurship, internet
of things, digital transformation, financial structures, and world-class cus-
tomer experiences. [email protected]
Magnus Mähring is Professor of Business Administration (Information Systems)

at the Stockholm School of Economics. He obtained his PhD at Stockholm
School of Economics. He sits on the Swedish government’s expert committee
for digital investments. His research concerns digitalization, particularly in
relation to organizational decision-making, transformation processes and work
practices. [email protected]
Frida Pemer is Assistant Professor in the Department of Management and Orga-

nization at the Stockholm School of Economics, where she also earned her PhD.
Her research centres upon how private and public organizations use professional
services, as well as what consequences digitalization will have for professional
336
A BOU T T H E AU T HOR S
service firms. Frida has published her research in journals including Human
Relations, Journal of Public Administration Research and Theory, Journal of Supply Chain
Management and Industrial Marketing Management. [email protected]
Sara Rosengren is a Professor in the Center for Retailing at the Stockholm

School of Economics. She is a Foundation of Ica Retailers’ Professor in Busi-
ness Administration, especially Retailing and Head of SSE’s Center for
Retailing. She holds a PhD from the Stockholm School of Economics. Sara’s
current research deals with consumer behaviour, marketing communications,
and retailing: three areas that are currently undergoing major transforma-
tions due to digitalization. [email protected]
Christopher Rosenqvist is an Associate Professor and Lecturer in the Center for

Digital Transformation and Innovation at Stockholm School of Economics
and at SSE Institute for Research. He holds an MSc in Media Technology
from Sweden’s Royal Institute of Technology as well as a PhD and Docent in
the area of Media Production & Business Development. His research concerns
the effects of new technologies upon the ecosystem of actors in different
industrial arenas. [email protected]
Tale Skjølsvik is Associate Professor in Computer Science at the Oslo Business

School: Oslo and Akershus University College. She holds a Ph.D. in Strategic
Management from BI Norwegian Business School and worked as a manage-
ment consultant at Bain & Company: Stockholm and Gemini Consulting:
Oslo. Her research includes: strategic management, digitalization, and purchas-
ing knowledge-intensive service. She has published papers in California Manage-
ment Review, The Service Industries Journal, Journal of Supply Chain Management and
Journal of Public Administration Research and Theory. [email protected]
Torkel Strömsten is an Associate Professor at the Stockholm School of Economics

and a Professor at the Royal Institute of Technology. His research lies in the
interface between management accounting and industrial networks. He is
particularly interested in how control tools are used in value creating processes.
He is also an advisor to start-ups and serves on advisory boards for private and
non-profit organisations. [email protected]
Martin Söndergaard works as a Research Assistant and holds a BSc in Business

and Economics at Stockholm School of Economics, as well as a BSc in
337
Psychology from Stockholm University. He won the Folke Kristensson Schol-

arship in 2016 for Outstanding Essay of the year in Marketing & Strategy for
his paper entitled Effective Mobile Display Advertising: The Impact of Product
Involvement and Creativity. Martin has plans to soon start his doctoral studies at
SSE. His main focus area is understanding consumer psychology through
statistical analysis. [email protected]
Robin Teigland is Professor of Business Administration in Strategic Information

Systems Management at the Stockholm School of Economics. She holds a B.A.
in Economics from Stanford, an M.B.A. from Wharton, an M.A. in Interna-
tional Studies from the University of Pennsylvania, and a PhD from the Stock-
holm School of Economics. Her research focuses on strategy, entrepreneurship,
social networks, and emerging technologies. Robin has five children and was one
of the 125 most influential women in Sweden in 2017. [email protected]
Karl Wennberg is Professor of Management at Linköping University and

Research Fellow at Ratio and the Stockholm School of Economics Institute
for Research (SIR). He obtained his PhD at the Stockholm School of Econom-
ics. His research concerns entrepreneurship and its societal implications.
[email protected]
Adam Åbonde is a PhD student at the Center for Media and Economic Psychol-
ogy (Department of Marketing and Strategy) at the Stockholm School of
Economics (SSE). His research interests revolve around consumer psychology
and behaviour, especially related to information in digital contexts. Adam has
a BSc in Business & Economics and MSc in Business & Management from
SSE. [email protected]
338
About the Language Editor
Karyn McGettigan is a Writer, Editor, and Language Specialist. She has led
communication projects in more than 180 countries, and has edited academic
manuscripts for some of the world’s leading researchers representing 55 pres-
tigious institutions in 20 different countries. Her writing has been translated
into 8 languages and her work has been cited in several peer-reviewed inter-
national journals. Karyn is also an International Correspondent and a Mem-
ber of Sweden’s Foreign Press Association.
Karyn holds an Honours B.A. in French Language and Literature from the
University of Western Ontario, an M.A. from the University of Victoria, a
degree in Broadcast Journalism & Media Communications from the British
Columbia Institute of Technology, a Language and Communications certifi-
cation from the Paris Institute of Language and Culture, and received an
Honours Scholarship to study at the University of Nice.
With dual citizenship in Canada and in Sweden, Karyn McGettigan is
based in Stockholm.
www.mcgettigan.se
339
An Assortment of Our Latest Publications
An assortment of publications from Stockholm School of Economics Institute
for Research and the Stockholm School of Economics. Books and disserta-
tions are published in the language which is indicated by the title. Books can
be ordered by e-mail from [email protected].
PUBLICATIONS SINCE 2014
2017
BOOKS
Einarsson, Stefan & Kallifatides, Markus. Parternas bolag. Ett diskussionsunderlag.
Kallifatides, Markus & Lerpold, Lin (ed). Sustainable Development and Business.
DISSERTATIONS
Freddi, Eleonora. Strategic Ignorance and Social Institutions.
Freij, Åke. Mastering the Impact of Regulatory Change--The Capability of Financial Services Firms to
Manage Interfaces.
Hosseini Tash, Fatemeh. Essays in Entrepreneurial and Household Finance.
Hultin, Lotta. Go with the Flow: Post-humanist Accounts of how Matter Matters in Organizational Change.
Jeremiah, Rupin. R&D Configurations and Innovation Outcomes. The Case of Swedish R&D Offshore.
Lagrelius, Anna-Maria. Doktorns dilemman: Att välja läkemedel med standarder som styrverktyg:
effekter och konsekvenser.
Laurin, Ebba. Box Paradox--How Key Account Management Contributes to Business Model Innovation.
Li, Jieying. Essays in Empirical Finance.
Lundgren, Gustaf. Essays on Job Market Screening, In-Group Bias and School Competition.
Muço, Arieda. On Anti-Corruption Tools in Developing Countries.
Owalla, Beldina. Women’s Entrepreneurial Identities. A Typology Based on Insights from
Entrepreneurship Programs in Two Different Contexts.
Petri, Henrik. Lexicographic and Social Choice, Takeovers, and Fixed Points.
Sebhatu, Abiel. Deregulation, Institutional Change, and Entrepreneurship in the Swedish Education
System: Intended and Unintended Effects of Competition.
Sokolovski, Valeri. Essays on Currency Risk and Financial Frictions.
Stigzelius, Ingrid. Producing Consumers: Agencing and Concerning Consumers to Do Green in
Everyday Food Practices.
Yetis Larsson, Zeynep. Open Entrepreneurship: Investigating Entrepreneurship in Open Source
Software Communities.
Wikberg, Erik. Artful Organizing: Essays on Places, Measurements, and Money Flows in the
Contemporary Cultural World.
341
A N A SSORTMEN T OF OU R LAT EST PU BLICAT IONS
Åhblom, Per. The Financial Performance--A study of how financial numbers become meaning ful.
Åkestam, Nina. Understanding Advertising Stereotypes. Social and Brand-related Effects of
Stereotyped versus Non-Stereotyped Portrayals in Advertising.
Ählström, Jenny. ”Alla ska behandlas med respekt”. Uppkomst och utveckling av narrativet om
mänskliga rättigheter i globala leverantörskedjor.
2016
BOOKS
Einarsson, Stefan. Medlemmar, ideologi och strategi. En studie av IOGT-NTO 1970–2009. Forskning
i Fickformat.
Schuster, Walter. Finansiell Redovisning.
Sjöstrand, Sven-Erik. Nordic Corporate Governance. An extensive in-depth study of corporate governance
and board practices in 36 large companies.
Sjöstrand, Sven-Erik. Nordisk bolagsstyrning. En närstudie av ägarstyrning och styrelseprocesser i 36
storbolag.
Wahlund, Richard (red.) Risker och riskhantering. I näringsliv och samhälle.
DISSERTATIONS
Amberg, Niklas. Firms and Banks in the Macroeconomy: Empirical Essays using Microeconometric Methods.
Ceh, Ana Maria. Public Debt, Private Liquidity and Welfare.
Chen, Jun. Essays on Information Acquisition.
Efendic, Nedim. Expatriate Entrepreneurship: The Role of Accelerators in Network Formation and
Resource Acquisition.
Facchinello, Luca. On the Role of Information in Educational Choice.
Forsell, Eskil. Experimental Testing of Old and New Hypotheses in Economics.
Fröberg, Emelie. Seeking Alpha – and financing it. Empirical studies of the impact of information acquisition
behavior, market beliefs and risk attitude on fund performance among equity fund managers in Sweden.
Gillqvist, Anna-Stina. Conversations on Accounting Practices--A Study of an Enforcement Body in a
Time of Regulatory Change.
Heidari, Mahdi. Essays on Financial Market Anomalies and Investment Strategies.
Karmaziene, Egle. Essays in Empirical Finance.
Karlberg, Gabriel. Styrning av outsourcad offentlig service. En balansakt mellan fasthet och flexibilitet.
Karsberg, John. Reception, reception, reception. The effects of receiver context on advertising effectiveness.
Larsson von Garaguly, Joacim. Vasaloppet – Resan från skidtävling och skidlöpare till produkter och
kunder. En studie om kommersialisering och professionalisering.
Lopez Aliouchkin, Ricardo. Essays in Financial Economics.
Lundin, Erik. Empirical Essays on Strategic Behavior in the Electricity and Water Sectors.
Murtic, Adis. Soaking Up Knowledge: A Multi-level Analysis and Conceptualization of Absorptive Capacity.
Ringbo, Joel. In search of the optimal grocery store layout. Examining the effect of placement order for
complementary product categories in the store environment.
Thörnqvist, Tomas. Essays in Household Finance.
342
Töndevold Liljedal, Karina. Communicated Consumer Co-creation: Consumer Response to Consumer

Co-creation in New Product and Service Development.
Özbek, Nurgül. Entering a Global Play: Insights into Swedish Small Life science Firms’ Legitimation
in International Networks.
2015
BOOKS
Bay, Charlotta. Finansens folkdräkt Om att översätta ekonomi så folk förstår. Forskning i Fickformat.
Sölvell, Örjan. On Strategy & Competitiveness. 10 recipes for analytical success.
Yttermyr, Olga. Skolans nya värld Att organisera en kundvalsmarknad – med grundskoleutbildningen
i fokus. Forskning i Fickformat.
DISSERTATIONS
Berg, Hanna. Faces of Marketing: Examining Consumer Responses to Depictions of People in Marketing.
Christner, Carl Henning. The Valuable Organisation--A Study of How Activities are Calculated,
Controlled and Made Valuable.
Di Casola, Paola. Essays in Macroeconomics and Institutional Economics.
Khapko, Mariana. Essays in Financial Economics.
Koptyug, Nikita. Essays on Consumers, Risk and Rationality.
Luistro Jonsson, Marijane. Cooperating for Sustainability: Experiments on Uncertainty, Conditional
Cooperation and Inequality.
Milonas, Kristoffer. Essays in Empirical Finance.
Morgulis-Yakushev, Sergey. Exploring Fit: Essays on the Role of Organizational Cultural Fit for
Knowledge Transfer.
Nilsson Altafi, Sofia. Evaluations as Mirrors and Co-constructors: An Empirical Investigation of a
Microfinance Rating Agency and Its Rating Practices from 1999–2014.
Shir, Nadav. Entrepreneurial Well-Being: The Payoff Structure of Business Creation.
Sichlimiris, Spyridon. Public Debt, Business Cycles and Institutions.
Wehrmüller, Simon. Credit, Labor and the Great Recession.
Xia, Qing. Creditor Rights and Innovation: Evidence from China.
2014
BOOKS
Beckerman, Carina. Skepparen först, båten sedan. En bok om Marcus Wallenbergs värld och synen
på ledarskap.
Jacobsson, Bengt & Sundström Göran. Demokrati och förvaltning. En festskrift till Rune Premfors.
Näslund, Lovisa. Får jag lov? Förtroendeskapande bland konsulter och teaterregissörer. Forskning
i Fickformat
Sjöstrand, Sven-Erik & Hammarkvist, Karl-Olof. Om ägarstyrning i statligt ägda företag.
Ingår i ”Ekonomiskt värde och samhällsnytta – förslag till ny statlig ägarförvaltning.”
Expertbilaga i Ägarutredningen (SOU 2012:14).
Söderlund, M. (red) Marknadsföring och påverkan på konsumenten.
343
DISSERTATIONS
Bernergård, Axel. Best replying, to yourself and others.
Bird, Miriam. The Impact of the Family on Entrepreneurial Outcomes: The Role of Social Embeddedness.
Brattström, Anna. Trust in a Product Development Context: Drivers, Dynamics and Consequences.
Breckenfelder, Johannes. Empirical Essays in Financial Economics.
Crosta, Alberto. The Effects of Credit Rating and Watchlist Announcements on the U.S. Corporate
Bond Market.
de Rezende, Rafael B. Essays on Macro-Financial Linkages.
Elger, Paul. Essays on Taxation and Consumption.
Faragó, Ádám. Essays on Disappointment Aversion in Portfolio Choice and Asset Pricing.
Hagbjer, Eva. Navigating a Network of Competing Demands--Accountability as Issue Formulation and
Role Attribution Across Organisational Boundaries.
Hartwig, Markus. Interest Alignment, Delegation of Authority and Economic Rent – A Collection
of Empirical Papers.
Hederos Eriksson, Karin. Essays on Inequality, Gender and Family Background.
Junker, Sven-Olof. Att skapa gemenskap: hur beslut fattas i en EU-myndighet.
Lazarczyk Carlson, Ewa. Essays on Electricity Markets: Information and Trading.
Löfgren, Angelika. International Network Competitiveness: Technical and Foreign Market Knowledge
Development in International SME Networks.
Mengistu, Andualem. Empirical essays on the monetary markets of developing countries.
Molin, Jonas. Business Streamlining: Toward a Substantive Theory of the Management of Outsourced
Business Processes.
Nilsson, Andreas. Financing of Nonprofits and Social Enterprises.
Rosenström, Martin. Att skapa en marknad: marknad och politisk styrning i symbios?
Sandberg, Anna. Empirical Studies on Sources of Inequality.
Schnitzler, Jan. Empirical Essays in Finance.
Svahn, Mattias. Pervasive Persuasive Games: The Case of Impacting Energy Consumption.
Vardardottir, Arna. Family and Friends: Essays on Applied Microeconometrics.
Yildirim, Mustafa. Incentives in Contests.
344
Digitalization has arrived.
Digitalization disrupts markets. Changes in power and structures

in a fast-paced environment demands strategic and insightful
change. A change leaders must act upon.
The impact upon organisations is multi-dimensional and profound,

affecting both internal and external processes and structures in new
and unexpected ways. This book serves as a tool to support
managers and other stakeholders in pursuing digital transforma-
tion. An inspiring collection of chapters from 27 scholars across
various academic disciplines provide several insights, frameworks,
and perspectives that will help you leverage and govern organisatio-
nal change and digital transformation.
This inspiring collection of current research can assist you in facing

key challenges in today’s organisations, in the quest to adapt to
ever-evolving business environments. This book examines new
demands and behaviours, and discusses how businesses need to
adapt and re-organise in order to bridge the gap to the digital
customer. These visions and actions on digitalization can help
corporations and organisations discover new ways of earning
money and delivering value.
This is just the beginning.
ISBN 978-91-86797-31-7

Managing Digital Transformation Med Omslag PDF

Uploaded by

Copyright:

Available Formats

Managing Digital Transformation Med Omslag PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Managing Digital Transformation Med Omslag PDF

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What challenges does digitalization pose for organizations?

What challenges does digitalization pose for organizations?

M ANAGING

Per Andersson, Staffan Movin,

Karyn McGettigan, Language Editor

Chair: Professor Richard Wahlund

Managing Digital Transformation

Stiftelsen Marknadstekniskt centrum, MTC

This is the topic of this book, which we dedicate to him.

Digitalization: Different Perspectives

2. Reaping Value From Digitalization in Swedish Manufacturing Firms:

3. Digital Platforms: A Critical Review of the Core Concepts 65

The Digital Customer

5. Digital Trace Data: Which Data Should we Collect

6. Managing Digital Media Investments 133

8. Robotisation of Accounting in Multi-National Companies:

10. Explaining the Behaviour of News Consumption 203

11. Digital Transformation Supporting Public Service Innovation:

Business Models and Ecosystems

13. Digitalization, Collective Intelligence,

14. AgTech and the City: The Case of Vertical Farming

About the Authors 333

Stockholm, January 2018

PART 2: THE NE W DIGITAL CUSTOMER

PART 3: THE RE-ORGANISATION IN ORDER

PART 4: BUSINESS MODELS AND ECOSYSTEMS

Staffan Movin, Stiftelsen Marknadstekniskt Centrum

Strategic Challenges of Digital

“The current generation of entrepreneurs is unlike any other. Empowered by

Our second example is from the medical technology industry. A business

The company thought it could do something about the masses of machine

technology, engineering, and mathematics. The small company’s mission is

Companies report on a set of major strategic challenges when engaging in

About the Underlying Research

themes, which included the following issues: technical, customer exchange

Table 1.1: Cases of Digital Transformation Processes: Studied in 15 Different Areas

Ten Digitalization Challenges

1. M ANAGING THE IMPORTANT ROLES AND

technologies and platforms require actors to create new “meanings” to prod-

2. M ANAGING BIG DATA:

3. COPING WITH USER-CENTRIC SYSTEMS:

of consumers or organisational customers start sharing products and begin

4. GO-TO-M ARKET AND SCAL ABILIT Y:

ings is often experienced as being challenging due to the uncertainties regard-

5. M ANAGING DIGITALIZ ATION AND SERVICE TR ANSFOR M ATION:

vehicle” based upon new digital infrastructures; and, public transportation

6. SHAPING NE W CO-OPER ATIVE BUSINESS MODELS

partnerships associated with broader sets of cooperating organisations. This

7. M ANAGING THE NE W DIGITAL ENTREPRENEURS

8. CRE ATING NE W BUSINESSES ACROSS

9. RE-ORGANISATION OF DIGITALIZING CUSTOMERS:

10. M ANAGING THE POLITICAL AND

policy making – when it comes to digitalization, in general, and issues con-

Digitalization Challenges are Connected

Implications: Pressures on Internal

organisations, digitalization often requires administrations to explore new

NE W SKILLS, RESOURCES, AND INTERNAL CAPABILITIES