Big Data and Analytics: Vincenzo Morabito

Vincenzo Morabito
Big Data and

Analytics
Strategic and Organizational Impacts
Big Data and Analytics
Vincenzo Morabito

Strategic and Organizational Impacts
123
Vincenzo Morabito
Department of Management and Technology
Bocconi University
Milan
Italy
ISBN 978-3-319-10664-9 ISBN 978-3-319-10665-6 (eBook)

DOI 10.1007/978-3-319-10665-6
Library of Congress Control Number: 2014958989
Springer Cham Heidelberg New York Dordrecht London

© Springer International Publishing Switzerland 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made.
Printed on acid-free paper
Springer International Publishing AG Switzerland is part of Springer Science+Business Media

(www.springer.com)
Foreword
Few organizations understand how to extract insights and value from the recent
explosion of “Big Data.” With a billion plus users on the online social graph doing
what they like to do and leaving a digital trail, and with trillions of sensors now
being connected in the so-called Internet of Things, organizations need clarity and
insights into what lies ahead in deploying these capabilities. While academic
scholars are just beginning to appreciate the power of big data analytics and new
media to open up a fascinating array of questions from a host of disciplines, the
practical applicability of this is still lacking. Big data and analytics touches multiple
disciplines ranging from sociology, psychology, and ethics to marketing, statistics,
and economics, as well as law and public policy. If harnessed correctly it has the
potential to solve a variety of business and societal problems.
This book aims to develop the strategic and organizational impacts of Big Data
and analytics for today’s digital business competition and innovation. Written by an
academic, the book has nonetheless the main goal to provide a toolbox suitable to
be useful to business practice and know-how. To this end Vincenzo as in his former
books has structured the content into three parts that guide the reader through how
to control and govern the innovation potential of Big Data and Analytics. First, the
book focuses on Strategy (Part I), analyzing how Big Data and analytics impact on
private and public organizations, thus, examining the implications for competitive
advantage as well as for government and education. The last chapter provides an
overview of Big Data business models, creating a bridge to the content of Part II,
which analyzes the managerial challenges of Big Data and analytics governance
and evaluation. The conclusive chapter of Part II introduces the reader to the
challenges of managing change required by an effective use and absorption of Big
Data and analytics, actually trying to complement IT and non-IT managers’ per-
spective. Finally, Part III discusses through structured and easy to read forms a set
of cases of Big Data and analytics initiatives in practice at a global level in 2014.
Use this book as a guide to design your modern analytics-enabled organization.
Do not be surprised if it resembles a large-scale real-world laboratory where
employees design and conduct experiments and collect the data needed to obtain
answers to a variety of questions, from peer influence effects, the influence of
v
vi Foreword
dynamic ties, pricing of digital media, anonymity in online relationships, to

designing next-generation recommender systems and enquiries into the changing
preference structures of Generation Y and Z consumers. This is a bold new frontier
and it is safe to say we ain’t seen nothing yet.
Ravi Bapna
Preface
Notwithstanding the interest and the hype that surround Big Data as a key trend as
well the claimed business potentiality that it may offer the coupling with a new
breed of analytics, the phenomenon has been yet not fully investigated from a
strategic and organizational perspective. Indeed, at the moment of writing this book,
apart from a series of articles that appeared on the Harvard Business review by
McAfee and Brynjolfsson (2012) and on MIT Sloan Management Review by
Lavalle et al. (2011) and Davenport et al. (2012), most of the published mono-
graphic contributions concern technical, computational, and engineering facets of
Big Data and analytics, or oriented toward high-level societal as well as general
audience business analyses.
An early joint academics-practitioners effort to provide a unified and compre-
hensive perspective has been carried out by the White Paper resulting from joint
multidisciplinary contributions of more than 130 participants from 26 countries at
the World Summit on Big Data and Organization Design held in Paris at the
Université Panthéon-Sorbonne during May 16–17, 2013 (Burton et al. 2014).
However, it is worth to be mentioned that since 2013 new editorial initiatives have
been launched such as, e.g., the Big Data journal (Dumbill 2013). Thus, following
up the insights discussed in (Morabito 2014), the present book aims to fill the gap,
providing a strategic and organizational perspective on Big Data and analytics,
identifying the challenges, ideas, and trends that may represent “food for thought”
to practitioners. Accordingly, each topic considered will be analyzed in its technical
and managerial aspects, also through the use of case studies and examples. Thus,
while relying on academic production as well, the book aims to describe problems
from the viewpoints of managers, adopting a clear and easy-to-understand
language, in order to capture the interests of top managers and graduate students.
Consequently, this book is unique for its intention to synthesize, compare, and
comment on major challenges and approaches to Big Data and analytics, being a
simple yet ready to consult toolbox for both managers and scholars.
In what follows we provide a brief overview, based on our previous work as well
(Morabito 2014), on Big Data drivers and characteristics suitable to introduce their
discussion also with regard to analytics in the further chapters of this book, whose
outline concludes this introduction.
vii
viii Preface
Big Data Drivers and Characteristics
The spread of social media as a main driver for innovation of products and services
and the increasing availability of unstructured data (images, video, audio, etc.) from
sensors, cameras, digital devices for monitoring supply chains and stocking in
warehouses (i.e., what is actually called internet of things), video conferencing
systems and voice over IP (VOIP) systems, have contributed to an unmatched
availability of information in rapid and constant growth in terms of volume. As for
these issues, an interesting definition of “Big Data” has been provided by Edd
Dumbill in 2013:
Big data is data that exceeds the processing capacity of conventional database systems. The
data is too big, moves too fast, or doesn’t fit the structures of your database architectures. To
gain value from this data, you must choose an alternative way to process it (Dumbill 2013).
As a consequence of the above scenario and definition, the term “Big Data” is
dubbed to indicate the challenges associated with the emergence of data sets whose
size and complexity require companies to adopt new tools and models for the
management of information. Thus, Big Data require new capabilities (Davenport
and Patil 2012) to control external and internal information flows, transforming
them into strategic resources to define strategies for products and services that meet
customers’needs, increasingly informed and demanding.
However, Big Data computational as well as technical challenges call for a
radical change to business models and human resources in terms of information
orientation and a unique valorization of a company information asset for invest-
ments and support for strategic decisions. At the state of the art the following four
dimensions are recognized as characterizing Big Data (IBM; McAfee and Bry-
njolfsson 2012; Morabito 2014; Pospiech and Felden 2012):
• Volume: the first dimension concerns the unmatched quantity of data actually
available and storable by businesses (terabytes or even petabytes), through the
Internet: for example, 12 terabytes of Tweets are created everyday into improved
product sentiment analysis (IBM).
• Velocity: the second dimension concerns the dynamics of the volume of data,
namely the time-sensitive nature of Big Data, as the speed of their creation and
use is often (nearly) real-time.
• Variety: the third dimension concerns type of data actually available. Besides,
structured data traditionally managed by information systems in organizations,
most of the new breed encompasses semi-structured and even unstructured data,
ranging from text, log files, audio, video, and images posted, e.g., on social
networks to sensor data, click streams, e.g., from Internet of Things.
• Accessibility: the fourth dimension concerns the unmatched availability of
channels a business may increase and extend its own data and information asset.
• It is worth noting that at the state of the art another dimension is actually con-
sidered relevant to Big Data characterization: Veracity concerns quality of data
and trust of the data actually available at an incomparable degree of volume,
Preface ix
velocity, and variety. Thus, this dimension is relevant to a strategic use of Big
Data and analytics by businesses, extending in terms of scale and complexity the
issues investigated by information quality scholars (Huang et al. 1999; Madnick
et al. 2009; Wang and Strong 1996), for enterprise systems mostly relying on
traditional relational database management systems.
As for drivers, (Morabito 2014) identified cloud computing as a relevant one,
besides social networks, mobile technologies, and Internet of Things (IoTs). As
pointed out by Pospiech and Felden (2012), at the state of the art, cloud computing
is considered a key driver of Big Data, for the growing size of available data
requires scalable database management systems (DBMS). However, cloud com-
puting faces IT managers and architects the choice of either relying on commercial
solutions (mostly expensive) or moving beyond relational database technology,
thus, identifying novel data management systems for cloud infrastructures (Agrawal
et al. 2010, 2011). Accordingly, at the state of art NoSQL (Not Only SQL)1 data
storage systems have been emerging, usually not requiring fixed table schemas and
not fully complying nor satisfying the traditional ACID (Atomicity, Consistency,
Isolation, and Durability) properties. Among the programming paradigms for
processing, generating, and analyzing large data sets, MapReduce2 and the open
source computing framework Hadoop have received a growing interest and
adoption in both industry and academia.3
Considering velocity, there is a debate in academia about considering Big Data
as encompassing both data “stocks” and “flows” (Davenport 2012). For example, at
the state of the art Piccoli and Pigni (2013) propose to distinguish the elements of
digital data streams (DDSs) from “big data”; the latter concerning static data that
can be mined for insight. Whereas digital data streams (DDSs) are “dynamically
evolving sources of data changing over time that have the potential to spur real-time
action” (Piccoli and Pigni 2013). Thus, DDSs refer to streams of real-time infor-
mation by mobile devices and IoTs, that have to be “captured” and analyzed real-
time, provided or not they are stored as “Big Data”. The types of use of “big” DDSs
may be classified according to those Davenport et al. (2012) have pointed out for
Big Data applications to information flows:
1
Several classifications of the NoSQL databases have been proposed in literature (Han et al.
2011). Here we mention Key-/Value-Stores (a map/dictionary allows clients to insert and request
values per key) and Column-Oriented databases (data are stored and processed by column instead
of row). An example of the former is Amazon’s Dynamo; whereas HBase, Google’s Bigtable, and
Cassandra represent Column-Oriented databases. For further details we refer the reader to
(Han et al. 2011; Strauch 2010).
2
MapReduce exploit, on the one hand, (i) a map function, specified by the user to process a key/
value pair and to generate a set of intermediate key/value pairs; on the other hand, (ii) a reduce
function that merges all intermediate values associated with the same intermediate key (Dean and
Ghemawat 2008). MapReduce has been used to complete rewrite the production indexing system
that produces the data structures used for the Google web search service (Dean and Ghemawat
2008).
3
See for example how IBM has exploited/integrated Hadoop (IBM et al. 2011).
x Preface
• Support customer-facing processes: e.g., to identify fraud or medical patients’

health risk.
• Continuous process monitoring: e.g., to identify variations in customer senti-
ments toward a brand or a specific product/service or to exploit sensor data to
detect the need for intervention on jet engines, data centers machines, extraction
pump, etc.
• Explore network relationships on, e.g., Linkedin, Facebook, and Twitter to
identify potential threats or opportunities related to human resources, customers,
competitors, etc.
As a consequence, we believe that the distinction between DDSs and Big Data is
useful to point out a difference in scope and target of decision making, and analytic
activities, depending on the business goals and the type of action required. Indeed,
while DDSs may be suitable to be used for marketing and operations issues, such as
customer experience management in mobile services, Big Data refer to the infor-
mation asset an organization is actually able to archive, manage, and exploit
for decision making, strategy definition, and business innovation (McAfee and
Brynjolfsson 2012).
Having emphasized the specificity of DDS, we now focus on Big Data and
analytics applications as also discussed in (Morabito 2014).
As shown in Fig. 1 they cover many industries, spanning from finance (banks
and insurance), e.g., improving risk analysis and fraud management, to utility and
manufacturing, with a focus on information provided by sensors and IoTs for
improved quality control, operations or plants performance, and energy manage-
ment. Moreover, marketing and service may exploit Big Data for increasing cus-
tomer experience, through the adoption of social media analytics focused on
sentiment analysis, opinion mining, and recommender systems.
As for public sector (further discussed in Chap. 2), Big Data represents an
opportunity, on the one hand, e.g., for improving fraud detection as tax evasion
control through the integration of a large number of public administration
databases; on the other hand, for accountability and transparency of government
and administrative activities, due to the increasing relevance and diffusion of open
data initiatives, making accessible and available for further elaboration by con-
stituencies of large public administration data sets (Cabinet Office 2012; Zuiderwijk
et al. 2012), and participation of citizens to the policy making process, thanks to the
shift of many government digital initiatives towards an open government per-
spective (Feller et al. 2011; Lee and Kwak 2012; Di Maio 2010; Nam 2012).
Thus, Big Data seem to have a strategic value for organizations in many
industries, confirming the claim by Andrew McAfee and Brynjolfsson (2012) that
data-driven decisions are better decisions, relying on evidence of (an unmatched
amount of) facts rather than intuition by experts or individuals. Nevertheless, we
believe that management challenges and opportunities of Big Data need further
discussion and analyses, the state of the art currently privileging their technical
facets and characteristics. That is the motivation behind this book, whose outline
follows.
Preface xi
Sentiment Banks /
Analysis Insurances
Opinion Mining
Social Media Risk Analysis
Analytics Fraud detection
Recommender Threat Analysis
systems Credit scoring
…
Marketing/ BIG DATA Utilities /

Services and Analytics Manufacturing
Applications
Fraud detection
Tax evasion control
Reduction in Quality management
consumption of public and control
utilities Sensor Data Fusion
… …
Public Sector
Fig. 1 Big Data Applications. Adapted from (Morabito 2014)
Outline of the Book
The book argument is developed along three main axes, likewise. In particular, we
consider first (Part I) Strategy issues related to the growing relevance of Big Data
and analytics for competitive advantage, also due their empowerment of activities
such as, e.g., consumer profiling, market segmentation, and new products or ser-
vices development. Furthermore, the different chapters will also consider the stra-
tegic impact of Big Data and analytics for innovation in domains such as
government and education. A discussion of Big Data-driven Business Models
conclude this part of the book. Subsequently, (Part II) considers Organization,
focusing on Big Data and analytics challenges for governance, evaluation, and
managing change for Big Data-driven innovation. Finally (Part III), the book will
present and review case studies of Big Data Innovation Practices at the global level.
Thus, Chap. 8 aims to discuss examples of Big Data and analytics applications in
practice, providing fact-sheets suitable to build a “map” of 10 interesting digital
innovations actually available worldwide. Besides an introduction to the factors
considered in the choice of each innovation practice, a specific description of it will
be developed. Finally, the conclusion will provide a summary of all arguments of
the volume together with general managerial recommendations.
Vincenzo Morabito
xii Preface
References
Agrawal, D., Das, S., El Abbadi, A.: Big data and cloud computing: new wine or just new bottles?
Proc. VLDB Endow. 3, 1647–1648 (2010)
Agrawal, D., Das, S., El Abbadi, A.: Big Data and Cloud Computing: Current State and Future
Opportunities. EDBT, ACM. pp. 530–533. March 22–24, Sweden (2011)
Burton, R.M., Mastrangelo, D., Salvador F.(eds.): Big data and organization design. J. Organ. Des.
3(1), (2014)
Cabinet Office UK: Open Data White Paper—Unleashing the Potential. (2012)
Davenport, T.H., Barth, P., Bean, R.: How “big data” is different. MIT Sloan Manag. Rev. 54(1),
43–46 (2012)
Davenport, T.H., Patil, D.J.: Data scientist: The sexiest job of the 21st century. Harv. Bus. Rev.
October, (2012)
Dean, J., Ghemawat, S.: MapReduce: Simplified data processing on large clusters. Commun.
ACM. 51(1), 1–13 (2008)
Di Maio, A.: Gartner open government maturity model. Gartner (2010)
Dumbill, E.: Making sense of big data (editorial). Big Data. 1(1), 1–2 (2013)
Feller, J., Finnegan, P., Nilsson, O.: Open innovation and public administration: Transformational
typologies and business model impacts. Eur. J. Inf. Syst. 20, 358–374 (2011). doi: 10.1057/
Ejis.2010.65
Han, J., Haihong, E., Le, G., Du, J.: Survey on NoSQL database. 6th International Conference on
Pervasive Computing and Applications (ICPCA). pp. 363–366 (2011). doi: 10.1109/ICPCA.
2011.6106531
Huang, K.T., Lee, Y., Wang, R.Y.: Quality, information and knowledge. Prentice-Hall, Inc (1999)
IBM, Zikopoulos, P., Eaton, C.: Understanding Big Data: Analytics for Enterprise Class Hadoop
and streaming data, 1st edn. McGraw-Hill Osborne Media (2011)
IBM: What is big data?, http://www-01.ibm.com/software/data/bigdata/what-is-big-data.html.
Accessed 7 Jan 2015
Lavalle, S., Lesser, E., Shockley, R., Hopkins, M.S., Kruschwitz, N.: Big Data, Analytics and the
Path From Insights to Value. MIT Sloan Manag. Rev. 52(2), (2011)
Lee, G., Kwak, Y.H.: An open government maturity model for social media-based public
engagement. Gov. Inf. Q. 29(4), 492–503 (2012)
Madnick, S.E., Wang, R.Y., Lee, Y.W., Zhu, H.: Overview and Framework for Data and
Information Quality Research. J. Data Inf. Qual. 1, 1–22 (2009). doi: 10.1145/1515693.1516680
McAfee, A., Brynjolfsson, E.: Big data: The management revolution. Harv. Bus. Rev. 61–68
(2012)
Morabito, V.: Big data. Trends and Challenges in Digital Business Innovation, pp. 3–21 Springer,
Cham Heidelberg New York Dordrecht London (2014)
Morabito, V.: Trends and Challenges in Digital Business Innovation. Springer (2014)
Nam, T.: Citizens’ attitudes toward open government and government. Int. Rev. Adm. Sci. 78(2),
346–368 (2012)
Piccoli, G., Pigni, F.: Harvesting external data: The potential of digital data streams. MIS Q. Exec.
12(1), 143–154 (2013)
Pospiech, M., Felden, C.: Big data—A State-of-the-Art. AMCIS 2012. (2012)
Strauch, C.: NoSQL databases. Lect. Notes Stuttgart Media. 1–8 (2010)
Wang, R.Y., Strong, D.M.: Beyond accuracy: what data quality means to data consumers.
J. Manag. Inf. Syst. 12(4), 5–33 (1996)
Zuiderwijk, A., Janssen, M., Choenni, S.: Open Data Policies: Impediments and Challenges. 12th
European Conference on eGovernment (ECEG 2012). pp. 794–802, Barcelona, Spain (2012)
Acknowledgments
This book is the result of the last two years of research, where several people are
worth to be acknowledged for their support, useful comments and cooperation.
A special mention to Prof. Vincenzo Perrone at Bocconi University, Prof. Vallabh
Sambamurthy, Eli Broad Professor at Michigan State University, and Prof. Franco
Fontana at LUISS University as main inspiration and mentors.
Moreover, I acknowledge Prof. Giuseppe Soda, Head of the Department of
Management and Technology at Bocconi University, and all the other colleagues at
the Department, in particular Prof. Arnaldo Camuffo, Prof. Anna Grandori, Prof.
Severino Salvemini, and Prof. Giuseppe Airoldi, all formerly at the Institute of
Organization and Information Systems at Bocconi University, who have created a
rich and rigorous research environment where I am proud to work.
I acknowledge also some colleagues from other universities with whom I’ve had
the pleasure to work, whose conversations, comments, and presentations provided
precious insights for this book: among others, Prof. Anindya Ghose at New York
University’s Leonard N. Stern School of Business, Prof. Vijay Gurbaxani at
University of California Irvine, Prof. Saby Mitra at Georgia Institute of Technology,
Prof. Ravi Bapna at University of Minnesota Carlson School of Management,
George Westerman at MIT Center for Digital Business, Stephanie Woerner at MIT
Center for Information Systems Research, Prof. Ritu Agarwal at Robert H. Smith
School of Business, Prof. Lynda Applegate at Harvard Business School, Prof. Omar
El Sawy at Marshall School of Business, Prof. Marco de Marco at Unversità
Cattolica del Sacro Cuore di Milano, Prof. Tobias Kretschmer, Head of Institute for
Strategy, Technology and Organization of Ludwig Maximilians University, Prof.
Marinos Themistocleous at the Department of Digital Systems at University of
Piraeus, Prof. Chiara Francalanci at Politecnico di Milano, Wolfgang König at
Goethe University, Luca Giustiniano at LUISS University, Prof. Zahir Irani at
Brunel Business School, Prof. Sinan Aral at NYU Stern School of Business, Prof
Nitham Mohammed Hindi and Prof. Adam Mohamedali Fadlalla of Qatar Univer-
sity, Antonio de Amescua and Román López-Cortijo of Universidad Carlos III de
Madrid and Ken and Jane Laudon.
Furthermore, I want to gratefully acknowledge all the companies that have
participated to the research interviews, case studies, and surveys.
xiii
xiv Acknowledgments
In particular, for the Financial Institutions: Agos Ducato, Banca Carige, Banca
Euromobiliare, Banca Fideuram, Banca d’Italia, Banca Mediolanum, Banco
Popolare, Banca Popolare dell’Emilia Romagna, Banca Popolare di Milano,
Banca Popolare di Sondrio, Banca Popolare di Vicenza, Banca Popolare di Bari,
Banca Sistema, Barclays, BCC Roma, BNL-BNP Paribas, Borsa Italiana, Carip-
arma Credit Agricole, CACEIS Bank Luxemburg, Carta Si, Cassa Depositi e
Prestiti, Cassa di Risparmio di Firenze, Cedacri, Che Banca!, Compass, Corner
Bank, Credito Emiliano, Deutsche Bank, Dexia, HypoVereinsbank, Istituto Cent-
rale delle Banche Popolari Italiane, ING Direct, Intesa SanPaolo, Intesa SanPaolo
Servitia, Istituto per le Opere Religiose, Luxemburg Stock Exchange, JP Morgan
Chase, Key Client, Mediobanca, Monte Titoli, Banca Monte dei Paschi, Poste
Italiane, SEC Servizi, Société Européene de Banque, Standard Chartered, Royal
Bank of Scotland, UBI Banca, Unicredit, Unicredit Leasing, Veneto Banca and
WeBank.
For the Insurance sector: Allianz, Assimoco, Aspe Re, Cardif, Coface, Ergo
Previdenza, Europe Assistance, Assicurazioni Generali, Groupama, Munich RE,
Poste Vita, Reale Mutua, Novae, Sara Assicurazioni, UnipolSai, Vittoria Assicur-
azioni and Zurich.
For the Industrial Sector: ABB, Accenture, Acea, Aci Informatica, Acqua Minerale
S. Benedetto, Adidas, Alpitour, Alliance Boots, Amadori, Amazon, Amplifon, Anas,
Angelini, ArcelorMittal, Armani, Astaldi, ATAC, ATM, AstraZeneca, Arval,
Auchan, Audi, Augusta Westland, Autogrill, Autostrade per l’Italia, Avio, Baglioni
Hotels, BMW, BASF, Barilla, Be Consulting, Benetton, Between, Business Inte-
gration Partners, Brembo, Bravo Fly, BskyB, BSH, BOSH, Boeing Defence,
Cementir, Centrica Energy, Cerved, Chiesi Farmaceutici, CNH Industrial, Coca Cola
HBC, Coop Italia, Costa Crociere, D’Amico, Danone, Daimler, De Agostini, Diesel,
Dimar, Dolce and Gabbana, General Electric, Ducati, Elettronica, Edipower, Edison,
Eni, Enel, ENRC, ERG, Fastweb, Ferservizi, Fincantieri, Ferrari, Ferrovie dello Stato,
FCA, Finmeccanica, GlaxosmithKline, GE Capital, GFT Technologies, Grandi Navi
Veloci, G4S, Glencore, Gruppo Hera, Gruppo Coin, Gruppo De Agostini, Gtech,
Gucci, H3G, Hupac, Infineon, Interoll, Il Sole24Ore, IREN, Istituto Poligrafico e
Zecca dello Stato, ITV, Kuwait Petroleum, La Perla, Labelux Group, Lamborghini,
Lavazza, Linde, LBBW, Levi’s, L’Oréal, Loro Piana, Luxottica, Jaguar Land Rover,
Lucite International, MAN, Magneti Marelli, Mapei, Marcegaglia, Mediaset,
Menarini, Messaggerie Libri, Miroglio, Mondelez International, Mossi & Ghisolfi,
Natuzzi, Novartis, Oerlikon Graziano, OSRAM, Piaggio, Perfetti, Pernod Ricard,
Philips, Pirelli, Porsche, ProSiebenSat1, Procter & Gamble, Prysmian, RAI, Rexam,
Rolex, Roche, Retonkil Initial, RWE, Saipem, Sandoz, SEA, Seat PG, Selex, Snam,
Sorgenia, Sky Italia, Schindler Electroca, Pfizer, RFI, Telecom Italia, Telecom Italia
Digital Solution, Telecom Italia Information Technology, Tenaris, Terna, Trenitalia,
Tyco, TuevSued, Telefonica, Unilever, Unicoop Firenze, Virgin Atlantic, Volks-
wagen, Vodafone and Wind.
For the Public Sector: Agenzia per l’Italia Digitale, Comune di Milano, Regione
Lombardia and Consip.
Acknowledgments xv
I would especially like to acknowledge all the people that have supported me
during this years with insights and suggestions. I learned so much from them, and
their ideas and competences have inspired my work: Silvio Fraternali, Paolo
Cederle, Massimo Milanta, Massimo Schiattarella, Diego Donisi, Marco Sesana,
Gianluca Pancaccini, Giovanni Damiani, Gianluigi Castelli, Salvatore Poloni, Milo
Gusmeroli, Pierangelo Rigamoti, Danilo Augugliaro, Nazzareno Gregori, Edoardo
Romeo, Elvio Sonnino, Pierangelo Mortara, Massimo Messina, Mario Collari,
Giuseppe Capponcelli, Massimo Castagnini, Pier Luigi Curcuruto, Giovanni Sor-
dello, Maurizio Montagnese, Umberto Angelucci, Giuseppe Dallona, Gilberto
Ceresa, Jesus Marin Rodriguez, Fabio Momola, Rafael Lopez Rueda, Eike Wahl,
Marco Cecchella, Maria-Louise Arscott, Antonella Ambriola, Andrea Rigoni,
Giovanni Rando Mazzarino, Silvio Sperzani, Samuele Sorato, Alberto Ripepi,
Alfredo Montalbano, Gloria Gazzano, Massimo Basso Ricci, Giuseppe De Iaco,
Riccardo Amidei, Davide Ferina, Massimo Ferriani, Roberto Burlo, Cristina
Bianchini, Dario Scagliotti, Ettore Corsi, Luciano Bartoli, Marco Ternelli, Stewart
Alexander, Luca Ghirardi, Francesca Gandini, Vincenzo Tortis, Agostino Ragosa,
Sandro Tucci, Vittorio Mondo, Andrea Agosti, Roberto Fonso, Federico Gentili,
Nino Lo Banco, Fabio Troiani, Federico Niero, Gianluca Zanutto, Mario Bocca,
Marco Zaccanti, Anna Pia Sassano, Fabrizio Lugli, Marco Bertazzoni, Vittorio
Boero, Carlo Achermann, Stefano Achermann, Jean-Claude Krieger, Reinhold
Grassl, François de Brabant, Maria Cristina Spagnoli, Alessandra Testa, Anna
Miseferi, Matteo Attrovio, Nikos Angelopoulos, Igor Bailo, Stefano Levi, Luciano
Romeo, Alfio Puglisi, Gennaro Della Valle, Massimo Paltrinieri, Pierantonio
Azzalini, Enzo Contento, Marco Fedi, Fiore Della Rosa, Dario Tizzanini, Carlo
Capalbo, Simone Battiferri, Vittorio Giusti, Piera Fasoli, Carlo di Lello, Gian
Enrico Paglia, George Sifnios, Francesco Varchetta, Gianfranco Casati, Fabio
Benasso, Alessandro Marin, Gianluca Guidotti, Fabrizio Virtuani, Luca Verducci,
Luca Falco, Francesco Pedrielli, Riccardo Riccobene, Roberto Scolastici, Paola
Formaneti, Andrea Mazzucato, Nicoletta Rocca, Mario Breuer, Mario Costantini,
Marco Lanza, Marco Poggi, Gianfranco Ardissono, Alex Eugenio Sala, Daniele
Bianchi, Giambattista Piacentini, Luigi Zanardi, Valerio Momoni, Daniele Panigati,
Maurizio Pescarini, Ermes Franchini, Francesco Mastrandrea, Federico Boni,
Mauro Minenna, Massimo Romagnoli, Nicola Grassi, Alessandro Capitani, Mauro
Frassetto, Bruno Cocchi, Marco Tempra, Martin Brannigan, Alessandro Guidotti,
Gianni Leone, Stefano Signani, Domenico Casalino, Fabrizio Lugoboni, Fabrizio
Rocchio, Mauro Bernareggi, Claudio Sorano, Paolo Crovetti, Alberto Ricchiari,
Alessandro Musumeci, Luana Barba, Pierluigi Berlucchi, Matthias Schlapp, Ugo
Salvi, Danilo Gismondi, Patrick Vandenberghe, Dario Ferri, Claudio Colombatto,
Frediano Lorenzin, Paolo Trincianti, Massimiliano Ciferri, Danilo Ughetto, Tiberio
Strati, Massimo Nichetti, Stefano Firenze, Vahe Ter Nikogosyan, Giorgio Voltolini,
Andrea Maraventano, Thomas Pfitzer, Guido Oppizzi, Alessandro Bruni, Marco
Franzi, Guido Albertini, Massimiliano De Gregorio, Vincenzo Russi, Franco Col-
lautti, Massimo Dall’Ora, Fabio De Ferrari, Mauro Ferrari, Domenico Solano, Pier
Paolo Tamma, Susanna Nardi, Massimo Amato, Alberto Grigoletto, Nunzio Calì,
Gianfilippo Pandolfini, Cristiano Cannarsa, Fabio Degli Esposti, Riccardo
xvi Acknowledgments
Scattaretico, Claudio Basso, Mauro Pianezzola, Marco Zanussi, Davide Carteri,

Giulio Tonin, Simonetta Iarlori, Marco Prampolini, Luca Terzaghi, Christian
Altomare, Pasquale Tedesco, Michela Quitadamo, Dario Castello, Fabio Boschiero,
Aldo Borrione, Paolo Beatini, Maurizio Pellicano, Ottavio Rigodanza, Gianni
Fasciotti, Lorenzo Pizzuti, Angelo D’Alessandro, Marcello Guerrini, Michela
Quitadamo, Dario Castello, Fabio Boschiero, Aldo Borrione, Paolo Beatini, Pier-
luigi De Marinis, Fabio Cestola, Roberto Mondonico, Alberto Alberini, Pierluca
Ferrari, Umberto Stefani, Elvira Fabrizio, Salvatore Impallomeni, Dario Pagani,
Marino Vignati, Giuseppe Rossini, Alfio Puglisi, Renzo Di Antonio, Maurizio
Galli, Filippo Vadda, Marco De Paoli, Paolo Cesa, Armando Gervasi, Luigi Di
Tria, Marco Gallibariggio, David Alfieri, Mirco Carriglio, Maurizio Castelletti,
Roberto Andreoli, Vincenzo Campana, Marco Ravasi, Mauro Viacava, Alessio
Pomasan, Salvatore Stefanelli, Roberto Scaramuzza, Marco Zaffaroni, Giuseppe
Langer, Francesco Bardelli, Daniele Rizzo, Silvia De Fina, Paulo Morais, Massi-
miliano Gerli, Andrea Facchini, Massimo Zara, Luca Paleari, Carlo Bozzoli, Luigi
Borrelli, Marco Iacomussi, Mario Dio, Giulio Mattietti, Alessandro Poerio, Fabrizio
Frustaci, Roberto Zaccaro, Maurizio Quattrociocchi, Gianluca Giovannetti, Pier-
angelo Colacicco, Silvio Sassatelli, Filippo Passerini, Mario Rech, Claudio Sordi,
Tomas Blazquez De La Cruz, Luca Spagnoli, Fabio Oggioni, Luca Severini,
Roberto Conte, Alessandro Tintori, Giovanni Ferretti, Alberta Gammicchia, Patri-
zia Tedesco, Antonio Rainò, Claudio Beveroni, Chiara Manzini, Francesco Del
Greco, Lorenzo Tanganelli, Ivano Bosisio, Alessandro Campanini, Giovanni Pie-
trobelli, Pietro Pacini, Vittorio Padovani, Luciano Dalla Riva, Paolo Pecchiari,
Francesco Donatelli, Massimo Palmieri, Alessandro Cucchi, Riccardo Pagnanelli,
Raffaella Mastrofilippo, Roberto Coretti, Alessandra Grendele, Davide Casagrande,
Lucia Gerini, Filippo Cecchi, Fabio De Maron, Alberto Peralta, Massimo Perni-
gotti, Massimo Rama, Francisco Souto, Oscar Grignolio, Mario Mella, Massimo
Rosso, Filippo Onorato, Stefan Caballo, Ennio Bernardi, Aldo Croci, Giuseppe
Genovesi, Maurizio Romanese, Daniele Pagani, Derek Barwise, Guido Vetere,
Christophe Pierron, Guenter Lutgen, Andreas Weinberger, Luca Martis, Stefano
Levi, Paola Benatti, Massimiliano Baga, Marco Campi, Laura Wegher, Riccardo
Sfondrini, Diego Pogliani, Gianluca Pepino, Simona Tonella, José González Osma,
Sandeep Sen, Thomas Steinich, Barbara Karuth-Zelle, Ralf Schneider, Rüdiger
Schmidt,Wolfgang Gärtner, Alfred Spill, Lissimahos Hatzidimoulas, Marco
Damiano Bosco, Mauro Di Pietro Paolo, Paolo Brusegan, Arnold Aschbauer,
Robert Wittgen, Peter Kempf, Michael Gorriz, Wilfried Reimann, Abel Archundia
Pineda, Jürgen Sturm, Stefan Gaus, Andreas Pfisterer, Peter Rampling, Elke
Knobloch, Andrea Weierich, Andreas Luber, Heinz Laber, Michael Hesse, Markus
Lohmann, Andreas König, Herby Marchetti, Rainer Janssen, Frank Rüdiger Poppe,
Marcell Assan, Klaus Straub, Robert Blackburn, Wiebe Van der Horst, Martin
Stahljans, Mattias Ulbrich, Matthias Schlapp, Jan Brecht, Enzo Contento, Michael
Pretz, Gerd Friedrich, Florian Forst, Robert Leindl, Wolfgang Keichel, Stephan
Fingerling, Sven Lorenz, Martin Hofmann, Nicolas Burdkhardt, Armin Pfoh, Kian
Mossanen, Anthony Roberts, John Knowles, Lisa Gibbard, John Hiskett, Richard
Wainwright, David Madigan, Matt Hopkins, Gill Lungley, Simon Jobson, Glyn
Acknowledgments xvii
Hughes, John Herd, Mark Smith, Jeremy Vincent, Guy Lammert, Steve Blackledge,
Mark Lichfield, Jacky Lamb, Simon McNamara, Kevin Hanley, Anthony Mead-
ows, Rod Hefford, Stephen Miller, Willem Eelman, Alessandro Ventura, David
Bulman, Neil Brown, Alistair Hadfield, Rod Carr and Neil Dyke.
I would especially like to gratefully acknowledge Gianluigi Viscusi at College of
Management of Technology (CDM)-École polytechnique fédérale de Lausanne
(EPFL), Alan Serrano-Rico at Brunel Univeristy, and Nadia Neytcheva Head of
Research at the Business Technology Outlook (BTO) Research Program who
provided me valuable suggestions and precious support in the coordination of the
production process of this book. Furthermore, I acknowledge the support of
Business Technology Foundation (Fondazione Business Technology) and all the
bright researchers at Business Technology Outlook (BTO) Research Program that
have supported me in carrying out interviews, surveys, and data analysis: Florenzo
Marra, Giulia Galimberti, Arianna Zago, Alessandro De Pace, Matteo Richiardi,
Ezechiele Capitanio, Giovanni Roberto, Alessandro Scannapieco, Massimo Bellini,
Tommaso Cenci, Giorgia Cattaneo, Andrada Comanac, Francesco Magro, Marco
Castelli, Martino Scanziani, Miguel Miranda, Alice Brocca, Antonio Attinà,
Giuseppe Vaccaro, Antonio De Falco, Matteo Pistoletti, Mariya Terzieva and
Daniele Durante.
A special acknowledgement goes to the memory of Prof. Antonino Intrieri who
provided precious comments and suggestions throughout the years.
Finally I acknowledge my family whose constant support and patience made this
book happen.
Vincenzo Morabito
Contents
Part I Strategy
1 Big Data and Analytics for Competitive Advantage . . . . . . . . . . . . 3

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Competitive Advantage Definition: Old and New Notions . . . . . 4
1.2.1 From Sustainable to Dynamic . . . . . . . . . . . . . . . . . . 5
1.2.2 From Company Effects to Network Success. . . . . . . . . 6
1.3 The Role of Big Data on Gaining Dynamic
Competitive Advantage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3.1 Big Data Driven Target Marketing . . . . . . . . . . . . . . . 6
1.3.2 Design-Driven Innovation . . . . . . . . . . . . . . . . . . . . . 8
1.3.3 Crowd Innovation. . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 Big Data Driven Business Models . . . . . . . . . . . . . . . . . . . . . 10
1.5 Organizational Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.5.1 Skill Set Shortages . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5.2 Cultural Barriers. . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5.3 Processes and Structures . . . . . . . . . . . . . . . . . . . . . . 13
1.5.4 Technology Maturity Levels . . . . . . . . . . . . . . . . . . . 13
1.5.5 Organizational Advantages and Opportunities . . . . . . . 13
1.6 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.7 Recommendations for Organizations . . . . . . . . . . . . . . . . . . . . 17
1.7.1 Ask the Right Questions . . . . . . . . . . . . . . . . . . . . . . 17
1.7.2 Look Out for Complementary Game
Changing Innovations . . . . . . . . . . . . . . . . . . . . . . . . 18
1.7.3 Develop Sound Scenarios . . . . . . . . . . . . . . . . . . . . . 18
1.7.4 Prepare Your Culture . . . . . . . . . . . . . . . . . . . . . . . . 18
1.7.5 Prepare to Change Processes and Structure . . . . . . . . . 19
1.8 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
xix
xx Contents
2 Big Data and Analytics for Government Innovation. . . . . ....... 23

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 23
2.1.1 New Notions of Public Service:
Towards a Prosumer Era? . . . . . . . . . . . . . . . . . . . . . 24
2.1.2 Online Direct Democracy . . . . . . . . . . . . . . . . . . . . . 25
2.1.3 Megacities’ Global Competition . . . . . . . . . . . . . . . . . 25
2.2 Public Service Advantages and Opportunities. . . . . . . . . . . . . . 26
2.2.1 New Sources of Information: Crowdsourcing . . . . . . . . 26
2.2.2 New Sources of Information:
Internet of Things (IoTs) . . . . . . . . . . . . . . . . . . . . . . 27
2.2.3 Public Talent in Use . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2.4 Private–Public Partnerships . . . . . . . . . . . . . . . . . . . . 31
2.2.5 Government Cloud Data . . . . . . . . . . . . . . . . . . . . . . 31
2.2.6 Value for Money in Public Service Delivery . . . . . . . . 32
2.3 Governmental Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3.1 Data Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3.2 Data Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3.3 Privacy, Civil Liberties and Equality. . . . . . . . . . . . . . 34
2.3.4 Talent Recruitment Issues . . . . . . . . . . . . . . . . . . . . . 35
2.4 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.5.1 Smart City Readiness . . . . . . . . . . . . . . . . . . . . . . . . 39
2.5.2 Learn to Collaborate . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.5.3 Civic Education and Online Democracy . . . . . . . . . . . 41
2.5.4 Legal Framework Development . . . . . . . . . . . . . . . . . 41
2.6 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3 Big Data and Education: Massive Digital Education Systems . . . . . 47

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.1.1 From Institutionalized Education to MOOCs . . . . . . . . 49
3.2 MOOC Educational Model Clusters . . . . . . . . . . . . . . . . . . . . 51
3.2.1 University-Led MOOCs . . . . . . . . . . . . . . . . . . . . . . 51
3.2.2 Peer-to-Peer MOOCs . . . . . . . . . . . . . . . . . . . . . . . . 52
3.3 The Role of Big Data and Analytics . . . . . . . . . . . . . . . . . . . . 54
3.4 Institutional Advantages and Opportunities from MOOCs . . . . . 55
3.5 Institutional Challenges from MOOCs. . . . . . . . . . . . . . . . . . . 57
3.6 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.7 Recommendations for Institutions . . . . . . . . . . . . . . . . . . . . . . 62
3.8 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Contents xxi
4 Big Data Driven Business Models . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.2 Implications of Big Data for Customer Segmentation . . . . . . . . 69
4.3 Implications of Big Data as a Value Proposition . . . . . . . . . . . 69
4.4 Implications of Big Data for Channels . . . . . . . . . . . . . . . . . . 70
4.5 The Impact of Big Data on Customer Relationships . . . . . . . . . 71
4.6 The Impact of Big Data on Revenue Stream . . . . . . . . . . . . . . 72
4.7 The Impact of Big Data on Key Resources
and Key Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.8 The Impact of Big Data on Key Partnerships . . . . . . . . . . . . . . 74
4.9 The Impact of Big Data on Cost Structures . . . . . . . . . . . . . . . 75
4.10 Organizational Advantages and Opportunities . . . . . . . . . . . . . 76
4.11 Organizational Challenges and Threats . . . . . . . . . . . . . . . . . . 77
4.11.1 Creativity and Innovation Capability Deficit . . . . . . . . 77
4.11.2 Interrogating Big Data . . . . . . . . . . . . . . . . . . . . . . . 77
4.11.3 Plug and Play Architectures. . . . . . . . . . . . . . . . . . . . 78
4.12 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Part II Organization
5 Big Data Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

5.1 Introduction to Big Data Governance . . . . . . . . . . . . . . . . . . . 83
5.1.1 Big Data Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.1.2 Information Governance Disciplines . . . . . . . . . . . . . . 87
5.1.3 Industries and Functions . . . . . . . . . . . . . . . . . . . . . . 90
5.2 Big Data Maturity Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.2.1 TDWI Maturity Model . . . . . . . . . . . . . . . . . . . . . . . 91
5.2.2 Analytics Business Maturity Model . . . . . . . . . . . . . . 93
5.2.3 DataFlux Data Governance Maturity Model . . . . . . . . . 94
5.2.4 Gartner Maturity Model . . . . . . . . . . . . . . . . . . . . . . 95
5.2.5 IBM Data Governance Maturity Model . . . . . . . . . . . . 96
5.3 Organizational Challenges Inherent with Governing
Big Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.4 Organizational Benefits of Governing Big Data . . . . . . . . . . . . 99
5.5 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.7 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6 Big Data and Digital Business Evaluation . . . . . . . . . . . . . . . . . . . 105

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.2 Digital Business Evaluation Using Big Data . . . . . . . . . . . . . . 106
xxii Contents
6.3 Organizational Advantages and Opportunities . . . . . . . . . . . . . 108

6.3.1 Customer Value Proposition . . . . . . . . . . . . . . . . . . . 109
6.3.2 Customer Segmentation. . . . . . . . . . . . . . . . . . . . . . . 110
6.3.3 Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.3.4 Customer Relationship . . . . . . . . . . . . . . . . . . . . . . . 111
6.4.1 Key Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.4.2 Privacy and Security . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.4.3 Cost Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.5 Cases Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.6.1 Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.6.2 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
6.7 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7 Managing Change for Big Data Driven Innovation . . . . . . . . . . . . 125

7.1 Introduction: Big Data—The Innovation Driver . . . . . . . . . . . . 125
7.2 Big Data—The Key Innovative Techniques . . . . . . . . . . . . . . . 126
7.2.1 Integration of Data Platforms . . . . . . . . . . . . . . . . . . . 127
7.2.2 Testing Through Experimentation . . . . . . . . . . . . . . . . 128
7.2.3 Real-Time Customization . . . . . . . . . . . . . . . . . . . . . 128
7.2.4 Generating Data-Driven Models . . . . . . . . . . . . . . . . . 128
7.2.5 Algorithmic and Automated-Controlled Analysis . . . . . 129
7.3 Big Data: Influence on C-Level Innovative Decision Process . . . 129
7.3.1 Stimulating Competitive Edge . . . . . . . . . . . . . . . . . . 130
7.3.2 Predictive Analytics: Data Used to Drive Innovation. . . 130
7.4 The Impact of Big Data on Organizational Change. . . . . . . . . . 132
7.4.1 An Incentivized Approach . . . . . . . . . . . . . . . . . . . . . 133
7.4.2 Creating a Centralized Organizational ‘Home’ . . . . . . . 133
7.4.3 Implementing the Changes—First Steps . . . . . . . . . . . 135
7.5 Methodologies for Big Data Innovation. . . . . . . . . . . . . . . . . . 135
7.5.1 Extending Products to Generate Data . . . . . . . . . . . . . 135
7.5.2 Digitizing Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.5.3 Trading Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
7.5.4 Forming a Distinctive Service Capability. . . . . . . . . . . 136
7.6 New Big Data Tools to Drive Innovation . . . . . . . . . . . . . . . . 137
7.6.1 The Hadoop Platform . . . . . . . . . . . . . . . . . . . . . . . . 137
7.6.2 1010DATA Cloud Analytics . . . . . . . . . . . . . . . . . . . 137
7.6.3 Actian Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
7.6.4 Cloudera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
7.7 Models of Big Data Change . . . . . . . . . . . . . . . . . . . . . . . . . 139
7.7.1 Big Data Business Model . . . . . . . . . . . . . . . . . . . . . 139
7.7.2 The Maturity Phases of Big Data Business Model . . . . 139
7.7.3 Examples of the Business Metamorphosis Phase . . . . . 142
Contents xxiii
7.8 Big Data Change Key Issues . . . . . . . . . . . . . . . . . . . . . . . . . 143

7.8.1 Storage Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.8.2 Management Issues. . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.8.3 Processing and Analytics Issues . . . . . . . . . . . . . . . . . 144
7.9.1 Data Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.9.2 Information Extraction . . . . . . . . . . . . . . . . . . . . . . . 146
7.9.3 Data Integration, Aggregation, and Representation . . . . 146
7.10 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
7.11 Recommendation for Business Organizations . . . . . . . . . . . . . . 149
7.12 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Part III Innovation Practices
8 Big Data and Analytics Innovation Practices . . . . . . . . . . . . . . . . . 157

8.1 Introduction . . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 157
8.2 Sociometric Solution. . . . . . . ........ . . . . . . . . . . . . . . . . 158
8.2.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 158
8.2.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 159
8.3 Invenio . . . . . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 160
8.3.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 160
8.3.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 161
8.4 Evolv . . . . . . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 161
8.4.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 162
8.4.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 163
8.5 Essentia Analytics . . . . . . . . ........ . . . . . . . . . . . . . . . . 163
8.5.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 164
8.5.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 164
8.6 Ayasdi Core . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 165
8.6.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 165
8.6.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 166
8.7 Cogito Dialog . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 167
8.7.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 167
8.7.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 168
8.8 Tracx . . . . . . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 168
8.8.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 169
8.8.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 169
8.9 Kahuna . . . . . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . 170
8.9.1 Developer . . . . . . . . ........ . . . . . . . . . . . . . . . . 170
8.9.2 Applications . . . . . . ........ . . . . . . . . . . . . . . . . 171
xxiv Contents
8.10 RetailNext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

8.10.1 Developer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
8.10.2 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8.11 Evrythng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8.11.1 Developer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8.11.2 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
8.12 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
9.1 Building the Big Data Intelligence Agenda . . . . . . . . . . . . . . . 177
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Acronyms
ACID Atomicity, Consistency, Isolation, and Durability

AI Artificial Intelligence
API Application Programming Interface
B2B Business to business
B2G Business to government
BI Business Intelligence
BM Business Model
BMI Business Model Innovation
BS Bachelor of Science
CD Compact disc
CEO Chief Executive Officer
CIO Chief Information Officer
CMO Chief Marketing Officer
CRM Customer Relationship Management
CSFs Critical Success Factors
CTO Chief Technology Officer
CxO C-level Manager
DDS Digital data stream
DG Data Governance
ERP Enterprise Resource Planning
EU The European Union
GPS Global Positioning System
HR Human Resources
ICT Information and Communication Technology
IoTs Internet of Things
IP Intellectual Property
IP address Internet Protocol address
IPO Initial public offering
IT Information technology
KPIs Key performance indicators
MIS Management Information Systems
xxv
xxvi Acronyms
MOOCs Massive open online courses

MS Master of Science
NoSQL Not Only SQL
OER Open educational resources
OLAP Online analytical processing
P2P Peer 2 Peer
PC Personal computer
QR code Quick Response Code
R&D Research and Development
RFID Radio-frequency identification
ROI Return on investment
SMEs Small and medium enterprises
SQL Structured Query Language
UK The United Kingdom
UN The United Nations
US The United States of America
VOIP Voice over Internet Protocol
Part I
Strategy
Big Data and Analytics for Competitive
Advantage 1
Abstract
The role of this chapter is to introduce the reader to the utilization of big data for
achieving competitive advantage. It begins by clarifying current notions of
competitive advantage in strategic literature and highlights the current organi-
zational challenges in taking advantage of the big data trend, as well as the
possible advantages and opportunities. Finally, a case study discussion provides
insights from practice and highlights points of attention, for those pursuing big
data-driven competitive advantage.
1.1 Introduction
The concept of competitive advantage has created kaleidoscope of perspectives

about its sources and mechanisms of generating it and destructing it, about what
influences it and how can companies plan for it. Yet, the very theories of com-
petitive advantage have been products of their times, explaining rather than pre-
dicting what made companies successful during the era of their development.
Understanding the strategic role of IT has been even more challenging, as strategic
literature ignored its importance as strategic asset until recently. Also, Management
Information Systems (MIS) literature on the organizational value of IT has been
confused amidst multitude issues of analysis, methodology, and measurement or
simply distinguishing between causal relationships and correlation (Drnevich and
Croson 2013).
However, e-commerce has changed our perception about the strategic impor-
tance of IT for the going concern of organizations and with the advent of big data,
another era of strategic game playing is likely, as rules of competition may change
yet again and so will our understanding of competitive advantage. Concurrent
social changes, for example, new ways of funding and valuing organizations as well
as virtual money such as bitcoin, may even change our understanding of the link of
competitive advantage to monetization.
© Springer International Publishing Switzerland 2015 3

V. Morabito, Big Data and Analytics,
DOI 10.1007/978-3-319-10665-6_1
4 1 Big Data and Analytics for Competitive Advantage
Before we discuss, however, the potential implication of big data on business

strategy and competitive advantage, let’s briefly review the key strategic perspec-
tives thus far.
1.2 Competitive Advantage Definition: Old and New

Notions
As popularized by Michael Porter’s back in the 80s, competitive advantage denotes

a company’s profit making superiority over its competitors (Porter 1985). Such
superiority was determined by a company’s position in its sector and its ability to
defend its position against challenges from competitors, new entrants, suppliers and
even changes in customers’ preference. In response, companies’ strategies aimed at
controlling every aspect of business activity and even plan to ensure that the
industry was safeguarded from new entrants. Long-term or exclusive contracts with
suppliers, covert price fixing, heavy advertising to orient consumer preferences,
heavy capital investments were few of the moves open to companies in the com-
petition chest board (Drnevich and Croson 2013). For example, the fierce compe-
tition between car manufacturers in the 90s leading to the concentration of the
sector into few large car manufacturing groups is characteristic of that era
(KPMG 2010).
Others suggest that competitive advantage can be obtained by efficient man-
agement of resources, including ones’ employees, or by reducing transaction costs
within the value chain. Anyway, successful governance required its own IT tools.
All types of Enterprise Resource Planning (ERP) systems are prime examples of IT
investments oriented to help managers at all levels attain resource efficiency; while
IT-assisted, vertical business models that ‘cut out the middleman’ was a great
example of reducing transaction costs. E-commerce phenomenal success over brick
and mortar organizations was based on this simple principle, enabling people to
shop for the best price product from the convenience of home at zero extra cost.
For those advocating competence as the source a competitive advantage, com-
panies should decide and focus on developing their key capabilities, a set of
valuable, rare, imperfectly imitable, non-substitutable resources, which could pro-
pel a business to such an advantageous position (Wu 2013). This set comprised
anything from exclusive or discriminatory access to capital resources (raw materials
or even expertise) to business process superiority (such as procurement, innovation
or marketing) to combinations of the above (KPMG 2010; Taylor and LaBarre
2008). This required companies to predict customers’ behaviors in the long-term
and commit in the development and retention of such key capabilities. Process
efficiency and knowledge management were considered particularly important from
this perspective as it was the combination of resources in particular ways that gave
companies an advantage over its competitors, and that requires know-how. The role
of IT from this perspective is to assist people to take more effective decisions and
do so efficiently.
1.2 Competitive Advantage Definition: Old and New Notions 5
With the advent of e-commerce and the rise of the “maverick” entrepreneur,
business ethos towards control vis à vis openness changed and flexibility as the
source of competitive advantage came to the forefront (Taylor and LaBarre 2008).
Yet, it is rooted in Schumpeter’s theories on creative destruction, where the
industries evolve as new companies come up with new ways of doing things, thus,
pushing old ones who cannot adapt out of the market (Schumpeter 1934). Hence,
for a company to survive or to success should build in the capability to renew itself.
From this perspective, companies with great operational flexibility that can quickly
respond to challenges coming from shifting customer preferences, new entrants’
changes of industry norms, and unpredicted competitors’ moves will survive. Even
better, innovative companies that can challenge the status quo and gain the profits
of first mover’s advantage and move on as market’s mature or cut down their loses
if ventures don’t prove successful.
Thus, the strategic game changed from a slow careful positioning on the com-
petition chessboard, to a fast action game, where speed is a crucial factor. Big data
driven business models will exacerbate this trend. Already, real-time, location-
based offers—the likes of Groupon—have proliferated and are now part of most
inner city dwellers’ life (Raice and Woo 2011).
Big data however poses technical challenges of storage and processing, and
while the load of necessary inflexible capital investments is now over, matters of
repurposing, sharing or turning infrastructure into a shared public good arise. The
recent trend of Shared services and Shared Clouds are typical examples of infra-
structure sharing and perhaps the same principle will be the norm for big data
storehouses, e.g., by Pharmaceuticals, where companies, academics and public
sector will openly utilize for common good (Schultz 2013).
Before we discuss the influence of big data on companies’ efforts to achieve
competitive advantage, we will briefly summarize the trends of our conceptions of
competitive advantage.
1.2.1 From Sustainable to Dynamic
Globalization and e-commerce are radically changing consumer preferences and a

flood of young Internet entrepreneurs that is driving discontinuous change business
environment has emerged. This change is becoming more ambiguous and difficult
to predict and plan for. For example Facebook has transformed the online retail
market as Skype has transformed telecommunications. Sure enough a key shift in
recent strategy theories is that competitive advantage is neither something a com-
pany owns nor something that it can safeguard. Emphasis has moved away from
controlling in strategic orientation (Schumpeter 1934) to position game-changer
innovation as key to business success.
1.2.2 From Company Effects to Network Success
Such unpredictably changing market conditions were defined by Williamson as

turbulent environments (O’Brien 1976), demand networking strategies to navigate
them successfully (Raice and Woo 2011; Schultz 2013). While collaboration
always had a position in business, the nature of it is ever-changing (Morabito 2014).
Conglomeration, lobbying, contractual and gentlemen agreements amongst large
group of companies dominated business life in the past; nowadays, a shift towards
open networks of suppliers, employees, customers has blurred the boundaries of
companies, shared value amongst networks or produced public goods. Interde-
pendence between companies and the public has intensified, as customers, suppliers
or simple product evangelists nurture and propagate company information and
referential credibility. For example, open source software has spread its paradigm
becoming the norm for certain domains: nowadays people advertise product or
services for free to their friends by clicking the share button, or recommend a friend
for a job on Linkedln at your spare time, and all this creates business value for free.
It, of course, creates a lot of social-media-derived big data about our social
behavior that companies are eager to understand the workings of, whether to nur-
ture and reward us or to control and manipulate us, will depend on their strategic
perspective (Morabito 2014).
1.3 The Role of Big Data on Gaining Dynamic

Competitive Advantage
The overarching disruptive power of big data demands that organizations engage
with it at a strategic level. However, how organizations will utilize this technology
trend with their existing business model will depend on their orientation. While,
consultants at A.T. Kearney argue that big data would have positive effects across
strategy and operations (Hagen et al. 2013). In summary, the advantages of utilizing
big data to obtain competitive advantage has been discussed into relation to big data
driven target marketing, design-driven innovation, and crowd innovation, all of
which will be discussed in detail in the following sections.
1.3.1 Big Data Driven Target Marketing
Big data can change the way companies identify and relate to their customer base.
Undoubtedly, companies can boost the old marketing strategies using new big data
tools and expertise. Market penetration strategies can leverage big data to feed
marketers information on how to keep existing customers and improve repetitive
sales. Likewise, new customer engagement techniques, like gamification, promise
improved loyalty levels (Paharia 2013).
1.3 The Role of Big Data on Gaining Dynamic Competitive Advantage 7
Cross-selling, for example, leverages a company’s affinity and knowledge of its

market to sell different products to the same people. Banks, for example, have
started not only to analyze vast amounts of their clients’ transactions in relation to
social media to understand their customer preferences, but also to create new
offerings to their clients. Citi, for example aggregates data from its global customer
base, to help corporate clients identify market trends (Lesk 2013).
Identifying, however, new market niches must be the real power of big data and
its real challenge. Companies no longer need to approach the market in large
demographic chunks. They can instead use emerging analytics to identify new
niches or even subdivide existing target markets into smaller more coherent groups
to unlock their potential (Adamson et al. 2012). Targeting becomes a matter of
aggregating multiple small niches. Combined with advancements in automated
marketing communications, we are heading towards the era of mass customization.
Thus, the ultimate big data promise for marketers is that of mass customization.
big data collection, synthesis and analysis promises to provide businesses with real
insight about its customers’ behaviors based on actual past and real-time, step-by-
step behaviors. Much like in virtual ethnography (Hine 2000), that seeks to
understand online communities through participation, social media data is collected
covertly. Because most of us post spontaneous opinions and views in social media,
information is free from response biases often encountered in surveys or under
focus group conditions. This gets rid of the cost of ineffective market research,
which can then mislead sales efforts, marketing plans and company strategies.
Sentiment analysis on what we post about areas of our lives on social media unveils
our attitudes and can lead to discovery about new product and service requirements
(Morabito 2014; LaValle et al. 2011). Fine grained behavioral analysis aspires to
feed predictive analytics, enabling marketers to spot deviations in our purchasing
patterns. An initial issue was corresponding social media views and opinions to
demographics, so one can understand who says what to profile information and
improve targeting. The ability to correlate social media data, for example tweets
with Facebook and LinkedIn profiles, has mitigated our doubts about the source of
information.
Thus, big data gives Marketing Relationship Managers an excellent tool to feed
us the right information, to influence us at the right moment towards making that
final decision to buy. There is now a whole sector of social media analytics dedi-
cated at eliciting these insights. Mashable.com alone offers 20 Application Pro-
gramming Interface (API) that can help people scan different channels, like
Facebook and Twitter or even texts in newspaper articles and blogs, for what people
want. In the same way, they can also get real-time feedback about how promotions
and other promotional activities are received (Provost and Fawcett 2013).
But data has also changed the way we can target people. We don’t need to group
people into large or even small target groups, we can target people directly based on
the electronic trails of our computer’s IP address about our lifestyle choices, from
what we buy to what we vote for to what we are interested in, where we are located,
what our demographic characteristics are.
While social media, however, dominate current discussions about the potential
of big data to provide companies with a competitive advantage, it is likely that
really differentiated business models will take advantage of design-driven innova-
tion relating, for example, to the Internet of Things (IoT), see also (Morabito 2014).
1.3.2 Design-Driven Innovation
The combination of big data sources with other emerging technologies can inspire
design-driven innovations. These innovations are disruptive game changers that
manage innovations that customers do not expect but they eventually love. Design,
in its etymological essence, means “marking sense of things” and design-driven
innovations are the R&D process for meanings (Verganti 2009).
For example Apple did not change how we make calls from our mobile, but
what we do with our mobile and how we think of it. For everything you want to do,
there is now an app for that, from designing color schemes for your baby’s room to
passing time playing angry birds to checking the news to watching a film to
measuring the dimensions of your rooms. An iPhone is not a phone anymore, it is a
multiple purpose tool (Verganti 2009) and Apple is not a phone-making company,
it’s the company that has changed our lives and most people love it for it. Products
are not seen as outputs of some faceless industrial process anymore, they are
symbols of the ethos and caliber of the people who designed it. Buying a product is
also a representation of who you are and who you favor. Hence, product innovation
is not just about products, it is the strategy of sharing common meanings with your
customers and being part of the community.
The same way product innovation is not just about products, but also about
sharing common meanings, business model design is also about sharing meanings
about what an organization stands for. For example, Asos.com is a fashion retailer
though is not just about cloths (Asos 2012). The company has invested in a mar-
ketplace site which is not just about fashion it is about fashion Democracy, enabling
anyone, anywhere in the world to sell fashion, to anyone, anywhere in the world
and for a 10 % commission per sale, it is a self-sustained business model as well
(Asos 2012).
What will constitute big data Design driven Innovations then? To date, “Big
data = Social Data” in most people’s minds, yet really transformative innovations are
likely to be inspired by the Internet of Things (IoT). Intelligent systems equipped
with sensors and decision support systems promise autonomous, rather than auto-
mated, innovations. Such intelligent systems change our paradigm, the very core
assumptions about what is possible, what is right and what is wrong. They promise
an “always-on, always-aware, always-connected, always-controllable” (Paharia
2013) machine-to-machine coordinated world. This will affect almost every aspect
of infrastructure as we know it. Such technologies will turn mundane everyday
things into novel offerings. Commuting, for example, may change dramatically over
the next few decades. Your future car may be more of a driverless taxi you pay for on
demand. You will be able to call it to pick you up and drive you to work as
1.3 The Role of Big Data on Gaining Dynamic Competitive Advantage 9
technology allows cars to communicate with other cars and the road infrastructure in
a safe way, and, of course, self-regulate their own green energy consumption and
storage as they will most like run by solar or hydrogen power (Neiger 2014; Griggs
2014). And perhaps, you won’t even have to own it! You will be able to hire it from
city stations, from companies such as, e.g., www.Car2go.com.
Since 2008 more things are connected to the internet than people, making it a
huge business opportunity. According to the UK government’s Department for
Business, Innovation and Skills (BIS) the global market for smart city solutions will
be at more than $400 billion annually by 2020. This may sound huge to some but it
is still a fraction of global infrastructure spending (Townsend 2013). Though, an
increasing movement of civic hackers, open-source technologies and open
government data are still working together in order to demonstrate the value of
smart technology to make cities more efficient, democratic, safer and sociable
(Townsend 2013). And while this may be a challenge for organizations who strive
for profit, it is extremely valuable to societies and possibly to social entrepreneurs
who strive for social impact and social change, rather than money.
Perhaps ideas for innovations won’t even come from within organizations and
institutions; perhaps we have entered an era of peer-to-peer innovation where ideas
and even solutions are crowdsourced and crowdfunded. The following section
describes how the locus of innovation has shifted over the years from an internal
process to becoming the creative engagement of communities of users, and how
thus its ownership and directions shifts from the organization to the community
(Morabito 2014).
1.3.3 Crowd Innovation
Big data can not only change how we approach the market with a product or
service, but also how we design the product to start with. Open innovation was
based on the premise that innovation ideas that can be useful for organizations may
lie outside the organization and companies should not restrict themselves from
harnessing these ideas for the sake of control (Chesbrough 2003). This perspective
suggests some very different principles about how a successful organization should
behave. For example, it abolishes the “non-invented here” notion to recognize talent
with useful ideas wherever these may come from (universities, suppliers, customers,
other companies, the public). Intellectual Property (IP) is a trading asset to be
bought and sold for profit. And IP can be a matter of co-creating with outsiders for
mutual benefit (Chesbrough 2003). Big data can take this conception into a whole
new level. Seen as product requirements, social media can be scanned for customer
complaints and product related wish lists. But it is not only that we can get better
insight into the market, we can also respond rapidly.
Open innovation is now facilitated by innovation intermediaries, like Innocen-
tive platform which match makes ‘solvers’ and companies with a problem seeking a
solution. Big companies can take advantage of these developments to outsource
such expertise, and they do. For example AstraZeneca, has set up an innovation
pavilion that hosts their challenges on Innocentive. As part of this, it has set up a
$100,000 innovation fund to source a solution for a Targeted Delivery of Oligo-
nucleotides that will improve their therapeutic effectiveness on tumors cells
(Innocentive 2014). But such outsourced expertise is available to smaller companies
too. While this is a good thing, outsourcing expertise is a great leveler between
large and small companies, and big data has given small organizations a leg up.
In the big data era, not only data and opinions are open, but so are ideas, even
business ideas! Innovation hubs pop up across the globe to provide support to
people with ideas to incubate new businesses offering, mentoring, and avenues to
funding. Crowdfunding sites have broadened the funding avenues and the funding
base even more, by enabling consumers to support these creative business ideas
directly. For example Kickstarter.com is a community of people working together
that is a crowdfunding platform to enables people to donate, pre-order or get a stake
in a company of their liking (Kickstarter 2014).
Anything from Art and Comic design to Food and Technology business ideas
are included. Lix, for example, a pen-like 3D printer idea has pledged for £30,000
only to collect £485,249 from 5,388 backers in 26 days, most of them early
adopters who pre-ordered the pen (LIX 2014).
Social media and big data feed off each other. Identified ideas, prototypes,
products and scenarios are discussed, developed and constantly updated in
collaboration within communities, and tested using against historical and real-time
data to predict market reactions (Choi and Varian 2012; Hafkesbrink and Schroll
2011). Using predictive analytics, for example innovators can get insights about
best case scenarios and comparisons of different alternatives (Kearney 2014).
1.4 Big Data Driven Business Models
The emergence of mobile phones is a potentially lucrative media platform for

marketers. Mobile devices have enabled context-specific, real-time marketing
communications for new types of middlemen, like daily deal sites, and take
advantage of these technological advancements. Groupon location-based services,
are an example of just that, aspiring to becoming the ultimate virtual marketplace,
where local people can find local deals on anything, anytime, anywhere (Styles
2014). They monitor the location of millions of their subscribers across the globe to
match them with local deals in their area based on their interests. With the pro-
liferation of such sites, deal aggregator sites, like Yipit offer one-stop shop of all
daily deal sites to customers (Raice and Woo 2011).
However, we have not yet become really creative with big data. So far, the
aspirations of most people are driven by old conceptions of adding value, still
seeking to use big data to do old things perhaps more efficiently and more effec-
tively, and certainly more flexibly, yet the same. With big data changing the rules
favoring those with technical and analytical skills, it is inevitable that technology
companies will diversify, entering traditional sectors whether on their own or by
1.4 Big Data Driven Business Models 11
acquiring smaller and aspiring companies in the market. Data storage and analysis
capability give a competitive advantage to companies who want to make an inroad
into other sectors.
In the sphere of retailing for example, online advertising and cross-selling seems
to be the key drivers for utilizing big data. It will not be surprising however if game
changing business models in the retail sector are driven by big technology com-
panies. IBM on April 2014 announced the acquisition of Fluid to develop a virtual
personal shopper mobile app, based on its intuitive Watson technology that can
interact with people in natural language (Dignan 2014). It does not take a big stretch
of imagination to understand the transformative potential of combining Artificial
Intelligence (AI) with big data analytics.
Furthermore, Google has made inroads into the travel market, ironically enough
partly funded by travel agencies themselves through online advertising expenditure.
Nearly 70 % of travel bookings are done online. Some 70–90 % of advertising by
online Tour Agencies is spent on Google. Google know how the competition
performs, owns the channel of reaching customers, employs the right caliber ana-
lysts, has the culture to keep them, immense data storage capability and a loyal
audience. With a capabilities storehouse like this, Google can make inroads into
every retail market it pleases (Brumley 2014).
Perhaps more impressive and overarching will be smart city innovations relating
to the “Internet of Things” (IoT). Data can be generated by sensors integrated into
anything we know, from garbage bins and bike wheels to water pipes and traffic
lights, combined with Artificial Intelligence (AI) technologies can form a network
of self-managing city infrastructure. No wonder IBM, Microsoft and Cisco are all
into the race of winning smart city pilot projects (Ratti 2014). It is the role of big
data in designing these disruptive innovations that will transform the way we live.
Do big companies have an advantage in this competitive space? Perhaps or
perhaps not. A competing paradigm is emerging that is more in line with ‘open’
conceptions of the prosumer, i.e. a customer that produce their own products or
services. And this is the conception of Service Mashups–compositions of service
modules put together by consumers themselves. Hence, the role of companies is to
put together the service modules in a way that can be easily combined with other
modules to form a service. Already businesses are working towards developing web
and cloud based interactivity environments where IoT services can be put together
(Im et al. 2013; Guinard and Trifa 2009).
1.5 Organizational Challenges
While businesses across industries recognize the imperative of big data, there are
many challenges that face the research and evolution in this field. The most
prevalent are skill set shortages, cultural barriers, processes and structures, and
technology maturity levels. These issues are discussed in what follows.
1.5.1 Skill Set Shortages
One commonly referred problem with respect to big data is having the right people.
Data scientists are often PhD graduates who combine mathematical and program-
ming skills, able to interrogate the databases to uncover trends and build predictive
models to check different scenarios (Kearney 2014; Davenport and Patil 2012).
Unlike large online retailers, such as Amazon, technology companies, such as
Google and Financial institutions, most organizations had not invested in such
expertise. While Data scientists are a rare commodity and very likely to be highly
pricey, assess to it may not be impossible, as outsourcing companies are popping
up. For example, Exerfy.com is a Harvard spin off, that specializes in resourcing
Data Scientists for organizations or undertake Analytics projects on their behalf
(Harvard Innovation Lab 2014). Perhaps, even the pool of expertise is not that
small. Big data opens a new employment avenue for mathematicians and data
modeling people working for financial institutions and perhaps students with strong
mathematical skills are likely to be drawn to this new field. Perhaps even open
source coders will create ever so user-friendly, predictive analytics freeware that
can be interrogated by none experts, in natural language. After all, some freeware
tools already exist for those who are statistically inclined.
1.5.2 Cultural Barriers
While market-driven innovations require insight, design-driven innovations call for

foresight! Strategic and Technology foresight go hand-in-hand. For company
boards who long viewed IT as a support function, the struggle to change their
attitude, attract and keep the right talent will be even harder. Spencer Stuart’s 2011
Index shows that the average age of the Standard and Poor’s 500 board members is
62.4 years (Bricker and Eckler 2014). C-level management in the US may have a
hard time shedding old assumptions and wholeheartedly fostering new ways of
conducting businesses.
In addition, in a world where competitive advantage is created and depleted
quickly, capability may be a matter of sourcing, investments, and alliances rather
internal evolution and development; thus, patterns of capability development may
be common in start-ups as well as technology giants (Fine et al. 2002).
Organizations that have learned to collaborate, they will now be more able to
compete. Innovation 3.0 (Hafkesbrink and Schroll 2011), for example, is based on
collaborative practices, where it is not companies but communities of interest and
communities of practice, with wide participation for suppliers, customers, even
competitors who seek a benefit in creating a common purpose. Still many orga-
nizations are struggling with internal animosities, silo mentalities and professional
territories. Big data has become a new reason to fight about, as marketing compete
against IT departments for big data ownership, and of course for the investment
budgets that go with it (Gardner 2014).
1.5 Organizational Challenges 13
1.5.3 Processes and Structures
For big data driven strategies to succeed they need to be implemented, and they
need to be implemented rapidly. This fundamentally means changing processes and
possible structures and architecture, with knock on effects on business and technical
architecture.
Particular after years of pursuing process standardization to affect operational
efficiencies, people and departments are now stuck in their ways. Job descriptions
are tight down to particular roles and so are reward systems and remuneration.
Departmental and directorial kudos depends on the budget allocate to them. Big
data driven innovations demand enterprise-wide collaboration, flexibility and
knowledge sharing. Inevitably this will require organizations to take a modular
approach to their structure, to enable them to mix and match accordingly. It will
also mean that company policies, remuneration and leadership orientation will have
to change or perish.
1.5.4 Technology Maturity Levels
To enter a field still in development, particularly one that depends a lot on emerging
technologies, one need to plan for the technology obsolescence and technological
skills renewal. The big data sphere both in terms of data collection, quality controls
and analysis is still in development and inevitably investments in any platform run
an inherent risk of becoming outdated or disrupted by new technologies. For
example, innovations on database infrastructure, such as orthogonal frequency-
division multiplexing (OFDM) (Werbach and Mehta 2014), can facilitate real-time
analytics, in ways that is not currently possible. Advances in in-memory computing
capability, such as Non-volatile memory (NVM) devices (Lankhorst et al. 2005),
can address energy consumption concerns and analytics speed (Chen et al. 2014).
Adoption of ubiquitous applications will transform not only what we can do with
technology but our attitude towards it, with implications on company, public policy,
and the organization of social life (Lesk 2013; Adamson et al. 2012).
1.5.5 Organizational Advantages and Opportunities
Much like with most disruptive innovations, embarking on big data utilization
projects will accrue a number of organizational benefits. Those benefits are
discussed below:
a. Improve decision making by lowering the cost of better quality information

analysis
b. Improve business performance by disseminating information more effectively
across the organization
c. Improve collaboration by developing a common, enterprise-wide business

intelligence, integrating views on identified business opportunities
d. Generate and pre-test value propositions utilizing advanced and discovery
analytics.
Others focus more on new opportunities that arise from the utilization of big data
and big data analytics. For example, Michael and Miller (2013) are arguing for the
opportunities that will arise from mining non-text data, such as videos, pictures,
and voice, as well as humans—machines and machine-to-machine interactions
(LaValle et al. 2011). We’ve also assisting at the steadily increasing of the amount
of data captured in bidirectional interactions, both people-to-machine and machine-
to-machine, by using telematics and telemetry devices in systems of systems.
Particularly, interesting is the impact of big data on Health-related industries.
Integrating and sharing different forms of biological information, from high-reso-
lution imaging such as X-rays, Computed Tomography (CT) scans, and Magnetic
Resonance Imaging (MRIs) to health records and lifestyle choices, expert com-
munities can get a better understanding of what makes us ill and what keeps us
healthy. According to Adrian Usher, Chief Information Security Officer of the of
the Skype division at Microsoft, particularly interesting for the sector will be the
integration of nanotechnology embedded in people, that will be utilized as a
monitoring and diagnostics tool (Shah 2013).
While big data visionaries talk about business advantages and opportunities,
others warn about its risks, particularly those of infringing on our privacy and
abusing our civil liberties, as well as being discriminated against (Lerman 2013).
Data breaches and security concerns are pertinent both in decentralized, ubiquitous
and in centralized cloud conditions; also concerns about privacy and confidentiality
are not characteristics about big data, but on social data utilization in general
(see digital business identity issues discussed in Morabito (2014)).
The real “elephant in the room” is that big data analysis seeks to make inferences
about who we are based on our online behavior as if this is the whole picture. In
addition, predictive analytics are modeled based on human theories about cause and
effect attributing perhaps the wrong labels to people. For example, Jeffrey Zaslow
gives an account of how he ‘wrestled’ with his Tivo machine algorithms to avoid
inaccurate stereotyping based on his TV recordings (Zaslow 2014). While in
entertainment this makes a funny account, things could get more serious in
healthcare.
1.6 Case Studies
In 2008, Groupon was the first organization to offer prepaid discount vouchers for a
variety of services, restaurant discounts, Spa experiences, and museum visits at
discounted prices of 50 % or more. Given its wide appeal, companies in the
business to consumer (B2C) sector used Groupon to reach new customers at
introductory prices and to create online buzz.
1.6 Case Studies 15
On the basis of its popularity, Google offered Groupon $6 billion and eventually
it was valued at $12.7 billion. Meanwhile, 400 similar websites began to copy the
business model. By February 2013, the company has failed twice to meet their own
earning predictions, thus, not translating revenues into profits, and their CEO is
fired (Lappin 2013). In year 2014, their share trades 40 % down from last year at $7
a piece almost a third its $20 Initial Public Offering price (Lappin 2011). Yet, it has
44 million customers and increasing sales. Selling local stuff seems to be its more
profitable segment with profit margins around 30 %. So how can big data help
Groupon ramp up its profits? In 2013, Groupon loses his Chief data scientist.
The Groupon team has been using open source data analysis software that was
appropriate for predictive analytics from a personal computer. Although, it could
not be used for Groupon’s total set of dataset information. This made responses to
be disconnected, slow and cumbersome, yet one Groupon analysts have spent too
much time configuring to their needs to let go.
Another, grappling issue for Groupon’s business model is cutting down on
overhead costs, i.e. management and administrative costs, which are around 50 %.
Fine grained, internal performance analysis of its operational costs can identify
where savings can be made to optimize operations at a global scale.
Point of Attention: Adopting big data and big data analytics does not
translate into competitive advantage as such. You need to have a clear
strategy, in the face of anticipated competition from copycats, whether to
protect your advantage or to utilize next generation technologies to better
adapt in changing market conditions.
Groupon invests in serving its most profitable market better. Taking advantage
of its large customer base, it has now invested in real-time, location-based services
to match customers’ position with nearby deals and offers.
Will that offer Groupon a distinctive competitive advantage? How long before
such technology can be copied by competitive sites, like Living Social (2014), or
deal aggregators, like Yipit (2014)? And what would stop, the likes of Amazon, to
enter the push market of geolocated deals to its own customer base?
Let’s consider now the case of pharmaceuticals. MIT professor Natasha Dow
Schull wrote in MIT Technology Review that technology companies are betting
that the affordable care Act’s mandate to cut costs will boost use of gadgets that let
people monitor their own conditions (Schüll 2014).
In a recent Digital Health Summit she noted that it was featured: “smart scales
and water bottles, digital pedometers, electronic skin patches, heart-rate-detecting
earphones, and an impressive collection of wristbands packed with sensors to log a
person’s steps, heart rate, sleep phases, and more…. The pitch was that a person
needing guidance on daily lifestyle decisions such as what to eat, when to sleep, and
how much to exercise could simply consult the data dashboard” (Schüll 2014).
These are only some of the wearable sensors that will be collecting data on our
health and provide biofeedback, transforming, thus, the end-to-end healthcare
supply chain and even the kind of therapeutic interventions available to people.
While these promise to change the competition landscape, western medicine is still
very much drug based and pharmaceuticals are getting into the big data game early.
A big collaboration project amongst pharmaceutical giants was announced in
2013, it is named ‘Project Data Sphere’ (LSC 2014). This platform allows
researchers to combine trials to create larger more helpful data sets, something they
have wished for a very long time. A goal of the Project Data Sphere is to spark
innovation through access to comparator arm data from historical cancer clinical
trials. The data can allow for more efficient research through improved trial design
and statistical methodology, reduced duplication and smaller trial sizes, as well as
the development of broader data standards (LSC 2014). Astrazeneca, Bayer,
Celgene, Johnson and Johnson, Pfizer, Memorial Sloan Kettering Cancer Center
and Sanofi committed to pulling together information resources to accelerate cancer
drug discovery (Comer 2014).
The idea is simple, pull together all information resources about clinical data and
make it available to the international community of clinical researchers. The target
is to beat them in the drug obsolescence race, constantly improving the offerings to
cancer patients.
SAS provides the platform and Sinequa the capability to index and analyze data
in different languages. The reason behind the collaboration is simple: improve the
rate of bringing new cancer medicine in the market and doing it while lowering
costs. What is the motivation of each of the big pharmaceutical to collaborate? To
understand this we need to understand the structure, current dynamics, and future of
the industry and where competitive advantage is coming from now and in the
future. To sell cancer medicine is consider as a business to business (B2B) and
business to government (B2G) activity. While the medicine is used by the public it
is actually governments and private insurers who decide which medicine they will
subsidize and avail to the public.
Cancer patients themselves have no knowledge and no say in this process, and of
course cancer medicine is not sold over the counter. Hence, price competition
happens at a B2B level. In most countries, insurers differentiate their offerings too
(Schultz 2013). Low paying customers get a lower contribution and highly paying
customers a higher contribution to drugs. Pharmaceuticals incentivize the inclusion
of their branded prescription drugs in insurer and government lists, with rebates,
i.e. money back discounts for quantity purchases, in order to compete with generics.
Generics are low cost, non-branded, copy cut medicine with the same or similar
therapeutic effects (Provost and Fawcett 2013).
With most governments shifting towards cutting down on healthcare costs,
generics have now 70 % of the market in most areas, pushing revenue growth of the
top pharmaceuticals down to 2–4 % in 2011–2012 from 11 to 14 % in 2003–2007
(Kaplan et al. 2013). While this is true in most areas, it is not for cancer treatment.
Why? Generic cancer drugs have low profit margins, hence there are only a few
1.6 Case Studies 17
companies producing them, leading to drug shortages when production issues arise.
So there is a market gap for cheap cancer medicine (Pellegrino Blog 2014).
In addition, the dominance of pharmaceuticals in health care may change.
Advances in other areas, such as high-end medical nanotechnology and biologics,
and a shift in focus from treatment to prevention will enable new types of entrants
will take away market share of the healthcare budget. Hence, it is in the interest of
pharmaceuticals to collaborate in order to compete, not only amongst themselves
but very likely with up and coming technology companies too.
Point of Attention: Big data and analytics can help industries redefine them-
selves and the value of product offerings to their customers. Big data may enable
traditional industries to make big leaps forward, but it will shift competition
within the sector to competitors outside the traditional industry boundaries.
Big data analytics in the sphere of bioinformatics can make collaboration

between companies in order to possible. To give you an example, the cost of
reading a human genome in 2001 was $95 m. In the near future, it will probably
cost $1,000 and it will be a matter of hours. Moreover, as wearable biosensors are
increasingly adopted, vast amounts of day to day contextualized biological data can
be aggregated. Such data can now be easily aggregated and correlated not only to
understand diseases better, but also to personalize treatments.
1.7 Recommendations for Organizations
Several recommendations arise regarding strategy in the big data era. Whilst, these
are neither new nor unique to big data-driven business modeling, they are however
accentuated in this era. Recommendations arise in the whole spectrum of strategy
development, from strategy formulation questions to synthesizing insights to
implementing strategic decisions, to ensuring the existence of contingency planning.
Big data (Junqué de Fortuny et al. 2013).
1.7.1 Ask the Right Questions
The big data era favors organizations with foresight and collaborative skills and
public good orientation. Big data can provide answers to many questions, so many
in fact that one can get lost in them. To this end, a focus on skills and education is
relevant to have a competitive exploitation of the opportunity coming from big data
and big data analytics. Companies with PhD level people in their boards are likely
to have a differential advantage over companies with good administrators. Why?
Simply because PhDs are trained to turn problems into relevant questions, design
different methods to answer them, understand the bias in these methodologies,
understand a plethora of data analysis and finally can turn such insights into
persuasive reports. Yet, some classic, strategic questions should be asked:
1. Does big data change your business model by creating new business opportunities
or change your relationship with suppliers and customers?
2. Can big data improve the value added to your customers by your offering?
3. Can big data help you identify new target customers and keep them?
1.7.2 Look Out for Complementary Game Changing

Innovations
Big data technologies do not develop in isolation. Other technologies develop in

parallel that will have synergistic effects with big data. Companies who want to take
advantage of the big data revolution will have to keep an eye on such technologies
and how they can combine to add value to their business model. For example,
embedding sensors into humans and animals will give a new meaning to ubiquitous
computing. Also transform the potential of industries, such as bioinformatics,
pharmaceuticals, health services, and even insurance.
Researchers at the University of Washington have now managed to establish
brain-to-brain communication via computer interfaces, where ‘person A’ was able
to send a brain signal via the Internet to control the hand motions of a ‘person B’
(Armstrong and Ma 2014). Which industries might be disrupted by combining big
data analytics with brain-to-brain interface communications might be a question for
organizations in the learning and health fields, for example, to envision.
1.7.3 Develop Sound Scenarios
Now more than ever, competitive analysis should be centered on customer needs. As
new business models arise, companies should be constantly on the lookout for
emerging competitive offering arising in unexpected fields. Sound scenario planning
will require not only an analysis of political, economic, social, technological, legal
and environmental factors, but also of paradigm shifts occurring by the interplay of
these factors. Big data discovery analytics can expose us to indications of such
changes, but it will be fundamentally our own open mindedness that will determine
whether we are able to synthesize them into comprehensive views of the future.
1.7.4 Prepare Your Culture
Cultural barriers have always been hurdles to major changes. Turning a company to
be data centric is a big change in its own right, particularly when we talk about
discovery. Discovery requires a bottom-up, inductive thinking. It requires people to
observe in order to develop hypotheses to test and this demands that they shed their
1.7 Recommendations for Organizations 19
preconceptions, stereotypes and need for clarity. Most organizations have selected
and conditioned their management to be otherwise; good efficient administrators,
some with better interpersonal skills and a flair for ‘big-picture’ thinking, yet not
many with the new skills required. In addition, employee assessments are structured
around productivity and efficiency, perhaps also relational skills, but certainly not
around innovativeness. If we want this paradigm shift, we need to review how old
systems work contrary to it.
1.7.5 Prepare to Change Processes and Structure
Discovering a new strategic choice will not lead to any advantage unless organi-
zations can act upon it. Making strategic units data-driven requires embedding data
analysts at the core of the team. And this alone is a big change. In addition, big data
can give first movers advantage to organizations only if they have the capability to
review their processes and restructure accordingly. Designing a modular enterprise
IT architecture that can be easily reconfigured will be fundamental. Doing so,
within a dedicated community of practice including the whole value chain, will be
even more profitable.
1.8 Summary
In this chapter we have discussed how big data can be utilized to achieve com-
petitive advantage. We opened this discussion by reviewing the evolution of our
understanding of competitive advantage and highlighted two advancements: on one
hand competitive advantage is a dynamic ever evolving effort, as opposed to a
sustainable asset. On the other hand shift from organizational to community
ownership of competitive advantage.
Then we illustrated how big data can affect different aspects of a business model
that offer add-on capabilities to organizations. In particular, we discussed the effects
of big data on marketing, innovation and business model design (Kearney 2014),
touching upon the different forms of collaborative organization upstream and lat-
erally to provide value added offerings.
Current challenges for existing organizations were highlighted as well as orga-
nizational advantages from embracing big data and business opportunities that open
up in the new era. Two case studies were discussed: Groupon and the Pharmaceu-
ticals big data consortium, from different points of view. We highlighted the
importance of scenario planning and the monetization of big data driven business
models, as well as the role of big data to set into gear big changes in different sectors.
The question as to whether big data will provide a differential advantage to one
company over another is still open. Whole sectors seem to embrace big data con-
currently leaving their weakest links behind to perish. In addition, technology
companies seem to have an advantage over others, as they have ample expertise to
deploy in data analysis. This leads to big technology providers making inroads to
previously traditional sectors, such as, e.g., travel and retail. Thus, big data is
fussing out industry boundaries and competition, perhaps even our understanding of
what competitive advantage is.
References
Adamson, B., Dixon, M., Toman, N.: The end of solution sales. Harv. Bus. Rev. 90, 60–70 (2012)
Armstrong, D., Ma, M.: Researcher controls colleague’s motions in 1st human brain-to-brain
interface. http://www.washington.edu/news/2013/08/27/researcher-controls-colleagues-motions-
in-1st-human-brain-to-brain-interface/ (2014). Accessed 15 Nov 2014
Asos: Asos discover fashion online. https://marketplace.asos.com/ (2012)
Innocentive: AstraZeneca Challenge: Targeted delivery of oligonucleotides. https://www.
innocentive.com/ar/challenge/9933013 (2014). Accessed 15 Nov 2014
Bricker & Eckler: Board Demographics: compare your board’s composition in terms of age,
gender, and ethnicity. http://www.acredula.com/details.aspx?id=105 (2014). Accessed 15 Nov
2014
Brumley, J.: Google just became a real problem for online travel agents. http://investorplace.com/
2014/04/goog-stock-google-travel-booking/#.VCPn_Pl_veg (2014). Accessed 15 Nov 2014
Chen, M., Mao, S., Liu, Y.: Big data: a survey. Mob. Netw. Appl. 19, 171–209 (2014)
Chesbrough, H.: Open Innovation: The New Imperative for Creating and Profiting from
Technology Xerox Parc: The Achievements and Limits of Closed Innovation, pp. 1–10.
Harvard Business School Press, Boston (2003)
Chesbrough, H.W.: The era of open innovation. MIT Sloan Manag. Rev. 44, 9 (2003)
Choi, H., Varian, H.: Predicting the Present with trends. Econ. Rec. 88, 2–9 (2012)
Comer, B.: After delays, project datasphere rolls out. http://blog.pharmexec.com/2014/04/09/after-
delays-project-datasphere-rolls-out/ (2014). Accessed 15 Nov 2014
Davenport, T.H., Patil, D.J.: Data scientist: the sexiest job of the 21st century. Harv. Bus. Rev. 90,
70–76 (2012)
Dignan, L.: IBM’s watson unit invests in fluid, eyes personal shopping applications. http://www.
zdnet.com/ibms-watson-unit-invests-in-fluid-eyes-personal-shopping-applications-
7000028580/ (2014). Accessed 15 Nov 2014
Drnevich, P.L., Croson, D.C.: Information technology and business-level strategy: toward an
integrated theoretical perspective. MIS Q. 37, 483–509 (2013)
Fine, C.H., Vardan, R., Pethick, R., El-hout, J.: Rapid-response capability in value-chain design.
Sloan Manag. Rev. 43, 69–75 (2002)
Gardner, J.: Who owns big data: The CMO or CIO?. http://www.wired.com/2013/06/who-owns-
big-data-the-cmo-or-cio/ (2014). Accessed 15 Nov 2014
Griggs, B.: The CNN 10: future of driving (2014)
Guinard, D., Trifa, V.: Towards the web of things: web mashups for embedded devices. In:
International world wide web conferences, pp. 1–8 (2009)
Hafkesbrink, J., Schroll, M.: Innovation 3.0: embedding into community knowledge—collaborative
organizational learning beyond open innovation. J. Innov. Econ. 7, 55 (2011)
Hagen, C., Ciobo, M., Wall, D., Yadav, A., Khan, K., Miller, J., Evans, H.: Big data and the
Creative Destruction of Today’s Business Models, pp. 1–18. Kearney Publishing, Chicago
(2013)
Harvard Innovation Lab: What is experfy?. https://www.experfy.com/ (2014). Accessed 15 Nov
2014
Hine, C.: Virtual ethnography. In: Fielding, N., Lee, R.M., Blank, G. (eds.) The Sage Handbook of
Online Research Methods, p. 179. Sage, London (2000)
References 21
Im, J., Kim, S., Kim, D.: IoT mashup as a service: Cloud-based mashup service for the internet of
things. In: Proceedings—IEEE 10th international conference on services computing, SCC
2013, pp. 462–469 (2013)
Junqué de Fortuny, E., Martens, D., Provost, F.: Predictive modeling with big data: is bigger really
better? Big Data. 1, 215–226 (2013)
Kaplan, W.A., Wirtz, V.J., Stephens, P.: The market dynamics of generic medicines in the private
sector of 19 low and middle income countries between 2001 and 2011: a descriptive time series
analysis. PLoS ONE 8, e74399 (2013)
Kearney, A.: Big data and the creative destruction of today’s business models—strategic IT article
—A.T. Kearney (2014)
Kickstarter.: http://www.kickstarter.com (2014). Accessed 15 Nov 2014
KPMG: Brand and ownership concentration in the European automotive industry (2010)
Lankhorst, M.H.R., Ketelaars, B.W.S.M.M., Wolters, R.A.M.: Low-cost and nanoscale non-
volatile memory concept for future silicon chips. Nat. Mater. 4, 347–352 (2005)
Lappin, J.: Groupon, Down 40 % in 2014, finally does something for shareholders. http://www.
forbes.com/sites/joanlappin/2014/05/05/groupon-down-40-in-2014-finally-does-something-
for-shareholders/ (2013). Accessed 15 Nov 2014
Lappin, J.: Deduced reckoning is assembling the mosaic! http://www.forbes.com/sites/joanlappin/
2011/02/04/deduced-reckoning-is-assembling-the-mosaic/ (2011). Accessed 15 Nov 2014
Lavalle, S., Lesser, E., Shockley, R., Hopkins, M.S., Kruschwitz, N.: Big data, Analytics and the
Path From Insights to Value. MIT Sloan Manag. Rev. 52(2), 21–31 (2011)
Lerman, J.: Big data and its exclusions. Stanf. Law Rev. 55, 55–63 (2013)
Lesk, M.: Big data, big brother, big money. IEEE Secur. Priv. 11, 85–89 (2013)
Life Sciences Consortium (LSC): Project data sphere, project data sphere. Accessed 15 Nov 2014
Living Social: Livingsocial https://www.livingsocial.com (2014). Accessed 15 Nov 2014
lixpen.com: LIX—The smallest 3D printing pen in the world. https://www.kickstarter.com/
projects/lix3d/lix-the-smallest-3d-printing-pen-in-the-world (2014). Accessed 15 Nov 2014
Michael, K., Miller, KW.: Big data: new opportunities and new challenges. Computer 46, 22–24.
Long Beach, California (2013)
Morabito, V.: Trends and Challenges in Digital Business Innovation. Springer International
Publishing, New York (2014)
Neiger, C.: 5 Future car technologies that truly have a chance. http://auto.howstuffworks.com/
under-the-hood/trends-innovations/5-future-car-technologies.htm#page=0 (2014). Accessed 15
Nov 2014
O’Brien, D.P.: Markets and hierarchies: analysis and antitrust implications (book review). Econ.
J. 86, 619–621 (1976)
Paharia, R.: Loyalty 3.0: How to Revolutionize Customer and Employee Engagement with Big
Data and Gamification. McGraw Hill Professional, New York (2013)
Pellegrino Blog: Generic drug shortage limits affordable cancer treatment options. http://www.
pellegrinoandassociates.com/generic-drug-shortage-limits-affordable-cancer-treatment-options/
(2014). Accessed 15 Nov 2014
Porter, M.E.: Competitive Advantage: Creating and Sustaining Superior Performance. The Free
Press, New York (1985)
Provost, F., Fawcett, T.: Data science and its relationship to big data and data-driven decision
making. Data Sci. Big Data. 1, 51–59 (2013)
Raice, S., Woo, S.: Groupon’s Boston problem: copycats. http://online.wsj.com/news/articles/
SB10001424052702303763404576420090000910026 (2011). Accessed 15 Nov 2014
Ratti, C.: Forget flying cars. Smart cities just need smart citizens. http://www.archdaily.com/
author/carlo-ratti/ (2014). Accessed 15 Nov 2014
Schüll, N.D.: Obamacare meets wearable technology. http://m.technologyreview.com/view/
526576/obamacare-meets-wearable-technology/ (2014)
Schultz, T.: Turning healthcare challenges into big data opportunities: a use-case review across the
pharmaceutical development lifecycle. Bull. Am. Soc. Inf. Sci. Technol. 39, 34–40 (2013)
Schumpeter, J.A.: The Theory of Economic Development: An Inquiry Into Profits, Capital, Credit,
Interest, and the Business Cycle. Harvard University Press, Cambridge (1934)
Shah, S.: Big data will go mainstream when nanotechnology is embedded into humans, says Skype
CISO. In: Computing, 26 Apr 2013, http://www.computing.co.uk/ (2013). Accessed 7 Jan
2015
Styles, K.: Groupon adds location-based deals for mobile users. http://mobilemarketingmagazine.
com/groupon-adds-location-based-deals-mobile-users/ (2014). Accessed 15 Nov 2014
Taylor, W.C., LaBarre, P.G.: Mavericks at Work: Why the Most Original Minds in Business Win.
HarperCollins, New York (2008)
Townsend, A.: A manifesto for smart city development. http://www.theeuropean-magazine.com/
anthony-townsend–2/8244-a-manifesto-for-smart-city-development (2013)
Verganti, R.: Design driven innovation. Harv. Bus. Sch. 40, 288 (2009)
Werbach, K., Mehta, A.: The spectrum opportunity: sharing as the solution to the wireless crunch.
Int. J. Commun. 8, 22 (2014)
Wu, M.: Towards a stakeholder perspective on competitive advantage. Int. J. Bus. Manag. 8,
20–29 (2013)
Yipit: Yipit has every deal in your city. http://www.yipit.com (2014). Accessed 15 Nov 2014
Zaslow, J.: If TiVo thinks you are gay, here’s how to set it straight. http://online.wsj.com/articles/
SB1038261936872356908 (2014). Accessed 15 Nov 2014
for Government Innovation 2
Abstract
This chapter discusses the transformation of the public service provision model
due to big data, and in particular due to public engagement in the context of open
government initiatives. We outline the changing role of governments in
societies, and the technological enablement towards direct online democracy and
active citizen engagement, as well as the utilization of big data enabled
governance as a competitive advantage for attracting resources and talent to
maintain a global smart megacity status. To this end, this chapter discusses the
utilization of (a) new sources of data, such as Crowdsourcing, Internet of Things,
(b) engage public talent, (c) institutionalize private–public partnerships and
(d) seeks for new models of value-for-money public provision, but also the
challenges that big data present us with respect to data ownership, data quality,
privacy, civil liberties, and equality, as well as public sector’s ability to attract
big data analyst talent. We demonstrate different aspects of this discussion
through two case studies: Barcelona Smart City and Haiti’s emergency support
during the 2010 earthquake disaster.
2.1 Introduction
This chapter discusses the impact of big data in the public sphere on public service
provision and new opportunities for public service organization and structure that
may transform the role of governments in societies.
The utilization of ICT to improve public sector services has started with the
whole e-government discussion. Transforming government services using ICTs has
been a complex and costly task, often associated with the automation of public
services and business systems integration. While e-government projects focused on
operational efficiency, initiatives such as Open Government efforts sought to foster
public service transparency, civic participation, and inter-departmental collabora-
tion. This could be achieved by sharing public sector infrastructure, seamless
information sharing with other agencies, bundling core competencies to improve

DOI 10.1007/978-3-319-10665-6_2
24 2 Big Data and Analytics for Government Innovation
service delivery and engaging external entities, such as universities and businesses
(Executive Office of the President of USA 2014).
While these changes definitely seek to effect efficiencies, they are also qualitative
in nature, changing fundamentally the nature of the relationship between govern-
ments and citizens. Big data initiatives come to underpin their progress. Since 2012,
both EU and the US are seeking ways, though legislative and policy changes, to
remove obstacles in the use of big data which promise greater effectiveness with
lower costs in the public sector (Nagy-Rothengass 2013). Civic participation via
social media, for example, can also reduce the cost of public service delivery.
Crowdsourcing information on potholes, for example, can cut down on inspection
costs. Big data also promise clockwork provision of public service. Intelligent assets,
such as intelligent traffic lights can notify a central asset management system about
their state of maintenance ahead of time and larking issues in their working condi-
tion, so repair work can be streamlined without disruption of service (Thomas 2013).
Before we go on to elaborate on the role of big data in civil life, it is important to
understand some underlying shifts in the role of Governments and its relationship
citizens.
2.1.1 New Notions of Public Service: Towards

a Prosumer Era?
Everything about civil service, even its very naming “service”, has emphasized the
transactional relationship between citizens and government. The relationship is
simple. Civilians pay taxes and in exchange they are served in various fields, health,
education, road maintenance and the like.
Recently, however, a new understanding of public provision put citizens in the
role of partners. The key idea is that the pursuit of public ends is the responsibility
of everybody—private and nonprofit entities, the public, and government. Part-
nership with citizens and community involvement was a big part of the 2010
Manifesto of the Conservative party in UK under the auspices of the “Big Society”
program (Conservative Party 2010). The US Open Government initiative,
announced only a year before in 2009 also seeks foster civic participation
(McDermott 2010). Both are premised on the idea of enabling people to take care of
themselves and of each other. Social media and smartphones can facilitate the
interaction between citizens and governments on the go. They can also amplify the
communication and engagement of public though communities of interest. To
combat crime, for example, citizens need to coalesce with police in monitoring and
reporting suspicious activities. Recently, this is also happening in other areas too of
civic responsibility.
Application developers, such as Citysourced.com have developed applications
enabling citizens and residents to report and provide information to local govern-
ment about all sorts of civic issues, from potholes to graffiti, fly tipping, broken
pavements or street lights. People can do so anonymously or not, they can upload
2.1 Introduction 25
photos, and pin them on a street map. The report is sent to councils and there is a
tracking of progress on the issue online (CitySourced Inc. 2014). This is a typical
example of how technology has facilitated citizens to play the role of council
inspectors and this is a free service to the community and to the government too, as
it minimizes inspection costs. Charities and interest groups work together to
amplify the message. Cyclist communities, for example, have a big interest in
potholes as it is a big nuisance for them, so pothole reporting is promoted by
cycling charities and associations (The National Cycling Charity 2014).
2.1.2 Online Direct Democracy
And even more fundamental change might take place due to social changes and
technology advancements; one that aspires to give citizens decision making power
on social issues, much like the type of direct democracy of ancient Greece. PartyX
(Nelson et al. 2015), is such an initiative. They seek to take advantage of develop-
ments in online collective decision making, to involve every stakeholder in
political decision making. As this is currently on beta version and used only at local
level for local issues, it does not fall under the big data agenda as yet. Should this
kind of technology however go into adoption phase and used to debate global
issues, then we can start seeing big data making inroads into the political and
legislative sphere. Real-time, big volume, unstructured information aside, global
political debating will add another dimension of interest in our discussion of big data;
that of ‘multilingualism’. Dealing with multilingualism is a dimension already
high in EU agenda (Nagy-Rothengass 2013).
2.1.3 Megacities’ Global Competition
Since 2011, more people live in cities than in rural areas for the first time in human
history. Megacities, i.e. cities larger than 10 million people, are an emerging
phenomenon. According to the UN, the number of megacities will have grown from
five in 1975 to 26, with 24 of them located in the developing world (United Nations
Department of Economic and Social Affairs 2006). Megacities are not a local or
national issue. They will affect the future prosperity and stability of the entire world
as they will shape the balance of power of national economies in a global world,
affect population mobility and configuration of talent, and will influence the social
and political dynamics of the world (United Nations Department of Economic and
Social Affairs 2006). Megacities have a functional and a symbolic role. What would
UK be without London? And what would The Emirates be without Dubai?
Megacities are not just key instruments of social and economic development at
all fronts but also harbors of social innovation for the private and public sector due
to their unique dynamics. Megacities are an attractive proposition for those seeking
a better quality of life in terms of a higher standard of living, better jobs, fewer
hardships, and better education. In a globally competitive environment Megacities

compete for capital resources including global talent. They typically face a 5 %
population growth rate, which challenges the quality of living indices (such as
security, cost of living, mobility, employment, environmental) that put pressure on
urban infrastructure and public policy. To raise their attractiveness, megacities need
to improve on those indices (United Nations Department of Economic and Social
Affairs 2006; Mostashari et al. 2011). Hence, Megacity Mayors face unique
dilemmas, primarily on how to raise standards of living across a number of well-
being indices in the face of high population growth rates, while compete in the
global environment.
Smart city infrastructure, the bundle of Internet of Things solutions for city
infrastructure management and intelligent infrastructure-citizen interfaces are con-
sidered to provide a way forward. The quantity of data produced and the criticality of
infrastructure management will raise the bar for big data analytics and management.
We will discuss this opportunity as part of the content in the next section.
2.2 Public Service Advantages and Opportunities
2.2.1 New Sources of Information: Crowdsourcing
Crowdsourcing is becoming an increasingly common term and opens new avenues

for creating free public value, civic engagement, and transparency. It can take many
forms. ‘Crowdreporting’, for example, is a common form of crowdsourcing in the
public sphere, at the moment, and in line with the new conception of citizen as a
partner. The “SeeClickFix.com” is a typical example (SeeClickFix 2014). It is an
online service designed to help citizens report non-emergency issues in their
neighborhood, via a web interface, Facebook or smartphone apps. The issue han-
dling process is tracked online. After the issue is reported, it is tracked online the
same way logistics companies track the delivery of packages to their destination,
only that information is published via Twitter and Facebook to inform the public
(SeeClickFix 2014). “Nothing new” one might say, reporting issues like this could
be done in the past using other means, like calling the council or writing a letter.
What is so different after all? I guess the answer should be immediacy and trans-
parency, and perhaps non-evasiveness.
The public does not need to go out of their way to report such issues anymore.
There is an app for that or they can just log into Facebook. There is no wait on the
telephone to reach an operator, there is no time consuming writing of memo. The
process is blended into our everyday social life. Now everybody with a smartphone
can go around and report civic issues. In addition, direct feedback and traceability
can give people satisfaction and a sense of achievement that they have contributed
to common good. In the past, reported information was not acknowledged and
follow-up information was non-existent. So, direct interaction between public and
government agencies in this civic reporting encapsulated three governmental goals:
2.2 Public Service Advantages and Opportunities 27
(i) to engage the public into civic life, as citizens actively engage in the process; (ii)
to decrease the cost of civil service, as citizen engagement is voluntary and free of
charge; and (iii) to improve transparency of public service processes, as the issue
handling process is now traceable online at par with private organizations (Vicini
and Sanna 2012).
The creation of public goods via crowd reporting is not, however, the sole
privilege of government agencies. Weather underground, for example, combines
crowd-sourced human observation with weather station data to establish a new level
of accuracy within weather reporting. Weather data is assimilated from 2,000
weather stations maintained by the Federal Aviation administration, 26,000 stations
part of the Meteorological Assimilation Data Ingest System (MADIS) and a 16,000
of personal weather stations adhering to quality controls and standards. Coupled
with crowd observations and meaning scientific analysis from meteorologists
provide valuable insights for the co into the science behind the data and the rela-
tionship between weather and climate change (The Weather Channel Inc. 2014).
This blend of human insight with private and public sensor systems, gives rise to
the idea of the Internet of Everything (Danova 2014), the merge of structured and
unstructured data in a variety of forms, from textual to pictures, videos and audio
material.
Crowdreporting can also take the form of feedback. For example, the “Did You
Feel it” service in the US surveys people on a number of earthquake parameters
regarding their experience of particular earthquake incidents. Enriching numerical
descriptors with empirical data can enrich knowledge of qualitative descriptors such
as intensity (USGS 2014). Using crowd feedback to train intelligent Internet of
Things (IoTs) technology utilizes the wisdom of the crowds in artificially intelligent
public systems. We discuss this idea of public organization in the subsequent
section, when we introduce the concept of Cognitive city as a potential domain for
big data analytics application.
2.2.2 New Sources of Information: Internet of Things (IoTs)
While there is not a commonly agreed definition, the Internet of Things (IoTs)
refers to the network of intelligent devices which include sensors to measure the
environment around them, actuators which physically act back into their environ-
ment such as opening a door, processors to handle and store the vast data generated,
nodes to relay the information and coordinators to help manage sets of these
components (Zhang and Mitton 2011).
IoTs have made way into utilities, smart homes, healthcare and wellbeing
applications, and they are expected to proliferate into other areas, such as
commuting and transport (as shown in Fig. 2.1).
IoTs will push our data storage, connectivity and architecture limits to a new
high. The socio-economic implications for how will live our lives might be huge.
For example, McKinsey Global Institute (Manyika et al. 2011) reports a 300 %
2,00,00,000
1,80,00,000
1,60,00,000
1,40,00,000
1,20,00,000
1,00,00,000
80,00,000
60,00,000
40,00,000
20,00,000
Connected Cars Wearables Connected Tvs Internet of Things

Tablets Smartphones PCs
Fig. 2.1 The internet of everything adapted from Danova (2014). The number of devices in use
globally are shown on the y-axis
increase in connected machine-to-machine devices over past 5 years and a sheer

80–90 % decline in microelectronics pricing which is anticipated to lead in 1 trillion
more ‘things’ connecting to the Internet across industries such as manufacturing,
health care, and mining with a potential $36 trillion cost saving in operating costs
(Manyika et al. 2011).
While big emphasis is given currently into the development of intelligent assets
by equipping them with sensors that beam information about their properties and
conditions, what really makes up the internet of things is their distributed, pur-
poseful collaboration, and this requires architecture, i.e. organization and structure,
in line with the ethics of transparency (Bentley et al. 2014). Hence most of such
technologies are clustered around common purpose for example smart city tech-
nologies, or smart car technologies to denote what drives the logic of their archi-
tecture. One of the trendiest such clusters is, of course, Smart City, which suggests
an advanced connectivity through a highly networked city infrastructure via
intelligent assets. The promise of course is a fertile foundation and over time
capable to attract new businesses and investments and further urban growth and
socio-economic development (URBACT 2012).
While the Internet of things can provide efficient use of resources, it nevertheless
lacks in responsiveness and agility. To instill responsiveness into city management,
we may be required to take a step further into developing cognitive city systems.
Decisions about cities need to be based on principles, values, and thereby quali-
tative metrics as to how people want to live their lives, and what well-being means
to them. Hence, a smart city is as different to a cognitive city, as a robotic arm is to

social anthropoid. A platform that can combine citizens’ active engagement into
decision making about local government’s decisions and community well-being.
Given that architecture is central to our notion of the Internet of things, we can
view cognitive cities as providing the principles for humanizing their design. Not
only sense and perceive but also learn, memorize, recall relevant experiences in
order to adapt their responses accordingly. Almost like biofeedback, people can
provide real-time feedback teaching the system how to behave. After all, cities seek
to improve their quality of life of their citizens; and quality is in the eye of the
beholder. People can train machines to include feature extraction, classification and
clustering, all of which they can then be perfected through successive optimization
of their algorithm (Warner 2011).
Smart city sensors can therefore be trained by people to learn, memorize, recall
relevant experiences, who teaches them and who makes the final value judgment on
things? This is an area of debate still in its infancy, but one that breeds big legal,
political, and ethical questions about the future of social life. EU for example,
supports initiatives that foster social innovation and inclusiveness together with
economic innovation and environmental sustainability to engage citizens in public
service co-creation and local authority support. Most critical to crowdsourcing
success is the feeling by participants that their efforts were considered and that
results came from the initiative (Evans-Cowley 2011). This requires moderators
who are competent social networkers, able to used crowdsourcing tools that link
and provide feedback easily, as well as a change in processes used to develop
governmental services (Brabham 2009).
2.2.3 Public Talent in Use
Consultation is nothing new in public life. In developed nations it is actually

institutionalized. City and regional planners are always looking for ways to engage
the public in the process of planning. As a consequence Crowdsourcing of ideas
and solutions has emerged as a web-based model to help in solving difficult
challenges. This Section explores the use of crowdsourcing to support problem
solving in planning. A case study of designing a planning curriculum using
crowdsourcing is highlighted. In this case, crowdsourcing was a successful model
for generating creative ideas to support curriculum revision (Evans-Cowley 2011).
The study discussed in Evans-Cowley (2011) that people chose to participate for
altruistic reasons, such as an opportunity to contribute to the community, to con-
tribute their knowledge, and that they wanted to be part of a conversation on the
topic. Thus, public involvement is a central concern for urban planners. Consid-
ering the difficulties inherent in the typical public involvement process, the
challenge for planners is how best to implement such programs. The Web can be
used to exploit collective intellect among a population in ways one-to-one meetings
(Brabham 2009).
Consequently, at the state of the art, for example, (Brabham 2009) argues that
the crowdsourcing model is appropriate for enabling citizen participation in public
planning projects. Starting with an exploration of the challenges faced by public
participation in urban planning projects, (Brabham 2009) supports the argument of
the Web as an appropriate technology for harnessing far-flung genius. Then, it
concludes with an exploration of crowdsourcing in a hypothetical neighborhood
planning example, together with a description of the challenges of implementing
crowdsourcing.
Also, governments gradually use the Internet to aid in transparency, account-
ability, and public participation activities, and there is growing interest in innova-
tive online problem-solving to serve the public good. The crowdsourcing model
influences the collective intelligence of online communities for particular purposes.
To develop better tools that engage the public, it is important to understand how
and why people participate in these kinds of activities. In 2009, for example, the
Federal Transit Administration in US used crowdsourcing for public participation in
transit planning (The Next Stop Design project). Based on interviews with 23
participants, they analyzed the motivations of those participants to engage the
project (OGP 2014).
Taking the above issues into account, it is worth noting that while data is the raw
material of knowledge, it is really the interpretation of such data that can be acted
upon by providing insights, foresight, knowledge or skill, etc. Interpretation of
information requires a schema, a way of looking at data and drawing conclusions.
In modern societies this was traditionally the job of experts. In the digital era, this is
changing! Printing empowers individual ownership of ideas, the digital era
empowers co-ownership. Wikipedia is a prime example of how data and infor-
mation is organized and interpreted in a collaborative way online to provide an
understanding on each subject matter through a constant ever evolving debate. The
idea is premised on Marshall McLuhan motto that “the medium is the message”, in
that a certain medium facilitates certain interpretations that other media do not
(Bentley et al. 2014).
To this end, online communities have become an important source for knowledge
and new ideas. Making “big data” available to a large number of analysts means that
more ideas can converge on how to mine such data. A platform for doing so is
Kaggle (2014), a world’s leading an online platform, which operates as a knowledge
broker between companies aiming to outsource predictive modeling competitions
and a network of over 100,000 data scientists (Kaggle Inc. 2014). Currently, clients
include companies such as General Electric (GE), Ford, Facebook and Microsoft,
and health service organizations, such as Heritage Provider Network, in California
who seeks to develop a predictive algorithm that can identify patients who will be
admitted to a hospital within the next year. Such predictive analytics in the service of
health care provision can inform the development of more accurate Public health
budgets and cost cutting plans (Pinsent Masons LLP 2013).
2.2.4 Private–Public Partnerships
Private–Public partnerships can be seen to emerge in almost every aspect of public

service. With the privatization of most utilities and the trend towards outsourcing,
much of the public sector in advanced societies is run by the private organizations.
Countries in the periphery of Europe, such as Greece are going through the teething
pains of making this transition now. With global competition on public service
management, we are entering of course a new phase of relations between public and
private organizations. That of partnership; the pursuit of a long standing mutually
beneficial and loyal relation that will ensure long-term planning and prosperity
(Miller 2013).
Big Data management has a core role to play in supporting decisions on all of
these partnerships, which is why progressive governments are aiming to mobilize
and support in the field. The event organized by the World Resource Institute on
“Public–Private Partnerships for Open Government” in London 2013 is charac-
teristic. It sought to share the task of mobilizing citizen engagement, developing
more transparent and accountable governments (Stefan and Kisker 2011). This has
implications for data ownership and data management at all public service fronts
from healthcare and natural resources to publicizing data on contracting, govern-
ment expenditure infrastructure, and government aid to third parties. One aspect
that deems these partnerships fundamental is the lack of expertise within the public
sector and the ability to vet and support them. Another, of course, is the upfront
investment in infrastructure, big data clouds for example, that large organizations
can afford with the view to recoup them though long-term government contracts.
Yet another reason is the ability and urge of private companies to experiment with
new technologies in less sensitive contexts and provide governments with ‘safe’
technology options that are not like to raise public criticism.
Due to the power to profile people and triangulate information about individuals,
big data analytics offers deals with some fundamental concerns of public services.
Identity management, for example, is a big issue for most public services, from tax
collection to health provision to parking ticket collection. Big data analytic tech-
nologies may make it possible for public service organizations to combat fraud,
improve data breach monitoring and authentication (Zhang and Chen 2010). On the
other hand, getting it wrong may have some serious repercussions. We discuss this
in more details in Sect. 2.3.
2.2.5 Government Cloud Data
Governments have slowly embarked on Cloud computing to tackle the much to be

desired transparency, participation, and collaboration amongst its agencies but also
with the public. The idea, the concept, and the term, that is cloud computing, have
passed into common currency in an ambiguous manner. The very concept
characterized by three main entities—Software, Hardware and Network to describe
the fusion of Virtualization, Grid computing, Utility computing and Web

technologies that result in new models of IT service delivery (Clark et al. 2013).
Cloud computing permits central governments to uniformly cover the whole
country with e-government solutions, independently of divergence of local
administrative units that may be better or worse prepared to provide e-services.
Service-oriented architecture facilitates provision of compound services covering
whole customer processes, where a customer may be a citizen or an enterprise. The
roll out of such systems can happen simultaneously and cost efficiently, as licensing
and support can be negotiated on the whole.
By now consumers, corporations, and governments are used to store their data to
“the cloud” so they can be accessed from any device, anywhere anytime. According
to Lerman (2013) over 69 % of Americans now use webmail services, store data
online, and utilize online applications. This trend is only going to continue, with
industry analysts predicting clouds to be $40 and $160 billion over the next few
years, not accounting for the internet of things applications (Lerman 2013). With
the wider adoption of cloud computing, the term C-Government was coined to
connote agility due to its virtualization, scalability due to grid computing and the
simplicity of Web 2.0. Clouds need to be interconnected to make it easier for users
to switch between cloud service providers as well as the providers to supply infinite
resources (Yiu 2012).
While the relevance of clouds for governments and their potential, e.g., for
voting information systems has been greatly welcomed, enthusiasm has been
curbed by identify vulnerabilities involved in the digitalization of government
transactions and the electoral process, thus surfacing issues with trust and
transparency (Manyika et al. 2011; Al-khouri 2012). The issues will be dealt with in
later sections.
2.2.6 Value for Money in Public Service Delivery
The whole move from paper-filling to e-government services was to facilitate cost
cutting partially though integration across public services. Big Data offers Gov-
ernments the possibility to do so without going through the pains of Business
systems integration.
The UK alone, according to the Head of Digital Government Unit, Chris Yu,
estimates that some 16–33 billion per year can be saved by taking advantage of big
data in the public sector, which can lead to £250–£500 per capita gains. This is
calculated based on a number of initiatives having to do with performance
management to improve the overall efficiency of government operations, reduce
fraud and error, and tax collection alone (Laney 2014). McKinsey Global Institute
estimates the potential savings at a European level to mount up to €150–€300
billion a year (Yiu 2012).
A number of analytical tools can also mine data regarding citizen sentiment
towards public services to provide feedback and highlight opportunities to
customize service delivery by helping employees better understand the needs of

each citizen. This is no different to how commercial organizations use it to
customize their service to paying customers. Providing quality service to citizens
has been government agenda for most developed nations and in particular those
who want to attract international talent. For example, predictive analytics in the area
of health informatics, particularly epidemiology, can also be a huge help for gov-
ernments who need to gear up for crisis management. The more accurately people
can predict the spread of disease the more cost effective prevention and treatment is
expected to be, the less the disruption to public life.
Such benefits accrue without further changes in the current public service
structures, yet even more benefits may accrue if we consider the potential of
adopting predictive analytics to advance crime prevention and reduce policing
resources or introducing smart grid technology to improve the efficient use of utility
resources, (gas, electricity and water).
2.3 Governmental Challenges
2.3.1 Data Ownership
With data ownership come great responsibility for its management, storage, use and
misuse. But in the age of open (public) data who really bears such responsibility?
Who is the legal guardian of it and what is written in the implicit contract between
civilians and government regarding its use and protection? In one sense all public
data is private data in that it is either civilians personal or lifestyle information or
relates to the functioning of the public service which is nevertheless accountable to
the public, at least in democratic settings. In that sense, governments and public
organizations are custodians of our data, granted permission to use it in exchange of
providing us with public services and promote public good (Gudipati et al. 2013).
Take the United States Patent and Trademark Office online database, for example,
which contains over 8 million patents and 16 million filings dating back to Samuel
Hopkins’s 1790 and receives 200,000 patent applications and 100,000 trademark
applications each year (Hoffman and Podgurski 2013). Who does this information
belong too? Obviously, the patent holders and submitters. Yet, such information
would be useless for both the person and the country unless the US government
safeguards its existence and integrity. Indeed, personal data can be created by a
variety of sources from people, machines, devices, and the rightful owner of such
data is the authority which can verify its veracity as being the ‘True Owner’ of the
data (Al-khouri 2012).
This will bear interesting questions when devices will monitor our behavior and
condition beaming information about our whereabouts and state. While devices are
susceptible to error, people are susceptible to deception and self-deception.
Governments and public agencies will have to make rules to decide how to handle
the inconsistencies.
2.3.2 Data Quality
Big Data can amplify the repercussions and implications of poor data quality, and is
a particularly important issue for governments and civilians alike. Recorded data
can be flawed (erroneous, miscoded, fragmented, or incomplete) due to workload
pressures and user interface workarounds (Junqué de Fortuny et al. 2013). Data
should be checked for completeness, conformity, consistency, accuracy, duplica-
tion, and integrity, and good practices around data quality do exist. In private
organizations, non-quality data can be ignored from consideration, without com-
promising the integrity of analysis or affecting the user. For example, if a retailer
does their analysis only on the ‘clean’ data of their customers to profile and predict
future sales, the customer is not greatly affected (Crawford 2013). The same cannot
be true for the health service, for example, particularly if ‘unclean’ data are
characteristic of a locality (e.g. a local hospital). Poor data quality can result from
integrating data sources. Data issues can also emerge from the integration, feder-
ation or conglomeration of data, and given the variety and volume of big data,
testing this data can be a big task. Various big data testing procedures have started
to emerge, using grid processing technologies, such as Hadoop, to support a timely
processing (Crawford 2013).
While, data stewardship and data testing procedures can deal with the “Garbage
in” problem, another issue is lurking in the seams of public sector decision-making;
and this is all forms bias [selection bias, confounding bias, and measurement bias
(Kerr and Earle 2013)]. Confounding occurs when we correlate two phenomena,
which has been studied independently (Howarth 2014; IQ Analytics 2014). For
example, pulling together smoking habits from Facebook profiles and associating
this with youth diabetes statistics, might erroneous speak of causal relationship that
could be mediated by another factor, for example unemployment or heredity or the
family’s economic status. Selection bias can come from all possible sources and
most importantly due to differential devise use amongst different countries, ages
and socioeconomic groups. Kate Crawford principle research in Microsoft and
Visiting professor of MIT, gives brilliant accounts on the hidden biases in big data
in Harvard Business Review blog (Crawford 2013). Given the sensitivity of public
services that can bear life threatening and legal implications, treating information
with utmost rigor is fundamental. Thus, we need to outline and progress the con-
versations on regulatory and other interventions to address data analysis difficulties
that could result in invalid conclusions and unsound public health policies
(Microsoft 2014).
2.3.3 Privacy, Civil Liberties and Equality
Privacy and civil liberties are the two sides of the same coin. User profiling is the
process of collecting information about a user in order to construct their profile. The
information in a user profile may include various attributes of a user such as
2.3 Governmental Challenges 35
geographical location, academic and professional background, membership in

groups, interests, preferences, opinions, etc. More, ominous applications include
cell-phone tracking and the proposed creation of a national biometric database.
As Kerr and Earle (2013) argue profiling individuals on the basis of their health,
location, electricity use, and online activity raise risks of discrimination, exclusion
and loss of control. When these involve access to public services, repercussions are
exacerbated. The promise of big data is based on prediction with the view to
preempt possible threats. If, for example, one can predict increased burglaries in an
area, local government can increase policing of this area to preempt such incidents
(Barcelona City Council 2014).
Preemptive action is based on prediction and prediction on predictive algorithm
based on social information and this curtails civil liberties replacing proof with risk
estimates. With increased predictive capability comes increased responsibility to
avoid such threats that can make governments more conservative in how they
approach social risks. To what extent can governments afford to let unemployed
youth roam the streets freely, once we have established that it is highly probable to
commit petty theft or drug dealing? On the other hand, shall we detain or curtail the
freedom of youth to meet and socialize and lead them to isolation and depression?
To what extent policemen should be informed about such correlations and would
that lead to discrimination of every unemployed youth, who treated with suspicion
may even turn to crime as self-fulfilling prophesy theory would predict?
Also, Professor Lerman (2013), for example, raises another issue; the issue of
equality regarding public treatment of people and groups who do not fully partic-
ipate in the information society, because they don’t have the means, time or appetite
for. Statistics about the digital divide show great variation in digital engagement
from country to country, age group participation, socio-economic class, urban or
rural living and, of course, between countries (Lerman 2013). Some interesting
2014 statistics about the global internet and social media use can be found at the
social media community blog hub (Kemp 2014). The risk here is that governments
may come to rely so much on big data that they forget to ensure that they engage
people to understand their needs and considered these during decision-making.
2.3.4 Talent Recruitment Issues
Given the scarcity of data analysts talent in the market, the public sector will have a
hard time attracting such talent as permanent stuff. In the UK, for example, public
sector organizations are obliged to recruit below market going rates to justify
Human Resource (HR) expenses (PageGroup 2014). In addition, the once upon the
time job security and fringe benefits of working for the public sector increasingly
disappear, making public sector organizations even less attractive. With large
organizations such as banks, insurances, large online retailers and consultancies
competing for such resources, governments will have a hard time attracting and
keeping such talent. On the other hand, governments could use University talent.
A resource much underutilized particularly in Europe, despite the depth and breadth
of skills relevant to the public sector (Campos 2008).
On the other hand, there is always training as an option to up skilling public
sector staff. In February 2014, for example, the UK government announcement that
£150,000 of government funding would be dedicated to Open data training of more
than 150 Public Sector employees (Gangadharan 2013).
2.4 Case Studies
Barcelona embarked on the smart city journey 10 years ago in an informal fashion
resulting in many smart city projects now dispersed in various departments across
the city, currently being collated under a single program. The 22@ Barcelona region,
once in need of redevelopment, has been transformed into a living test site for
piloting new technologies (Department for Business Innovation and Skills 2013).
Xavies Trias, mayor of Barcelona since 2011, has recognized the importance of
digital technologies for the future prosperity of the city. In his words in outlining his
commitment states that Barcelona “…should not waste the opportunity we have to
apply these new technologies to improving people’s quality of life, by generating a
new “economy of urban innovation” based around smart cities. This is another of
our future commitments” (Department for Business Innovation and Skills 2013).
To progress this agenda he formed Urban Habitat, a government wide man-
agement structure to promote collaboration across water, energy, human services
and environment agencies. Whereas, housing and urban planning were also
grouped together. To further cement cross agency collaboration, the Smart City
Personal Management Office, oversaw all projects with a smart city aspect. While
there are over 100 projects with a smart city angle, 13 are highlighted as strategic
for the smart future of Barcelona, tackling the necessary infrastructure to support
smart city applications, kit out city assets with intelligent sensors, and define smart
city public services. To this end, the telecommunications network is revamped to
integrate fiber optic networks, and Wi-Fi networking, public and a centralized
management system enabling the interoperability and prioritization of mobility,
public transport and urban infrastructure, applying concepts such as priority and
intermodality to make more efficient and sustainable mobility in cities. This is
underpinned with intelligent data project collating information from smart assets
and public service organizations with the view to opening these up to the public.
New public services are progressed such as energy projects relating to the urban
lighting of Barcelona, creating microgrids to create local generation and con-
sumption of green energy, telemanagement of irrigation urban green spaces and
electric car mobility options, as well as, smart parking options to enable speedy
parking avoiding unnecessary city traffic. Citizens will have contactless and mobile
apps to use city services. Some projects focus more generally on a mentality change
around smart city agendas. The O-Government project, for example, seeks to gain
support for Open Government, strategy and roadmaps and improve transparency,
2.4 Case Studies 37
open data and civic participation. The “Citizen compromise to sustainability

2012–2022” seeks to gain definition and traction for a city roadmap that can pro-
vide a more equitable, prosperous and self-sufficient environment to its people
(Department for Business Innovation and Skills 2013).
The regeneration of the 22@ Barcelona region was a public-private partnership
were companies, universities, research, and communities work in close proximity
with municipal leaders to exchange knowledge and streamline innovation, but also
ensure inviting and engaging urban planning by subsidizing housing and developing
green spaces. Local and international, private and public funding was used for
infrastructure development and the testing of new public services. The government
facilitated access to public funds by institutionalizing InnoActiva, a consultancy
agency which to supports private companies to make their case to public authorities
and institutions. Following the Silicon Valley cluster model, it is setting up clusters
in areas that they can develop a competitive advantage. Hence, 22@ is oriented
to attracting talent and expertise in Media, Information and Communication
Technologies, medical technologies, energy and design.
As to big data and analytics issues, open data is actually a core part of Barcelona’s
smart city event. Public and business access to information such as election results,
population, public facilities, or economy sits in a public repository called Open Data
BCN. Microsoft for example utilized data relating to a town festival called
“La Merce” as a pilot to providing improved crowd management solutions. To this
end, data feeds from social media, credit card transactions, web site visits, customer
service inquiries, GPS data, traffic status, weather data, and parking was collected
and analyzed. These data sought to gain insights about people’s perceptions of the
festival’s entertainment and food venues, citizen interests, people mobility patterns,
and medical and crime incidents that can help the planning and management of the
next event (Vienna University of Technology 2014).
The city also pilots the provision of services based on mobile identification
technologies. Through a smartphone app, citizens can access information about
parking tickets and car towing destinations, request public subsidies for nonprofit
activities, and the like, providing a proof of concept and of technology and getting
the necessary public engagement to move to the next level (Davenport and Prusak
1997). Consequently, public transport smart apps are ahead of their generation due
to popular demand.
In Barcelona transport information is based on a hyper-reality app. Anyone can
obtain information on bus stop locations, location, lines and even be directed to it
by simply pointing a smartphone camera in any direction, working wonders for
citizens new to an area, tourists and even blind people who can be oriented towards
their target destination using voice directions or Microsoft (2014).
Public engagement also manifest in developmental work for the smart city.
Sentilo, for example, is an open source sensor and actuator platform sponsored by
the Barcelona City Council, and designed by an open community.
Point of Attention: Barcelona’s open-source, smart city platform, engages

local talent in smart city development, ensures technology and provider
independence and data stewardship and remains with the public, under its
stewardship and safeguards civil liberties.
With the view to establish the city’s reputation as a smart city, Barcelona also
drives the Smart City Protocol initiative which seeks to connect global cities in pilot
projects to address common challenges (Bain and Sentilo 2014). Barcelona is a
small cosmopolitan city with the vision to grow and an exemplar for smart city
development that remains open, transparent, and democratic through an exchange
of all capital resources from capital and infrastructure to knowledge and talent.
While for most, smart city applications are still considered a nice-to-have feature
in our city life, crisis management is the acid test for any smart application. All
emergency services share a common requirement, when it comes to information
management. They need to accurately analyze life critical, real-time information
from diverse sources, in order to deploy and manage emergency service workflows
(Gangadharan 2013).
During the Haiti earthquake in 2010, emergency services needed to be dispatched
to the area to support the government cope with the circumstances. InStedd, a com-
pany specializing in technology design for emergency services such as natural
disasters and diseases, offered support to the emergency services and people. Within
48 h the company has set up the telecoms infrastructure and gain buy in for setting up
an emergency response number. The company offered a message-integrated com-
munications from two mobile network companies; incoming aid requests were
received in Haitian Creole. These were routed to Riff/EIS for analysis (Alehegn 2010).
Riff has the capability to automatically extract features, classify data and tag data
and their metadata (e.g. source and target geo-location, time, route of transmission)
and before it can process it via algorithms. The analytics module can detect rela-
tionships between these extracted features within a collaborative space or across
different collaborative spaces. Riff can also combine information from GeoChat, a
collaboration tool geolocating human comments, observations and reports to make
information richer and relevant. Riff then shared information with Crowdflower
another workflow provider, handling the distribution of tasks to a bilingual volunteer
workforce for translation, tagging, and geocoding. Information was then forwarded
to Ushahidi, a website initially developed to map reports of violence in Kenya now
turned a global crowdsourcing platform with humanitarian goals. ‘Ushahidians’, as a
community of interest helped to map, accurately geotag information to provide
accurate coordinates to the search and rescue team on the ground (Meier 2012).
Point of Attention: Smart apps and open collaboration platform can

become the critical infrastructure platform for the application of big data and
analytics to disaster recovery.
2.4 Case Studies 39
The value of swarm intelligence-based approaches for workflow-based emer-

gency management systems has been outlined as far back as 2007 (Bentley et al.
2014). This was an example par excellence for a bundle of crowdsourcing services
combined to provide an emergency response information architecture working with
the added complication of bilingualism. Data quality is of paramount importance to
prioritize calls and minimize erroneous dispatching of scarce rescue resources.
Timely and accurate information processing was life critical. International crowds
of volunteers were utilized and important safety critical decisions had to be taken on
the fly by the government and participating companies alike. The venture’s success
was based on companies’ technical capability and social responsibility, and
openness to collaboration with other companies and volunteers for the same cause.
Governments will have to find their feet and strike the right balance between
progress and the challenges of the big data era and redefine its relationship to the
public and to private capital in a world of global competition. Smart city has
become perhaps a pillar of competitive advantage for those who can grasp the
opportunity, while others will lag behind. As to these issues, in what follows we
point out some key factors for an effective application and exploitation of big data
and analytics in public sector digitalization, particularly, for smart cities and service
oriented initiatives.
2.5.1 Smart City Readiness
Each country will have to access the readiness of its cities to become and its
positioning in a smart cities global landscape. The European Smart Cities initia-
tives, audits cities on the basis of the six factors shown in Table 2.1 (Vienna
University of Technology 2014).
Smart city strategy at a national level is likely to be faced with budgetary
tensions between rural and urban development and a nation should have a vision
and a view of how to engage its people and private investors in the conversation.
Moreover, both local and central governments will have a ‘good cop, bad cop’ role
to play between role modeling the opening up information and effecting transpar-
ency and ensuring that information is safeguarded from abuse by involved parties.
In addition, smart cities will divulge responsibility for city services to machines,
partnerships with private companies and the public and this requires not only
educated citizens but also a change in mentality about civic responsibility from all
these parties.
Table 2.1 Factors for smart cities initiatives audit

Factors Description
Smart Innovative spirit and entrepreneurship, productivity, workforce flexibility,
economy ability to transform and international embeddedness
Smart mobility Local and international accessibility, availability of ICT infrastructure,
innovative transport systems
Smart Attractive natural conditions, pollution, environmental protection and
environment sustainable resource management
Smart people Level of qualification, affinity to life-long learning, social and ethnic
plurality, creativity, flexibility, cosmopolitanism and open-mindedness
Smart living Cultural facilities, health conditions, individual safety, housing quality,
education facilities, touristic attractively and social cohesion
Smart Participation in decision making, public and social services, transparent
Governance governance, political strategies and perspectives
2.5.2 Learn to Collaborate
Like most sociotechnical changes, to realize the benefits of big data and smart city
initiatives we need to change the way we do things. In particular, these technologies
require the diverse stakeholders collaborate. Government agencies and departments
have been traditionally separated by internal rivalries and financial competition
developing into an embedded silo mentality and culture. Information has been seen
as power and it has been hoarded to make people indispensable in the face of
downsizing, cost cutting, and other modernization attempts. Davenport and Prusak
highlighted such issue as far back as 1997 (Davenport and Prusak 1997).
Focus on efficiency and years of recruiting people and managers focusing on
cost cutting exercises have stripped the public sector from innovative human
resources and know-how and practices (Parker 2014). Public sector practices in
terms of renewing their staff and policies about paying their staff at the low end of
market prices makes it difficult for them to attract human talent, or indeed to
manage external associates who recruit such talent.
The public sector will need to rethink its internal recruitment processes to
employ smart people who will focus on creating successful partnerships with the
private sector and the public. If governments are to show the way of developing
smarter cities, they should orient themselves to attracting people with high quali-
fications, affinity to life-long learning, social and ethnic plurality, creativity, flexi-
bility, cosmopolitanism and open-mindedness. In addition, managing successful
collaborations will require new managerial skills. High partnering skills involve:
1. creating rapport via openness and self‐disclosure and feedback,

2. trust building through actions and words,
3. creative conflict resolution and problem solving,
4. appetite for change, and
5. welcoming interdependence (Dent 2006).
2.5 Recommendations for Organizations 41
Furthermore, partnering in the sphere of emerging technologies will also require

reviewing public sector procurement policies to allow wider participation in the
supplier pool and perhaps even participation of newly established technology
ventures that might be considered risky (Uyarra et al. 2014).
Finally, for Governments to continue to be relevant in a big data world, with
limited resources, they need to become smarter and this means fostering public
participation in decision making, public and social services, and making gover-
nance, political strategies and perspectives transparent and lean.
2.5.3 Civic Education and Online Democracy
The key aspirations of open government are the engagement of the public in the
political processes and their involvement in self-service public services. This will
require heightened levels of interest, knowledge, and maturity from the public, as
well as new modes of participation by governments. One means to achieve the
former is education. A United Nations review of such program in the US, showed
that such program had changed both people’s engagement levels and feeling of
adequacy to engage in the political process, but not people’s respect for different
political viewpoints, social cohesion and trust (United Nations Publication 2010).
Online participation in democratic processes can provide an affordable ways to
consult governments and take part in decision making in ways that it was not
possible before. Relevant initiatives spring up slowly in different countries. In
January 2014, California, for example, institutionalized the California Report Card,
mobile-friendly web-based platform that encourages citizens to engage in the
deliberative process via chat rooms where they would enter their own suggestions
but also rate others’ suggestions (Newsom and Goldberg 2014).
2.5.4 Legal Framework Development
Legal frameworks lag behind technological developments at all fronts of virtually

enabled living. The persistence and rise of Cyberbulling is a testament to that. Big
data profiling raises many issues regarding privacy, civil liberties, and equality, as
they were described above. With on demand public services via smart applications
entering the mix, such issues, particularly those of inclusion and exclusion from this
virtual world, can achieve another level of inequality. Thus, Governments need to
define new legal frameworks to regulate life and perhaps they even need to do so at a
global level, as internet engagement is a global phenomenon. Big data analytics can
be used as the tool to help international government bodies to analyze people’s
sentiments but also integrate best practices on such matters, but also to make law
more understandable by its law enforcement groups and the public (Morabito 2014).
2.6 Summary
This chapter discussed the impact of big data in the context of public service
provision and new opportunities for public service organization and structure that
may transform the role of governments in societies. We started our analysis by
discussing developments in public service provision, which treats citizens as
prosumers (proactive consumers) of public service delivery, moves towards direct
online democracy, and finally, to active engagement and a global smart megacities
competition for resources and talent.
In this context, governments seek to gain an advantage by utilizing a) new
sources of data, such as Crowdsourcing, Internet of Things, b) engage public talent,
c) institutionalize private–public partnerships and d) seeks for new models of value-
for-money public provision. Despite its potential, the adoption of big data and
analytics are not without challenges, particularly for central governments. Of par-
ticular interest are the challenges regarding data ownership, data quality, privacy,
civil liberties, and equality, as well as public sector’s ability to attract big data
analyst talent.
We showcased two case studies demonstrating how new forms of public service
provision. Barcelona Smart City provides an example par excellence of collabora-
tion between the private and public sector for regional redevelopment. Haiti’s
emergency support during the 2010 earthquake disaster demonstrates how big data
in the hands of passionate volunteers can organize and support with life-critical
emergency services, providing a life example as to what can be achieved through
the blend of human intuition and available big data integration and advanced
analytics. Like most sociotechnical changes, challenges reside in the social sphere of
technology acceptance and use, as well as with the regulation of such technology,
hence our recommendations are directed towards auditing readiness for Smart City
development, reskilling public servants with partnership management skills,
developing public’s mentality of civic participation and updating legal frameworks
to cope with developments in the big data area.
References
Alehegn, T.: Crisis mapping and collaboration between Western and African ICT developers for
Haiti quake. http://www.tadias.com/01/19/2010/crisis-mapping-and-collaboration-between-
western-and-african-ict-developers-for-haiti-quake-response/ (2010). Accessed 15 Nov 2014
Al-khouri, A.M.: Data ownership: who owns “my data”? Int. J. Manag. Inf. Technol. 2, 1–8 (2012)
Bain, M.: Sentilo—sensor and actuator platform for smart cities. https://joinup.ec.europa.eu/
community/eupl/document/sentilo-sensor-and-actuator-platform-smart-cities (2014). Accessed
15 Nov 2014
Bentley, R.A., O’Brien, M.J., Brock, W.: A: mapping collective behavior in the big-data era.
Behav. Brain Sci. 37, 63–76 (2014)
Brabham, D.C.: The public participation process for planning projects. Plan. Theor. 8, 242–262
(2009)
References 43
City Council: Mobile digital ID for Barcelona citizens. http://www.mobileid.cat/en (2014).

Accessed 15 Nov 2014
Clark, B.Y., Brudney, J.L., Jang, S.-G.: Coproduction of government services and the new
information technology: investigating the distributional biases. Public Adm. Rev. 73, 687–701
(2013)
CitySourced Inc.: Citysourced. http://www.citysourced.com (2014). Accessed 15 Nov 2014
Conservative Party: Building a big society. London, UK (2010)
Crawford, K.: The hidden biases in big data. http://blogs.hbr.org/2013/04/the-hidden-biases-in-
big-data/ (2013). Accessed 15 Nov 2014
Danova, T.: The internet of everything. http://www.businessinsider.com/the-internet-of-
everything-2014-slide-deck-sai-2014-2?op=1 (2014). Accessed 15 Nov 2014
Davenport, T.H., Prusak, L.: Information Ecology: Mastering the Information and Knowledge
Environment. Oxford University Press, New York (1997)
de Campos, A.: Literature on university-industry links: towards an integrated approach in the study
of influencing factors. Merit. Unu. Edu. 44, 0–32 (2008)
Dent, S.M.: Partnership Relationship Management (2006)
Department for business innovation and skills: global innovators: international case studies on
smart cities smart cities study (2013)
Evans-Cowley, J.: Crowdsourcing the curriculum: public participation in redesigning a planning
program. Soc. Sci. Res. Netw. Work. Pap. Ser. (2011)
Executive Office of the President of USA: Big data. Seizing Opportunities. Washington, D.C.
(2014)
Gangadharan, S.P.: How can big data be used for social good? http://www.theguardian.com/
sustainable-business/how-can-big-data-social-good (2013)
Gudipati, B.M., Rao, S., Mohan, N.D., Gajja, N.K.: Big data: testing approach to overcome quality
challenges structured testing technique. Infosys Labs Briefings 11, 65–73 (2013)
Hoffman, S., Podgurski, A.: The use and misuse of biomedical data : is bigger really better? Am.
J. Law Med. 39, 497–538 (2013)
Howarth, B.: Big data: how predictive analytics is taking over the public sector. http://www.
theguardian.com/technology/2014/jun/13/big-data-how-predictive-analytics-is-taking-over-
the-public-sector (2014)
IQ Analytics: big data training a priority for the public sector. http://www.itqanalytics.com/community/
blog/big-data-training-a-priority-for-the-public-sector#.VDJhABbILcs (2014). Accessed 15 Nov
2014
Junqué de Fortuny, E., Martens, D., Provost, F.: Predictive modeling with big data: is bigger really
better? Big Data 1, 215–226 (2013)
Kaggle Inc.: The home of data science. http://www.kaggle.com (2014). Accessed 15 Nov 2014
Kemp, S.: Social, Digital and mobile around the world. http://wearesocial.sg/blog/2014/01/social-
digital-mobile-2014/ (2014). Accessed 15 Nov 2014
Kerr, I., Earle, J.: Prediction, preemption, presumption how big data threatens big picture privacy.
Stanf. Law Rev. 66, 65 (2013)
Laney, D.: Who owns (really owns) “Big Data”. http://blogs.gartner.com/doug-laney/who-owns-
really-owns-big-data/ (2014). Accessed 15 Nov 2014
Lerman, J.: Big data and its exclusions. Stanford Law Rev. Online. 66, 55–63 (2013)
Manyika, J., Chui, M., Brown, B., Bughin, J., Dobbs, R., Roxburgh, C., Byers, A.H.: Big Data:
The Next Frontier for Innovation, Competition, and Productivity. McKinsey Global Institute,
Washington, DC (2011)
McDermott, P.: Building open government. Gov. Inf. Q. 27, 401–413 (2010)
Meier, P.: How crisis mapping saved lives in Haiti. http://newswatch.nationalgeographic.com/
2012/07/02/crisis-mapping-haiti (2012). Accessed 15 Nov 2014
Microsoft: City deploys big data bi solution to improve lives and create a smart-city template.
http://www.microsoft.com/casestudies/Microsoft-Excel-2010/City-of-Barcelona/City-Deploys-
Big-Data-BI-Solution-to-Improve-Lives-and-Create-a-Smart-City-Template/710000003415
(2014). Accessed 15 Nov 2014
Miller, H.E.: Big-data in cloud computing: a taxonomy of risks. Inf. Res. 18, 571 (2013)
Mostashari, A., Arnold, F., Maurer, M., Wade, J.: Citizens as sensors: the cognitive city paradigm.
International Conference & Expo on Emerging Technologies for a Smarter World, pp. 1–5.
IEEE (2011)
Nagy-Rothengass, M.: European Activities in the Area of Big Data. META-Forum, p. 27. Berlin (2013)
Nelson, S., Milberry, K., Richardson, J., Martin, M., Thomson, K.: PartyX. http://partyx.ca/
(2015). Accessed 8 Jan 2015
Newsom, G., Goldberg, K.: Let’s amplify California’s collective intelligence. http://citris-uc.org/
lets-amplify-californias-collective-intelligence-op-ed-ken-goldberg-gavin-newsom-california-
report-card/ (2014). Accessed 15 Nov 2014
OGP: Open Government Partnership. http://www.opengovpartnership.org (2014). Accessed 15
Nov 2014
PageGroup: Public sector salary survey: insight and trends. http://www.michaelpage.co.uk/salary-
survey/public-sector.htm#salary-results (2014). Accessed 15 Nov 2014
Parker, L.: Cuts mean a smaller public service must be smarter. http://www.governmentnews.com.
au/2014/06/cuts-mean-smaller-public-service-must-smarter (2014). Accessed 15 Nov 2014
Pinsent Masons LLP: Calls for EU public private “Big Data” partnership. http://www.out-law.
com/articles/2013/november/calls-for-eu-public-private-big-data-partnership/ (2013). Accessed
15 Nov 2014
SeeClickFix: Report neighborhood issues and see them get fixed. www.SeeClickFix.com (2014)
Stefan, R., Kisker, H.: Sizing the cloud. http://www.forrester.com/Sizing+The+Cloud/fulltext/-/E-
RES58161 (2011). Accessed 15 Nov 2014
The National Cycling Charity: Fill that hole. http://www.fillthathole.org.uk (2014). Accessed 15
Nov 2014
The Weather Channel Inc: Weather underground. http://www.wunderground.com (2014). Accessed 15
Nov 2014
Thomas, J.C.: Citizen, customer, partner: what should be the role of the public in public
management. Public Adm. Rev. 73, 786–796 (2013)
United Nations Department of Economic and Social Affairs: World Urbanization Prospects: The
2005 Revision (2006)
United Nations Publication: e-Government and New Technologies: Towards better citizen
engagement for development. Geneva, Switzerland (2010)
URBACT: Smart cities: Citizen innovation in smart cities. http://urbact.eu/en/projects/innovation-
creativity/smart-cities/homepage/ (2012). Accessed 15 Nov 2014
USGS: DYFI background—the science behind the maps. http://earthquake.usgs.gov/research/dyfi/
(2014). Accessed 15 Nov 2014
Uyarra, E., Edler, J., Garcia-Estevez, J.: Barriers to innovation through public procurement: a supplier
perspective. Technovation 34(10) 631–645 (2014). doi: 10.1016/j.technovation.2014.04.003
Vicini, S., Sanna, A.: How to co-create Internet of things-enabled services for smarter cities. The
First International Conference on Smart Systems, Devices and Technologies, pp. 55–61.
Stuttgart, Germany (2012)
Vienna University of Technology: European smart cities 3.0. http://www.smart-cities.eu/ (2014).
Warner, J.: Next steps in e-government. In: Proceedings of the 12th Annual International Digital
Government Research Conference on Digital Government Innovation in Challenging Times—
dg.o ’11, p. 177. ACM Press, New York, USA (2011)
References 45
Yiu, C.: The big data opportunity-making government faster, smarter and more personal, Policy
exchange. London, UK (2012)
Zhang, L., Mitton, N.: Advanced Internet of things. 2011 International Conference on Internet of
Things and 4th International Conference on Cyber, Physical and Social Computing, pp. 1–8
(2011)
Zhang, W., Chen, Q.: From E-government to C-government via cloud computing. 2010
International Conference on E-Business and E-Government, pp. 679–682. IEEE (2010)
Big Data and Education: Massive
Digital Education Systems 3
Abstract
This chapter discusses how massive digital education systems like MOOCs,
facilitate the distance learning aspects of formal education institutions and enable
a peer-to-peer learning. It discusses how MOOCs open new income streams for
traditional institutions through employment recruiting services, syndication, and
sponsoring, as well as by advertising income, selling student information to
potential employers or advertisers. To this end, this chapter first explains
MOOC educational models (peer-to-peer and institutional) and the role that big
data and analytics are playing in this context, highlighting institutional
advantages, opportunities and challenges. It then explains two cases studies
where big data and analytics played essential roles in the design and delivery of
the curricula.
3.1 Introduction
Distance learning has always been designed for those who did not have access to
formal learning institutions. Open Universities around the world have made such
markets their primary focus. Pedagogical material was adapted to reach students
with different socioeconomic backgrounds, learning habits, and studying patterns,
at remote areas or even close to home but limited in free time or incompatible
working schedules with those of institutions. Originally distance learning programs
employed a variety of audiovisual, information dissemination means, such as
television programs, CDs and the like but there was still lacking student-tutor and
peer-group interaction. It was really online social platforms that could re-instill the
social aspect of learning (Marques 2013).
Over the past 20 years, an academic revolution has taken place, unprecedented in
their global scope and diversity of institutions and people they affect, due to
advancements in information and communications technologies (ICTs) (Executive
2014; Yuan and Powell 2013). Education is a key factor to economic development
for nations and social mobility for the individual. Two factors have accentuated the

DOI 10.1007/978-3-319-10665-6_3
48 3 Big Data and Education: Massive Digital Education Systems
need for global demand in education. The skills currently required to compete in a
global environment are changing due to shifts in the economic model and com-
petition from emerging developing markets. ICT’s have given nations and insti-
tutions the opportunity to resource highly skilled people anywhere in the world and
individuals the opportunity to move up socially. This journey is not without
challenges. Developing countries for example have low digital technology pene-
tration. Africa, the Middle East, and Latin America/Caribbean constitute just
17.2 % of the world’s Internet users (Miniwatts Marketing Group 2014). Basic
technology education and affordability hold people back from developing necessary
technological skills. On the other hand, initiatives such as Negroponte’s One
Laptop Per Child (OLPC) project which sought to provide each child in the
development world with a low-budget computer slowly change this dynamic.
The use of ICT technologies in education has come under different guises.
Guri-Rosenblit (2010) mentions more than 20 different relevant terms, each
connoting slight different implementations of a fundamentally same thing
ICT-enabled distance learning:
• Internet mediated teaching,

• technology-enhanced learning,
• web-based education,
• online education,
• computer-mediated communication (CMC),
• telematics environments,
• e-learning,
• virtual classrooms,
• I-Campus,
• electronic communication,
• information and communication technologies (ICTs),
• cyberspace learning environments,
• computer-driven interactive communication,
• open and distance learning (ODL),
• distributed learning,
• blended courses,
• electronic course materials,
• hybrid courses,
• digital education,
• mobile learning,
• technology-enhanced learning.
Massive Open Online Courses (MOOCs) have recently come to be added to this
long list of different implementations of ICT-enabled distance learning. While
MOOCs facilitated the distance learning aspects of formal education institutions,
they also introduced an even more fundamental educational change; they opened up
participation to education by enabling a peer-to-peer learning. They offer the
possibility to thousands of people to voluntarily share knowledge with each other
3.1 Introduction 49
through open access to courses. Hence, they practically provide to anyone

interested fast-track, accessible, flexible, free or affordable courses, at different
levels from higher education to vocational training courses. Not only the courses
themselves add to the content variety of big data, but big data can have a leading
role to play in the production and delivery of MOOCs educational material.
In doing so, they are a key game changer in the education sector with pervasive
global implications, about nations’ economic capabilities, international relations
and global social mobility and welfare.
This chapter begins by discussing the implications of MOOCs in the educational
sphere and the role of big data in the design and implementation of MOOCs. It is
structured as follows.
3.1.1 From Institutionalized Education to MOOCs
MOOCs were an offshoot of formal education. MOOCs is rooted within the

ideology of open access to education without demographic, economic, and
geographical constraints which dates back in the early 20th century (Yuan and
Powell 2013). Figure 3.1 outlines the historical development of MOOCs thus far.
In 2001, MIT established the Open Courseware initiative (Massachusetts Insti-
tute of Technology 2014). The priority was to disseminate knowledge far and wide
where people can interact to learn and support each other in their journey to
knowledge. By 2007, MIT has put all their courses online and available to the
public. In the original announcement of the program to New York Times, MIT’s
president Dr Vest had predicted
2000-07 2008 2009 2010 2011 2012 2013
Open Future
Source Learn
Software
Connectivist
MOOC
cMOOC Udacity
Open Coursera
Content
Stanford
MOOC
(xMOOC)
MIT
edX
An influence
Directly related
Fig. 3.1 MOOCs and open education timeline. adapted from Yuan and Powell (2013)
I also suspect, in this country and throughout the world, a lot of really bright, precocious
high school students will find this a great playground. […] there will probably be a lot of
uses that will really surprise us and that we can’t really predict (Goldberg 2001).
Indeed, open courseware was a stepping stone to a much larger scale change.
Since then we have a proliferation of MOOC platforms and a number of highly
regarded institutions that have embarked on the very same journey. Coursera
(2014), for example, established in 2013 is a partnership between Stanford
University and 61 well-regarded universities. EdX (2014) is established by
Massachusetts Institute of Technology, École Polytechnique Fédérale de Lausanne,
The Hong Kong University of Science and Technology. Udacity (2014),
Peer 2 Peer (P2P) University, and FutureLearn (the UK Open University’s
MOOC platform) are other related platforms are just but a few other offshoots of
institution-led MOOCs.
However a number of community-based, open learning platforms have started to
pop-up with significant differences in their pedagogic approach and philosophy.
Peer 2 Peer University (2014), for example, is a grassroots initiative. It creates a
model for lifelong learning by facilitating learners to interrogate Internet material to
co-create their curriculum and in doing so taking responsibility for their own and
their teams learning.
According to the website of Peer 2 Peer University, openness, community and
peer learning are the key drivers and nothing falls beyond the agency of the learner.
Even the technology and process of learning can be open to change should the
learners wish to. The idea is to leave behind the hierarchical and role separation of
academics and students. Materials such as readings, videos and panel discussions
seek to provoke discovery, rather than transmit knowledge. They are designed in
self-contained modules to be consumed on demand, rather than follow a discipline.
Their motto stir away from institutional conventions, to create an environment of
“No judgment” where you “Put in what you want” and “Take what you can” an
environment where educators and learners accept that they will experiment and in
doing so “[they] will take risks. [they] will learn together. Things will be messy.
Things will be fun.” (Berger et al. 2014).
Other courses, such as learning a programming language might need more
direction and structure from experts, but emphasis is still on interaction, self-
learning responsibility and self-pacing. Udemy (2014) is another online learning
platform, hosting a variety of courses, from practical ones such as photography,
languages, and test preparation to more academic ones, such as philosophy,
humanities and science. Courses are not free, but affordable. Udemy hosts 16,000
such courses and engages 3million students. It keeps 50 % of revenues from each
course in exchange for marketing and technical support through the platform.
The differences between MOOCs business models are significant and so is the
invest interest in them, as Fig. 3.2 suggests, presenting the funding models of the
key MOOCs players in the market nowadays.
MOOCs are in search for business model validity and sustainability. For
Coursera for example, the business principles resemble those of Amazon. Create a
3.1 Introduction 51
VENTURE
NONPROFIT
CAPITAL
WORD
Andreessen Horowitz
Bill & Melind
New Enterprise Gates Foundation
Associates
Kleiner Perkins
Caufield & Byers
MOOCs
UNIVERSITIES
Hardvard
COMPANIES
Pearson MIT
Google University of
Pennsulvania
CalTech
Fig. 3.2 MOOCs platforms and investor interests. Adapted from (Davidson 2014)
hub of suppliers to attract a large number of students who will be attracted in

finding whatever they want in one place. For others, such as Simon Nelson of
FutureLearn (Wilby 2014), the name of the game remains to be seen, as the
potential of the technology has not been envisioned let alone realized. Udacity
focuses on employment and employ training teaming up with companies such as
Google, AT&T to offer degrees that are of interest (Udacity 2014). EdX is steering
its model away from course provision to becoming the manager of the technology
platform hosting MOOCs, and providing their platform management expertise as a
service to universities.
Not only is their market focus different, but so is their funding models too.
Venture Capitalists, philanthropists and national and educational institutions are all
happy to part with their funds to support the MOOCs agenda. Coursera (2014) has
accepted some $65 million, while Udacity (2014) another $15 million from venture
capitalists, and this bears a risk in terms of the openness of their business models as
their investors may demand the commercialization of courses for business profit
(Watters 2012; The Economist explains 2014).
3.2 MOOC Educational Model Clusters
3.2.1 University-Led MOOCs
University led MOOCs, also characterized as xMOOCs, represent the latest inno-
vation in formal education. Accredited institutions are now accepting MOOCs as
part of a degree, yet a wholly free online degree is not in effect. Tuition fees are
driven down because, while students still have to pay to certify for their degree,
they don’t pay for the operational costs of running the academic process, such as
supervision expenses, classroom and facilities maintenance and the like. When
MOOCs are able to provide free accredited online degrees, the University business
model will have fundamentally changed (Dellarocas 2013).
A feasible economic option for universities is virtualizing traditional teaching, by
recording of live lectures and presentations. Nothing new here e-Lectures are then
delivered over the Web using a Learning Management System (LMS). While
attendance levels may drop, student performance is not necessarily compromised
(Ottmann 2013). In addition, reaching student populations around the globe is not
only flattering for academics. but also a great way to decrease their face to face
teaching load. MOOCs are certainly attractive to both academics and university
management due to their ability to attract venture capital and grant money, often
linked to academic reward systems. Another source of potential income is advertising.
This is perhaps the reason why prestigious institutions have joined forces to offer their
modules and degrees from a single platform. The likes of Harvard, MIT, Cornell,
Berkeley, and others founded the non-profit edX to offer the option to “…take great
courses from the world’s best universities” or to “Take the world’s best courses online
for free” (Ottmann 2013). After all, there seems to be a gap in the market for low cost
educational provision. According to Laurillard et al. (2010), professor of learning
with digital technologies at the Institute of Education, University of London, student
loans are higher than credit card loans in the US, while 40 % of student debt will never
be repaid in the UK, and by 2025 demand will double to approximately 200 million
students per year as emerging economies demand access to higher education.
Perhaps this also indicates a deeper shift away from government budgets to more
self-sustainable forms of funding or to better links between tertiary education and
private capital, or even perhaps a shift in the mission and student reach of uni-
versities to exploit new technologies in order to educate the global population and,
in that way, create common worldviews.
The same process of student-expert can be observed in MOOCs with more
practical content where professionals or experts provide training and skills
development to novices. Many of the courses in Coursera (2014), are of this type.
3.2.2 Peer-to-Peer MOOCs
University-led, xMOOCs differ from peer-to-peer driven MOOCs, also referred to

as cMOOCs [i.e. connectivist MOOCs (Yuan and Powell 2013)]. They differ in
philosophy, delivery and intention and a certain ‘war’ of learning ideology and
practice is currently played out. The quickest way perhaps to understand cMOOCs
is to juxtapose them against their more traditional counterparts. Table 3.1 outlines
some of the basic differences.
3.2 MOOC Educational Model Clusters 53
Table 3.1 xMOOCs versus cMOOCs

xMOOCs cMOOCS
What is the ultimate Deliver content to larger/new Create a learning community,
goal? audiences; provide which can continue to learn
certifications; experiment together in the future
with new course; increase
access to top Northern
American universities or
provide free access to
education
What learning or Traditional teacher-centered Community approach: the
instructional theories approach: knowledge is learning process focuses on
are informing the transmitted from the collaborations between learners
instructor’s decisions? instructor to the student
What is the role of the Creates course content, Participates and facilitates other
instructor? pedagogical approach and learners to create content,
assessments learning goals, generate new
knowledge, etc.
What role does the The learner receives The learner co-creates the
learner play? knowledge, participates in pedagogical approach and
small work groups, and assessment
responds to tests and
assessments
How are learners Learners view content Learners create projects relating
building new developed by the instructor to course themes; share
knowledge? and apply it to problems or information and knowledge; help
projects defined by the each learn
instructor
How is learning Learners complete Learners share and self-assess
assessed? assessments that evaluate their learning paths
comprehension from the
instructor’s view
Who is creating the The instructor creates the The content goal is created by a
content? content core group of learners; content is
created by all participants
What types of Learners work in small Learners research and share
interactions are taking groups to complete relevant course content to achieve
place? assessments based on learning goal
provide course content
How flexible are the Learning path is set in Learning goals broadly set by
course path and the syllabus and curriculum core participants during the
course goals? planning material course, in response to the
community, on a week-by-week
basis
Source Crowley et al. (2013)
Instead of relying on traditional education models, P2P or cMOOCs seek to

motivate learning collaborations amongst students. In some ways, students are not
students in that there are not teachers either. cMOOCS are communities of learning
of like-minded people taking responsibility for and orchestrating their own learning
process. In doing so, they co-define their learning objectives, create content through
research, share themes, give feedback, insights, ideas and support. While a core
group tends to take more of the coordination activities, course goals and objectives
are fluid and determined in response to the community.
The idea of learning through research is not foreign to formal education. There
are plenty of Masters in Research usually a stepping stone to a PhD, and, of course,
Doctorate programs, the pinnacle of formal education, all using research as the
means or path to achieving knowledge. However, achieving a degree is a solitary
process where the student takes accountability for his or her own research to which
they present to some academic authority in exchange for a certification of their
ability. cMOOCs so far do not work like this, at least not yet. Why is this important
from a big data point of view? For starts, as there is no curriculum, information
from any source can be combined to improve learning. Anything from government
information to data generated by the internet of things, social media, academic
journals, observations from virtual games, countries statistics bureaus can be
blended, or in the speak of big data, mashed to create new knowledge. Such
blending might be criticized or supported by one’s learning group but inevitably
there is no academic authority to bind that search to specific sources, ideas or
frameworks. The world of internet is the learners’ oyster (Yeager et al. 2013).
At the moment, most cMOOcs are teachers and academics interested in learning
per se. P2PU (2014) is such an institution, where a team of experts has come
together to explore new ideas in curriculum development and design, research,
knowledge sharing, etc. and by doing so develop our knowledge about learning!
Will such form go mainstream? It is remained to be seen.
3.3 The Role of Big Data and Analytics
This section seeks to understand the role of big data in this new field of education. It
should be evident to the reader by now that cMOOCs make use of mashing dis-
parate sources of open information, whether big or small, to discover and create
new knowledge. But what is the role big data in xMOOCs?
Much like any other business, MOOCs need to develop market strategies for
attracting new students. Hence, big data can help with the market analytics within
the academic sphere. Indeed, such market research services for academic institu-
tions has now reserved its own term: Academic analytics (García and Secades
2013). Academic Analytics utilizes large data sets, statistical techniques, and pre-
dictive modeling to offer business intelligence to academic institutions to improve
on their customer experience. For example, education systems, especially MOOCs
with varying degrees of student engagement and completion rate, will require a
3.3 The Role of Big Data and Analytics 55
careful management of resources as well as seek cross selling opportunities of more

premium products. This will be particularly true as xMOOCs will try to capitalize
on their large student basis to satisfy their investors. Such predictive analytics are
likely to be integrated with workflow systems to automate or at least semi-automate
administration processes. For example, academic analytics can also inform a
decision-support system and a workflow system to automate the admissions
process. They can also be used to support students throughout the learning process.
Utilizing existing knowledge and models relating to student effort and success, they
can monitor students’ interaction with the system, alerting administrators when
student engagement patterns change in order to initiate communication with the
student, or indeed send a predetermined default email communication (Campbell
et al. 2007).
Another more fundamental use of big data analytics is its use as a learning tool.
This particular use of big data analytics has been defined as Learning Analytics
(Fournier et al. 2011). Learning analytics utilize individual learner-produced data
against known models about successful learner interactions and behaviors. People
in the learning analytics community propose six critical dimensions, both soft and
hard, along for designing such learning models. Soft issues relate to assumptions
about humans or the society, e.g., competences or ethics. Hard challenges on the
other hand relate to data and algorithms. In that sense, Learning analytics can be
defined as
…the area of discovery analytics which model of intelligent and learner-produced data to
discover information and social connections for predicting and advising people’s learning.
(Siemens 2010).
Such analytics have appeal to a number of educational stakeholders. For

example, individual learners can be assisted in reflecting on their achievements and
patterns of behavior in relation to others, academics can have early warnings about
students requiring extra support and attention and plan relevant interventions;
management and course leaders can be assisted in developing and marketing
attractive academic programs and curricula and promote them (Campbell et al.
2007).
3.4 Institutional Advantages and Opportunities

from MOOCs
According to Global Industry Analysts (Report and asp 2014), the global e-learning
market will reach $107 billion by 2015. Inevitably institutions want to leverage
their operations to capture market share. MOOCs open new income streams for
traditional institutions by not only relying on state education subsidies and student
tuition, but through employment recruiting services, syndication, and sponsoring, as
well as by advertising income, selling student information to potential employers or
advertisers (Report and asp 2014).
Traditional institutions, particularly universities, aspire to transform higher

education by expanding academic access on an unprecedented scale as means to cut
costs in already underfunded public universities. While others propose various
forms of ‘pay-as-you-learn’ teaching provision. One school of thought proposes
that MOOCs was an experiment that has served their purpose as a driver for
institutions to take a more strategic approach to online learning, which create new
opportunities for universities, by
• Utilizing openness as an online learning approach through the use of online

communities, including models for scalable provision that may generate
revenues, and goes beyond institutional boundaries.
• Inventing new Business models, such as applying the concepts of freemium and
premium offers into online learning, providing institutions with new ways of
thinking about marketing and income generation.
• Enabling Service Disaggregation that includes unbundling and re-bundling of
courses and delivery related services to offer premium educational services such
as paying for assessment and/or teaching support.
• Experimenting with new approaches for teaching and learning, by utilizing new
technologies.
• Creating new education models that cater for Learners with various studying
patterns across the globe.
• Developing internal capability by reviewing their technical infrastructure,
academic and support staff working practices (Yuan et al. 2014).
It is also important for academic institutions at all levels to increase learning

attainment without increasing their costs. Learning Analytics, along with educa-
tional data mining and teaching analytics can all be seen as three aspects of the
same solution to raising the education standards of the youth without necessarily
increasing the number of educators required, thus making institutions more cost
efficient. Learning analytics focus on capturing student behavior and correlating it
to achieving learning objectives, educational data mining seek to design predictive
analytics models for student attainment while teaching analytics helps educator
translate such findings into better course design and student support procedures and
interventions. As time goes by more and more online tools are being developed in
this area. According to Charlton and Mavrikis (2013), the Course Resource
Appraisal Model (CRAM), for example, allows educators to understand the impact
of different delivery modes (face-to-face, blended learning and larger-scale
delivery) on student learning and hence ease off transition or improve course
design. The authors (Charlton and Mavrikis 2013) also argue that the Maths-Whizz
can improve maths scores by online tutoring students and assessing score levels.
BlikBook focuses on modeling class engagement and substitutes lecturers’ student
support engagement by assisting students find relevant material. Table 3.2 provides
a brief overview of some of the most commonly used learning analytic tools
currently available and their impact on learning as presented in Charlton and
Mavrikis (2013).
3.4 Institutional Advantages and Opportunities from MOOCs 57
Table 3.2 Tools and their data

Tools and Learning analytics Educational data Teaching analytics
platforms perspective mining perspective perspective
Jenzabar Strong analytics overall Learning analytics and The tool provides
and systemic approach statistical models are visual representations
using large-scale data used to create to lecturers and support
from across the predictive analysis staff to inform which
organization students need support
and why
CRAM Provides visual Classification system Automated
(course representation of and relationships to visual models of the
resource teaching support activities impact on teaching and
appraisal across teaching and learning
model) learning analysis
Maths Domain-specific Uses analysis Uses a visual model to
whiz analysis that enables clustering models of provide information of
different types of insight static relative learning the teachers to use as
into the students’ points to determine part of their teaching
learning and relative maths age and
learning of maths contextual data
BlikBook Social peer group Student modeling and Student feedback to the
approach for exchange profiles for community teachers and more peer-
and group problem use led student groups
solving
MiGen Social grouping and Rule-based analysis Constant visual
outcome predictions of and other analytics of teaching
students computational and learning for
techniques to analyze teachers to instigate
task progression immediate intervention
on task progression and
collaborative activities
Adapted from Charlton and Mavrikis (2013)
A key issue that all educational systems have to solve in society today is young
people’s motivation to learn. Like many other fields, education may need to seek
outside their conventional boundaries for solutions. One area of interest for moti-
vating and maintaining student engagement may come from gamification, which
utilized sound game theory and game mechanics, as also mentioned in chapter one.
It is a potentially powerful method for driving user engagement, and not very far
from educational simulations (see for example, virtonomics (2014)).
3.5 Institutional Challenges from MOOCs
As it stands, MOOCs have a fundamental problem of identity and therefore issues

around their monetization efforts. Most MOOC start-ups do not appear to have clear
business models and like most ventures nowadays strive for achieving market share
and worry about profitability later. This is particularly true of xMOOCs, such as
Coursera, who seek to increase throughput. On the basis of registration numbers,

xMOOCs seek revenues for certification fees; recruitment services, sponsorships.
Yet, many institutions do not trust the quality of MOOC delivery yet; at least not
enough to offer their institutional degrees. Thus, MOOCs are still seen as an
interesting branding and marketing activity at present (Yuan 2013).
Indeed the quality of learning through MOOCs is uncertain. To start with there is
high dropout rate, often reported as high as 90 % (Liyanagunawardena et al. 2013).
For traditional institutions this rate would be alarming, if not fatal for their going
concern. Some suggest, however, that the same yardsticks for evaluating MOOCs
cannot hold true, simply due to the nature and stage of adoption of the new
technology.
Moreover quality issues need to be overcome, as is a certain attachment to ‘old
ways’ of teaching and old ways of evaluating teaching. Nevertheless, certain
incidents may harm the relationship between MOOCs, academics, and students. In
February 2013, for example, the Coursera/Georgia Tech course “Fundamentals of
Online Education” was cancelled, making students lose all their online contribu-
tions (Morrison 2013). In the same year, UC Irvine professor Richard McKenzie
abandoned his economics MOOC course in the middle due to disagreement over
how to best conduct the course. Michigan professor, Gautam Kaul caused waves
with his Coursera-run finance class by refusing to give students the correct answers
to assignments, “to avoid preparing new sets of questions with multiple versions to
allow [students] to attempt each one more than once. Handing out answers will
force us to do that.” Karen Head, Georgia Institute of Technology professor, raised
concerns around the technical functionality of MOOCs to serve her pedagogical
requirements (Watters 2013). Quality also suffers when student success rates are
measured. For example, while 74 % of students in traditional classes passed, only
51 % of those attended Udacity’s MOOCs equivalent program succeeded. To
address quality issues, for example, Udacity had turned to corporate training
(Watters 2013).
Another threat remains hidden in the potential success of MOOCs; that of
pedagogical hegemony. When and if education is concentrated in the hands of few
global institutions and driven by few academics who are able to reach out to large
numbers of students globally, then we might be creating a world where knowledge
is founded on the same principles, ideas and viewpoints, in the exclusion of others.
This will not only compromise the financial going concern of institutions but the
plurality of knowledge bases and hence the democratization of ideas. This is of
course not an issue for collaborative learning environments such as cMOOCs, as
the learning agenda is not controlled by an authority but co-created on the way by
participants.
cMOOCs, despite being a fundamental departure from existing, instructor-led,
educational models, and indeed better positioned to use big data as a tool for
knowledge discovery and integration, face their own challenges. They require a
degree of student maturity, motivation and responsibility in order for students to
take responsibility for their own learning and for that of others, thus, undertaking
the level of research and knowledge sharing required and expected by others. This
3.5 Institutional Challenges from MOOCs 59
may not be feasible at all ages or at least a different kind of socialization and
upbringing must precede it for primary and high school students to be able to
engage at that level, and national curricula mandate must be able to open up to
become compatible. Even in adult education context, people may find themselves in
the midst of difficult group dynamics such as conflict, different levels of commit-
ment. Certain skills need to be developed such as conflict resolution, team building,
and creative thinking.
Moreover, even web search is biased. The topology of the Web allows us to see
only few of the billion documents on a topic, and even when they find less popular
information they need the critical skills to evaluate it (Barabási 2003). Boyd (2010)
emphasized that information brokers, such as Google, filter information without
necessarily taking a critical point of view, and hoped that social media can play a
mediating role to balance this out (Kop 2011).
A new perspective seems to emerge, seeking the integration of instruction-led
modes of learning with more social forms of peer to peer learning to provide a more
holistic experience (see Fig. 3.3) (Crosslin 2014). Such solution might be a safer
option, but lacks imagination and still recycles old notions of educating.
Perhaps educators need to look into other aspects of social life to design
educational products. We also need to start differentiating MOOCs based on their
audience and purpose, which might be a vicious cycle as people have not decided
on the business model for MOOCs.
Dual Layer cMOOC / xMOOC
Form Group Work:

Groups Connectivist projects that are MOOC
cMOOC heavily student -centered . Completion
(ongoing)
Designed for problem -based
learning.
Orientation: Students or Al Students or Al

determines need determines need
Students or Al for instructivist for connectvist
Students determine modules interactions
which path to
take
Modules:
xMOOC Instructivist modules that are hevily instructor - centered.
Designed for direct reified knowledge transfer.
Fig. 3.3 Integrating cMOOC and xMOOC best practices. Adapted from Crosslin (2014)
3.6 Case Studies
HarvardX was formally launched in October 2012 and is co-hosted in EdX platform
along with MIT, Duke University and other institutions. Since its launch it had
1,331,043 registrants in 195 countries with various degrees of certification attain-
ment (Harvard University 2014). In its first year alone, over 500,000 students have
registered for HarvardX courses, which is larger than the number of Harvard
graduates in its 377-year history. Aside US that accounts for 36 % of enrolments,
the rest of its student cohort comes from 204 different countries. Amongst them,
Nigeria is the most enrolled country of Africa, Spain is the first in Europe, India is
the highest-enrolled country in Asia with almost 50,000 students, and Brazil in
South America. A potential barrier to global participation in enrolment may be
language as most students from global HarvardX enrolment comes from English-
speaking countries (Harvard University 2014). In China, for example, only 23,548
students enrolled in spite of its huge population of 1.3 billion. Interestingly certi-
fication attainment data do not follow the pattern of enrolment. Despite of
accounting for less than 2 % of enrolment each, Burkina Faso, Greece, and Georgia
are the countries with the highest rates of certificate attainment, indicating perhaps
differences in the culture, educational structure and conditions or personal moti-
vations for attending the course. However, certification is more likely among
registrants who enroll near the launch dates, but viewing likelihood is stable
through the run of the courses, indicating perhaps a premeditation on achieving a
particular educational goal (Harvard University 2014).
It is important also to note that the median age of students is 28 and most of them
already have a Bachelors or a Master’s degree. So HarvardX course seems to act as
an educational ‘top-up’, rather than a university substitute, and appeal more to the
already educated, mature university students as opposed to the uneducated masses
its proponents have been advertising.
Point of Attention: MOOCs are not substitutes for conventional universities

or other learning institutions. However, the use of big data analytics both in
designing curricula and in delivering MOOCs can provide new modes for
learning, merited their own attention.
The ultimate question is “Do MOOCs work?” Harvard representatives answer the
question with a poignant clarification: “Work for what?” (Summary et al. 2013). So
far, the presumption that MOOCs will enhance, replace and disrupt existing models
of higher education, was based on the assumption that students will utilize MOOCS
as a substitute for conventional education. They haven’t. MOOCs seem to be open,
virtual spaces for learning experimentation. For example, registrant activity within
courses is diverse. Certification attainment is one possible learning pathway and
there is an indication that they might benefit from synchronous course schedules and
the cohorts that they build. Most, simply watch videos or read text. Some registrants
3.6 Case Studies 61
dip into a couple of sessions in the first two weeks and then abandon it either to
register to other courses or sign up properly on a later instance of the course. Others
focus on self-assessment to test themselves (Summary et al. 2013).
MOOCs success was allegedly based on their adherence to well-established
educational and learning models. For these models to be relevant, however to the
new context, similar motivations, interactions, norms, expectations, should be
present; something not necessarily true when comparing MOOCs to conventional
education settings. For example, open enrolment periods and unrestricted use of
course resources raise important questions for analysis and design. Perhaps, we
need to examine MOOCs in their own right and not as substitute of universities or
other institutions. To this end, new metrics, far beyond grades and course certifi-
cation, would be necessary to capture the diversity of usage patterns and goals
(Summary et al. 2013).
The second case study we discuss in this Chapter is the one of Livemocha,
which is a free online language learning community, providing instructional
materials in 38 languages and a platform for speakers to interact with and help each
other learn new languages. It has approximately 12 million registered members
from 196 countries, with over 400,000 users visiting the site daily (Livemocha
2014). It was founded in 2007, by Shirish Nadkarni and Krishnan Seshadrinathan
and is based on the idea that you best learn a language when you interact with
native speakers. By 2010, it was announced as one of the top 50 websites by,
the New York Times and the Financial Times, and Time magazine (2010).
Livemocha’s online social language learning platform has followed a classic
disruptive innovation path through business model innovation, to transfer online the
process of learning a foreign language in a foreign country. The business model was
simple. Engage native speakers who want to learn another language for free and are
willing to help others to learn their own language; hence, native speakers alternate
between being students and tutors. Offer a platform of tools and techniques (such as
flashcards, quizzes) to help students with learning and memorizing. Maintain fair
access and reciprocity in the community with the exchange of virtual tokens and
rules often enjoyed by time banks and reciprocal feedback on both student per-
formance and the quality of the reviewer (Gartner 2013).
Point of Attention: Technology-enabled learning communities will generate

new types of unstructured big data relating to education and learning. New
norms in community education can provide new principles to model learning
analytics, based on principles of reciprocity and mutual benefit and gamification.
Income comes mainly from advertising, with approximately 700,000 recurring

users each month and approximately 1.85 million users reached within the first
15 months from its launch. In the next phase, Livemocha paired with Pearsons to
provide a premium product that bundled formal language learning methods with
social media interaction for a fee. The community remained loyal to its customer
base, still providing free courses, which helped it grow its community base to 5
million users by January 2010. Finally, the community engaged accredited teachers
on a pay as you go basis, where registered accredited teachers got paid for every
interaction they had with a student. Dedicated students were given access to the
premier content using their virtual tokens (Gartner 2013).
3.7 Recommendations for Institutions
Institutions willing to consider MOOCs at the current stage have to keep envi-
sioning the possibilities of MOOCs not as a single business model, but as a set of
business models centered around the notions of global online access, free or rela-
tively low fee, edification. The use of the term edification as opposed to education is
used here to connote both engagements in education: (i) those for obtaining formal
qualifications for employability purposes, and (ii) those for continuous development
through the pursuit of personal interests. With this in mind, interested institutional
parties should keep the following in mind during their search for a positioning in
the MOOC market.
• Different strands of MOOCs are very likely to develop depending on their

utilisation of big data.
• Big data analytics can be utilised as a marketing tool to target potential learners.
• Big data can be utilised as a learning tool to assist help—learning, by profiling
students learning patterns and to alert the education institution when human
intervention is required to motivate or support student learning.
• Within cMOOCs, big data analytics can be utilised as a tool to source and
integrate learning relevant learning material.
• xMOOCs are likely to transform into the MOOC equivalent of vocational
education, big data analytics can play a crucial role in matching industry-
required employment skills and curriculum development.
3.8 Summary
While MOOCs facilitated the distance learning aspects of formal education insti-
tutions, they also introduced an even more fundamental educational change; they
opened up participation to education by enabling a peer-to-peer learning. MOOCs
open new income streams for traditional institutions by not only relying on state
education subsidies and student tuition, but though employment recruiting services,
syndication, and sponsoring, as well as by advertising income, selling student
information to potential employers or advertisers. Big data analytics can enable the
personalization of the online learning process that was missing in previous online
instruction methods and facilitate this to happen at a global scale. Used as a ped-
agogical tool in learning analytics, big data along with educational data mining and
3.8 Summary 63
teaching analytics can all be seen as three aspects of the same solution to raising the
education standards of the youth without necessarily increasing the number of
educators required, thus making institutions more cost efficient. Utilizing existing
knowledge and education models relating to student effort and success, institutions
can use big data technologies to monitor students’ interaction with the system,
alerting administrators when student engagement patterns change in order to initiate
communication with the student, or indeed send a predetermined default email
communication. Big data in the form of learning analytics focus on capturing
student behavior and correlating it to achieving learning objectives, educational
data mining seek to design predictive analytics models for student attainment while
teaching analytics helps educator translate such findings into better course design
and student support procedures and interventions.
The reality however is quite different, MOOCs platforms do not share some of
the basic assumptions about student motivation and institutional norms, expecta-
tions, and obligations, so preexisting educational models are not a good fit for their
analysis. Moreover, as the Livemocha case suggested, MOOCs could utilize
gamification or other learning from online communities and worlds to ensure
continuous engagement and commitment.
References
Barabási, A.-L.: Linked: How Everything Is Connected to Everything Else and What It Means for
Business, Science, and Everyday Life (2003)
Berger, R.: Why should I take this MOOC anyway?, http://dlmooc.deeper-learning.org/about/why-
should-i-take-this-mooc-anyway/ (2014). Accessed 16 Nov 2014
Boyd, D.: Streams of content, limited attention: The flow of Information through social media.
Educ. Rev. 45, 26–36 (2010)
Campbell, B.J.P., Deblois, P.B., Oblinger, D.G.: Academic Analytics A New Tool for A New Era.
Educ. Rev. 42(4), 40–57 (2007)
Charlton, P., Mavrikis, M.: The Potential of Learning Analytics and Big data. Ariadne. (2013)
Coursera, http://www.coursera.org. Accessed 16 Nov 2014
Crosslin, M.: Designing a Dual Layer cMOOC/xMOOC, http://www.edugeekjournal.com/2014/
05/04/designing-a-dual-layer-cmoocxmooc/ (2014). Accessed 16 Nov 2014
Crowley, J.: cMOOCs: Putting Collaboration First, http://campustechnology.com/Articles/2013/
08/15/cMOOCs-Putting-Collaboration-First.aspx?Page=1, (2013). Accessed 16 Nov 2014
Davidson, C.: Major Players in the MOOC Universe, http://chronicle.com/article/The-Major-
Players-in-the-MOOC/138817/. Accessed 16 Nov 2014
Dellarocas, C.: Money Models for MOOCs. Commun. ACM 56, 25–28 (2013)
edX, http://www.edx.org. Accessed 16 Nov 2014
Executive Office of the President of USA: Big data: Seizing Opportunities., Washington, D.C.
(2014)
Fournier, H., Kop, R., Sitlia, H.: The value of learning analytics to networked learning on a
personal learning environment. Proceedings of the 1st International Conference on Learning
Analytics and Knowledge—LAK’11. p. 104 (2011)
García, O.A., Secades, V.A.: Big data and learning analytics: A potential way to optimize
elearning technological tools. Proceedings of the International Conference e-Learning 2013.
pp. 313−317 (2013)
Gartner: Business Model Innovation Examples in Education (2013)
Global Industry Analysts: ELearning: A Global Strategic Business Report, http://www.strategyr.

com/eLEARNING_Market_Report.asp (2014)
Goldberg, C.: Auditing Classes at M.I.T., on the Web and Free, http://www.nytimes.com/2001/04/
04/us/auditing-classes-at-mit-on-the-web-and-free.html (2001). Accessed 16 Nov 2014
Guri-Rosenblit, S.: Digital technologies in higher education: Sweeping expectations and actual
effects (2010)
Harvard University: HarvardX, http://harvardx.harvard.edu/. Accessed 16 Nov 2014
Kop, R.: The challenges to connectivist learning on open online networks: Learning experiences
during a massive open online course. Int. Rev. Res. Open Distance Learn. Distance Learn. 12,
19–38 (2011)
Laurillard, D.: Five myths about Moocs, http://www.timeshighereducation.co.uk/comment/
opinion/five-myths-about-moocs/2010480.article (2014). Accessed 16 Nov 2014
Livemocha, http://livemocha.com/. Accessed 16 Nov 2014
Liyanagunawardena, T.R., Adams, A.A., Williams, S.A.: MOOCs: A systematic study of the
published literature 2008-2012. Int. Rev. Res. Open Distance Learn. 14, 202–227 (2013)
Marques, J.: A short history of MOOCs and distance learning, http://moocnewsandreviews.com/a-
short-history-of-moocs-and-distance-learning/#ixzz36icU7Crp (2013). Accessed 16 Nov 2014
Massachusetts Institute of Technology: OCW is accessed by a broadly international population of
educators and learners., http://ocw.mit.edu/about/site-statistics/. Accessed 16 Nov 2014
Miniwatts Marketing Group: Internet Usage Statistics, http://www.internetworldstats.com/stats.
htm. Accessed 16 Nov 2014
Morrison, D.: How NOT to Design a MOOC: The Disaster at Coursera and How to Fix it, http://
onlinelearninginsights.wordpress.com/2013/02/01/how-not-to-design-a-mooc-the-disaster-at-
coursera-and-how-to-fix-it/ (2013). Accessed 16 Nov 2014
Ottmann, T.: MOOCs and/or eLectures; a means of virtualizing university education? Information
Technology Based Higher Education and Training (ITHET), 2013. International Conference
on. pp. 1−6 (2013)
P2PU, https://p2pu.org/en/. Accessed 16 Nov 2014
Siemens, G.: What are Learning Analytics?, http://www.elearnspace.org/blog/2010/08/25/what-
are-learning-analytics/ (2010). Accessed 16 Nov 2014
Summary, E., Findings, I., Among, D., Statistics, D.: HarvardX and MITx : The First Year of
Open Online Courses. (2013)
TIME Staff:Livemocha, http://content.time.com/time/specials/packages/article/0,28804,2012721_
2012929_2012923,00.html (2010). Accessed 16 Nov 2014
The Economist explains: The backlash against big data, http://www.economist.com/blogs/
economist-explains/2014/04/economist-explains-10 (2014). Accessed 16 Nov 2014
Udacity, http://www.udacity.com. Accessed 16 Nov 2014
Udemy, http://www.udemy.com. Accessed 16 Nov 2014
Virtonomics, http://virtonomics.com. Accessed 16 Nov 2014
Watters, A.: Top Ed-Tech Trends of 2013: MOOCs and Anti-MOOCs, http://hackeducation.com/
2013/11/29/top-ed-tech-trends-2013-moocs/ (2013). Accessed 16 Nov 2014
Watters, A.: Venture Capital and the Future of Open Education: FWK and MOOCs, https://www.
insidehighered.com/blogs/hack-higher-education/venture-capital-and-future-open-education-
fwk-and-moocs#ixzz36n6Kaqch (2012). Accessed 16 Nov 2014
Wilby, P.: Moocs, and the man leading the UK’s charge, http://www.theguardian.com/education/2014/
aug/19/moocs-man-leading-uk-foray-simon-nelson-futurelearn (2014). Accessed 16 Nov 2014
Yeager, C., Hurley-Dasgupta, B., Bliss, C.: cMOOCs and global learning: An authentic
alternative. J. Asynchronous. Learn. Netw. 17(2), 133–147 (2013)
Yuan, B.L., Powell, S., Olivier, B.: Beyond MOOCs: Sustainable Online Learning in Institutions
Beyond MOOCs : Sustainable Online Learning in Institutions. (2014). Accessed 16 Nov 2014
Yuan, L., Powell, S.: MOOCs and Open Education: Implications for Higher Education. JISC—
Cetis (2013)
Big Data Driven Business Models
4
Abstract
This chapter outlines the concept of ‘big data driven business model’ and utilizes
it to describe a set of businesses that rely on big data to achieve their key value
proposition and to substantially augment their value proposition to differentiate
themselves in order to gain competitive advantage. It describes the impact of big
data on each of the elements as identified in the Business model canvas. Also the
chapter discusses the potential of big data for mass customization and
personalization of product and services, as a value proposition in its own right,
on B2B and B2C logistics as well as for customer relationship management and
customer service. It also touches upon how big data has facilitated a shift in our
conceptions of utility as opposed to resources as the basis for socio-economic
value creation. Also, the chapter explores this issue a bit further by
understanding the implications of big data on partnerships, monetization, and
the opportunities and challenges it raises for accounting, budgeting and
performance metrics. In conclusion, the chapter acknowledges the synergistic
potential with other emerging technologies such as 3D printing, Robots, Drones,
self-driving cars, and the like.
4.1 Introduction
The focus of this chapter merits a more specific definition of the term Business
Model. Business models are a natural fit with IT-related business innovations.
Although rooted in transaction cost economics, it was really an Information
Communication Technology (ICT) related phenomenon. ICT made it feasible and
cost effective for businesses to collaborate with value networks in order to compete
(Don et al. 2000; Amit and Zott 2001). Bundling of product and services became
very popular during the 90s blurring the boundaries between industries.
So, what is a business model? Among the diverse available definitions (Amit and
Zott 2001), a business model can be seen as outlining the architectural logic of
business elements, such as a business structure, business processes, infrastructure, and

DOI 10.1007/978-3-319-10665-6_4
66 4 Big Data Driven Business Models
systems as well as finance options, i.e. how they all fit together to coordinate value
creation. It describes who buys the company’s products and services and why, how a
company organizes and which resources they utilize in order to finance, produce and
deliver their offerings, what they pay for doing so and how they get their money in.
Much like any model (Baden-Fuller and Morgan 2010), business models are
abstractions of real life and in this particular case, business models have been used
to describe various business phenomena. According to (Osterwalder et al. 2005)
people have used the term loosely to describe all real world businesses (e.g. the
capitalist model), or particular types of business with common characteristics (e.g.
the auction model), or a very particular real world business model (e.g. the Apple
model). Furthermore, according to (Osterwalder and Pigneur 2010), business
models comprise nine fundamental elements illustrated in Table 4.1.
While in the past, organizations relied on managers intuition to fill in the gaps of
scarce incomplete, poor information to make business decisions, the proliferation of
data generated every day through social media, cloud computing, and mobile
phones and soon the Internet of Things (IoTs) give managers a new headache, more
information that what they know what to do about. This combination of digital
intensity, connectivity, and big data provides a context of networked abundance
(Bharadwaj et al. 2013).
The purpose of this brief introduction of the concept was twofold. First it was
necessary in order to clarify how we are going to use the term in this chapter. We
are going to use the term ‘big data driven business models’, describing a set of
businesses which rely on big data to achieve their key value proposition and to
substantially augment their value proposition to differentiate themselves in order to
gain competitive advantage. Second, it would be useful to establish on the outset
the scope of this chapter. Most big data driven business models are currently
augmentations of existing value propositions (Hagen et al. 2013). Big data driven
business models are currently in development. Most businesses have introduced big
data in their projects portfolio on the basis of efficiency, which by default means
doing the same thing, with less money, and even fewer on effectiveness, i.e. doing
things better (Hagen et al. 2013). It will be some time before true visionaries
introduce business models whose unique selling point will rely on big data, and
Table 4.1 Nine fundamental elements of business models (Osterwalder and Pigneur 2010)
1 The customer segments the company serves
2 The value propositions that offer a solution to its customers
3 The channels through which the company communicates, distributes, and fulfils its sales
4 The customer relationships the company established and maintained with customers
5 The revenue streams accruing to the company from its customers
6 Key resources or assets required
7 Key activities, i.e. the tasks, processes and behaviors required
8 Key partnerships, which undertake responsibility for some of the above
9 Cost structure, i.e. the cost of maintaining and developing all of the above.
4.1 Introduction 67
very likely such value propositions to be augmented by other innovations also in

development in the social, business and technology spheres. So, within this chapter
we offer some projections along with descriptions of such scenarios.
Big data will have big implications for various fields (Hagen et al. 2013). Yet,
such changes seek to complement today’s sectors and business models. It is worth
noting that a business sector consists of a number of companies or business units
operating effectively under the same or similar business models. Table 4.2 shows
some of the indicated changes.
So far, there is a limited impetus for fundamental structural changes that place
big data to the core of company value creation as opposed to ‘nice-to-have’
capabilities. In that sense, the suggested changes seek to enhance existing business
models rather than disrupt them. For example, big data has been utilized to improve
existing business models as shown in Table 4.3.
Table 4.2 Existing business models currently enhanced by big data

1 Improving customer intimacy through customer profiling and personalized customer
service
2 Opening up product/service innovation by explicitly crowdsourcing ideas or analyze social
media generated opinions and feedback
3 Improving decision-making by generating business options based on larger amounts of
data, triangulated through various sources, filtered down to workable solutions in real time
4 Enabling real-time operations, utilizing smart assets, which can coordinate to carry out a
process or shift business processes instantly, while providing real-time tracking feedback
Table 4.3 Industries using big data to transform business models and improve performance in
many areas
Retail Manufacturing
Customer relationship Fraud detection and Product Process and quality
management prevention research analysis
Store location and layout Supply chain Engineering Distribution
optimization analytics optimization
Dynamic pricing Predictive
maintenance
Financial services Media and telecommunication
Algorithmic trading Fraud detection Network optimization Churn prevention
Risk analysis Portfolio analysis Customer scoring Fraud prevention
Advertising and public relations Energy
Demand signaling Sentiment analysis Smart grid Operational modeling
Targeted advertising Customer acquisition Exploration Power-line sensors
Government Healthcare and life sciences
Market governance Econometrics Pharmacogenomic Pharmaceutical
research
Weapon systems and Health Bioinformatics Clinical outcomes
counterterrorism informatics research
Adapted from Hagen et al. (2013)
These incremental changes will bring significant value to existing organizations

and change our customer expectations, which leave analytics lagging companies
behind in the competition game. Then again such models are not fundamentally
big-data driven, but big data enabled.
Fundamental changes can be envisioned by combining the analytical power of
big data analytics with new production technologies and new concepts about
production business value. For example, what business models can evolve by
combining the power of big data analytics, with other emerging technologies, such
as the ones shown in Table 4.4.
Will social services and primary care become obsolete by smart homes and
health tracking devices providing real-time information about the health status of an
individual, while artificially intelligent social robotics decide and administer
appropriate regular care to the elderly? Will bitcoin become part of the mix of or a
substitute for existing finance options? Can we organize centralized farming by
utilizing smart assets, drones and swarm robotics to tend to agricultural production.
Will the distinction between public and private transport systems disappear by
merging to create a new, on demand, door-to-door self-driven transport system? To
some of these ideas may seem far fetch, but for others some of these are expected to
become mainstream in the next 15−20 years. For example, pod based inner city
transport is already a reality in Masdar, Abu Dhabi (Feuilherade et al. 2014).
Real-time, big data analytics can transform production processes. For example,
big data analytics is utilized by General Electric as an ‘in process’ monitoring
mechanism to quality control highly sensitive and very expensive industrial 3D
printing processes of aerospace components where structural integrity is critical to
safety (Gereports 2013). For example, IBM’s Deep Thunder program is oriented
towards precision agriculture that combines microclimate predictive models with
remote drone monitoring to optimize agricultural processes, such as weeding,
spraying, watering and harvesting crops.
We continue this chapter by understanding the fundamental implications of big
data on each of the nine identified elements comprising a business model.
Table 4.4 Emerging

1 Home 3D printing
technologies suitable to be
combined with big data 2 Health and fitness monitors
analytics 3 Eye and hand tracking device
4 Drones
5 Self-driving cars
6 Bitcoin
7 (Ro)bots
8 Smart homes
4.1 Introduction 69
4.2 Implications of Big Data for Customer Segmentation
Big data has given the notion of ‘mass customization’ a new lease of life. Big data
can give companies the ability to target each customer individually based on their
preferences and purchasing habits, by integrating personal information about
website browsing, purchase histories, physical position, response to incentives, as
well as demographic information such as work history, group membership and
people’s views and opinions based on social influence and sentiment data. This data
enable ever finer targeting of content, offers, products and services, which can
deliver real and substantial returns, by increasing the probability of a final purchase
(Offsey 2014).
While mass customization is now enabled by big data, two even more disruptive
advancements can be facilities by new technologies. Making customizable products,
which can self-customize to suit the preference of their users. As products come with
embedded intelligence and machine-to-machine collaboration capabilities, they will
be able to undertake responsibility for serving their users. For example, a self-driving
car could double up as personal assistant modifying appointment schedules.
While big data is discussed primarily as a business tool, there is no reason why
big data cannot be utilized by consumers themselves to search and procure products
and services. Cognitive Code for example has invented advanced conversational
artificial intelligence that can interact in natural language with human beings while
interact in code with machines providing outputs and feedback in natural language
(Introducing SILVIA 2014). The perfect personal assistant for the electronic age! It
is a pity that companies such as Cognitive Code still orient their business model to
companies, such as call centers, as opposed to the public. Yet, I doubt it will take
long for consumer electronics giants, such as iPhone to build in such a capability.
4.3 Implications of Big Data as a Value Proposition
There are four big data ‘things’ that companies can sell for profit: raw big databases,
big data analytics services, big data experts, and big data technologies.
According to Forbes et al. (2014), many companies are geared towards selling
their customer data as a means to complement company revenues. While this may
bare a reputational risk for small companies, large organizations such as mobile
companies, banks, and airlines that collect huge amounts of information about our
whereabouts and shopping patterns are unfazed by such risks. However, raw data is
not of much use unless you analyze for a purpose. Banks are moving away from
selling raw data to developing higher value big data analytics services by recruiting
big data analytics experts to mine, analyze and synthesize data into consultancy
services.
Big data consultancy has become the must have service for Business Intelligence
consultancy providers. With big data analytics being harnessed in every aspect of
business activity, from strategy development to Human Resources (HR), and with
big data analysts being a scarce resource, the opportunity for consultancy work is
high. In fact, it would be impossible for a consultancy company to survive, without
offering big data consultancy services. Hence, all prestigious consultancy companies
as well as smaller boutique ones, offer clients big data analytics services.
Another big data business is related to data scientist recruitment services.
Recruiting businesses cannot shy away from the need to recruit and deploy data
scientist talent. Yet, big data analysts are currently a scarce resource. So new
business models are developing to serve the needs of companies. Kaggle (2014) for
example, uses a hackathon type competition model for undertaking predictive
modeling to data science projects but also doubles as a learning community for
those who want to attain data science expertise.
4.4 Implications of Big Data for Channels
With online shopping taken to new heights, business-to-consumer (B2C) delivery

options are open for discussion. A constraint in achieving high operational
efficiency in a distribution network occurs at the “last mile”, i.e. delivery at a
specific destination. Click and collect delivery systems seems to be the hottest trend
in B2C logistics, for example, in a town as London in the United Kingdom (Butler
2014). Click and collect seems to work at many levels, people do not need to be
somewhere waiting for their parcel, their parcel waits for them at a safe pick-up
point, companies don’t burden themselves with the cost of unsuccessful deliveries,
and citizens do not have to put up with the CO2 emissions of a fleet making rounds
delivering parcels.
Yet, there is scope for applying big data techniques to drive costs down even
further, while increase flexibility for customers. One way is to utilize real-time
optimization on delivery routing to maximizing the routing schedules of conven-
tional delivery fleet. Another approach would be to utilize combinatorial optimi-
zation to deliver product to people on the go. Combinatorial optimization can
process, for example, real-time information about recipient’s position and direction,
thus, deliveries can be schedules dynamically. This involves locating the paths of
both recipients and delivery vehicles in order to re-route vehicles on the go to the
next best point to meet recipients. Currently, the sequencing of deliveries is carried
out manually by drivers following each delivery. A big data driven sequencing
software can rapidly process information about deliveries, current traffic conditions,
loading bay availability, recipient availability to avoid unsuccessful deliveries,
thereby minimizing costs. As, big data have also facilitated a trend towards
“crowdsourcing” everything, why not utilize the crowd to deliver packages too?
Even in business-to-business (B2B) logistics, big data predictive modeling can
also take just-in-time operational efficiencies to a new level, reducing the space
required to store goods while ensuring business continuity. Predictive modeling can
now take into account not only customer purchasing patterns and feedback, but also
an increasing number of disruptions due to civil unrest, natural disasters, or even
4.4 Implications of Big Data for Channels 71
sudden economic developments. Big data technology and analytics can keep an eye
and keep track of developments in any of the critical risk factors that can
compromise the going concern of the organization from a variety of sources (e.g.,
social media, blogs, weather forecasts, news sites, stock trackers, and the like). It
can then alert and instigate risk management scenarios, such are fulfilling an order
via another country or route adding to the company’s resilience and flexibility.
4.5 The Impact of Big Data on Customer Relationships
Profiling customers utilizing big data capabilities to increase sales have been at the
forefront of adopting the new technologies. But how about customer service? With
so much information about customers, surely companies can improve customer
service to increase customer intimacy and loyalty. One of the key challenges for
customer service operators has been the lack of comprehensive customer infor-
mation at their fingertips. Unavailable data on the front line make customer service
representatives unable or slow to respond to customer issues and requests and offer
alternative options. Customer Relationship Management (CRM) projects have
made big investments towards addressing the matter, yet they have been costly,
high maintenance projects. Moreover, standardized siloed supply chain operations
made dealing with exception cases an operational nightmare, with such cases falling
usually between two stools or requiring expensive dedicated teams to handle them
appropriately.
Big data technologies, such as Hadoop, can relate isolated data silos via hyper-
fast in-memory analytics platforms, providing representatives with real-time
information to assist customers. However, how companies relate this to improve
customer service as opposed to sales is not yet clear. Big data customer relationship
projects are lagging behind, but this is expected. Customer relationship projects are
considered costly compared to their revenue generating potentials and tend to be
instigated later in the maturity lifecycle when the market begins to plateau and
companies seek to maximize value elicited from business assets.
Banks are likely to be the first to jump into this bandwagon, not only because
they have budget to do so, but also because they need to rebuild the trust of the
disenchanted public following the global financial crisis. Trends towards rebuilding
strong personal relationship can be seen in retail banks’ advertising themes.
Banorte, a Mexican bank, focuses strongly on its 13 million customer base and
capitalizes on new big data technologies to design banking services. Big data
enabled marketing automation can help the bank service individual customer needs
while keep the costs of marketing, creating a personalized experience at a return on
investment. Also, harnessing the power of automated personalized customer
service, requires business process reengineering throughout the entire organization
and rethinking of the customer journey (Wagle 2014).
Big data will also expand the use of automated call centers using natural
language recognition patterns and response, thus, making Artificial Intelligence
(AI) becoming increasingly more sociable in terms of responding to customer
queries using voice. Hence, companies with large call centers are likely to adopt
this next-generation AI enabled and big data informed speech solutions to drive
costs down as part of big data oriented business models.
4.6 The Impact of Big Data on Revenue Stream
Social media has enabled peer-to-peer transactions and, in effect, collaborative

consumption. Now hotels compete with Airbnb, a peer to peer platform, for holiday
makers’ cash (Riedy 2014). Airbnb supports the rental of spare bedroom space to
private travelers. Established in 2008, Airbnb lists more than 500,000 properties
worldwide and has hosted more than 8.5 million people increasing the value of the
company at US$2.5 billion (Riedy 2014). The financial crisis has changed social
values about re-use, providing impetus for the rise of a sharing economy emerges as
a potential threat to established businesses, as consumers are sharing things they
already have, then they don’t need to buy something new or so often. Blah blah cars
(2014), enables car sharing with someone who goes to their point of destination, by
match making people with cars with people in need for a ride. big data can make
these transactions safer, real-time and efficient. For example, car sharing does not
mean you hitching a ride with a complete stranger. Big data technology can
advance the safety of car sharing, via car monitoring, passenger and driver health
monitoring, and the like, offering precautionary measures and deterrents of common
social risks. In transaction cost economics terms, the ‘big brother’ capability of big
data can substitute institutions roles to monitor and ‘police’ adherence to rules and
regulations.
Another disruptive change in the revenue stream is electronic currencies. What
will the company of the future be paid with? LeHong, research vice president and
Gartner Fellow, also pointed to the cryptocurrency Bitcoin (Nakamoto 2008), as a
major disruptive force:
Typically only national governments could issue currency, but we’re seeing people put
more trust in a privately funded startup than in most Western governments (Turner 2013).
Of course, the regulatory environment for such currencies is still uncertain and
the risks currently not assessed in a systematic way. According to White & Case
(Duplat and Vercauteren 2014), the internet environment where virtual money is
held and traded is still risky, when trading platform or digital wallet could be
hacked, making owners lose their virtual money and systems reliability, particularly
with regard to the risk of fraud, actually not properly assessed. In addition, fluc-
tuations in the virtual currency exchange rate can result in substantial financial
losses, and there is no legal guarantee that virtual currency can be exchanged at any
time for its original value. Finally, virtual currency is not legal tender: no-one is
obliged to accept payment with virtual currency. While such fears are currently
4.6 The Impact of Big Data on Revenue Stream 73
justified, the future of Bitcoin payments depends on people’s acceptance and there
are strong proponents for its future.
According to professor Campell Harvey of Duke University (Card 2014),
Bitcoin, like all cryptocurrency, bears fewer risks than debit or credit card fraud and
hacking, and minimal cost of transacting compared to traditional methods adopted
by traditional institutions. Cryptocurrency relies on the power of computers to
enable real time exchange of ownership, verification of ownership, as well as the
ability to algorithmically design conditional contracts (Card 2014). In addition,
cryptocurrency-based transactions require triangulation of data to ensure security,
privacy, and trust without using centralized institutions. All of the above issues will
require the number crunching and user profiling power of big data analytics, which
will have a big role to play in analyzing the huge amounts of data that Bitcoin and
cryptocurrency systems will be generating. Companies will be able to issue and
guarantee their own currencies, which can then be spent in associated providers.
Accordingly, one can only imagine the power of retail suppliers such as Amazon
and Google. For example, Google Ventures has put money into Ripple, the open
currency exchange called Ripple, which allows users can trade any currency
without requiring any broker or third-party to facilitate the trades.
4.7 The Impact of Big Data on Key Resources

and Key Activities
Traditionally capital resources are manmade assets organized to generate income.

There are certain underlying assumptions in this definition that are likely to be
challenged in the future. First, resources presume ownership by the company, and
therefore their use and disposal at company will. Second, resources do not give
value unless they are used; hence it is utility that provides value, thus, neither
resource nor ownership per se. Third, to generate value, the utilization of resources
need to be organized into perform key business processes that deliver value, in the
form of a product or service desirable to a market. Big data can revolutionize all of
the above for business model innovation.
To start with, big data can revolutionize employment relationships, as more and
more companies can rely on crowdsourcing, thus, outsourcing a function once
performed by employees to an undefined (and generally large) network of people in
the form of an open call. This can take the form of peer-production when the job is
performed collaboratively. Crowdsourcing growing popularity is due to its cost
efficiencies and its global reach potential (Fretty 2014). As organizations rely more
and more on crowdsourcing, they will require a strong coordination mechanism,
capable to gather and reconcile often diverse information and views. After all,
businesses need to be able to judge data quality, find ways to overcome geo-
graphical differences and apply to organizational goals (Fretty 2014). Actually,
crowdsourcing and big data can revolutionize public service. The United Nations
(UN), for example, has started utilizing Mindjet’s SpigitEngage, a big data and
Crowdsourcing platform to devise best practice in tackling refugee issues, such as,
e.g., learning a new language in order to absorb new culture, finding jobs, and
gaining access to information and vital social services. Also, using gamification,
SpigitEngage platform allows the grassroots ideas to rise to the top avoiding status
quo bureaucratic barriers and gate keeping (Spigitengage 2014).
The above issues have also big implications about management structures and
managers. If the end to end decision making process can be coordinated by big data
algorithms, and so is the monitoring the implementation processes as well as the
dynamic allocation of people and resources, what is the purpose of expensive
organizational structures? This may seem farfetched from today’s reality where
formal organizations are still the dominant work organization structure, but flexing
the boundaries has already started and the trend seem irreversible.
Another resource impacted by big data is the capital itself, i.e. the financing of
business ventures. Social media and big data have enabled the merging of sales and
financing through crowdfunding. A quick study of Kickstarter projects and asso-
ciated pledges will demonstrate that most people effectively pre-buy the product or
service (Jeffries 2013). In that sense, consumers act as financiers of their wanted
products, and at the same time as early adopters of it. Take for example, Coolest
Cooler (a portable cooler, integrating a cooler, a blender for outdoor party occa-
sion), the most funded Kickstarter project who reached financing heights of more
than $13 million. More than 57,000 out of its 62,642 backers were people who
wished to pre-purchase the product, creating the market and the revenue for the
company way ahead of its production (Grepper 2014).
Finally, big data not only pose enormous challenges and technology investments
to access, process, and analyze the massive amounts of data, but requires companies
to also make corresponding changes on business processes to capitalise on it. For
example, Apple capitalised on market and supply chain data when they improved
supply chain management practices to become twice as fast as the average company
in the electronics market (Bharadwaj 2013).
4.8 The Impact of Big Data on Key Partnerships
While we already see partnerships in travel and hospitality to scale operation by

sharing reservation systems, loyalty programs, and cross-selling opportunities, we
will increasingly see the formation of different firms to pull together their strengths,
organized through plug-and-play architectures to create competitive advantage
(Bharadwaj et al. 2013).
By enabling real-time coordination monitoring and feedback, big data can provide
a sense of security to those involved in supply chain collaborations. Previously,
supply chain risks were moderated by policies/procedures and contractual agree-
ments that acted as deterrent and provided some security assurances against negli-
gence, deception and deceptive practices. Real-time information analysis of supply
chain coordination data could flag issues ahead of time, highlight underlying reasons
for delays and offer contingency planning alternatives (Zage and Glass 2013).
4.8 The Impact of Big Data on Key Partnerships 75
Further developments in advanced predictive big data analytics are likely to lead
to the development of complex decision making scenarios based on a variety of
stress tests, in multiple market conditions, against, numerous key performance
indicators (KPIs). Such business simulations, can be used to design and redesign
supply chain operations (Groves et al. 2014). For example, by simulating different
market and competitor scenarios, big data can assess business agility and resilience
of a given supply chain. Price volatility in the commodities and currency fluctua-
tions has made long-term procurement a risky business. For example, organizations
buffer themselves from the risks of late delivery and inability to fulfill orders by
planning for a minimum safety stock, while price competition on scarce resources
and raw material often leads to bidding for more quantities to ensure sufficient
orders. Allegedly such risks can be reduced by big data analytics by improving
price predictions or bid efficiency. The Internet of Things (IoTs) will also improve
the quality of information about the location of physical things in the supply chain.
Yan et al. (2014) envision a Cloud of things supply chain solution to facilitate
resource sharing and participant collaboration in the whole supply chain life cycle,
where supply chain condition perception, heterogeneous network access conver-
gence, and resource “servitization” co-exist.
4.9 The Impact of Big Data on Cost Structures
With crowdfunding, crowdsourcing, and big data-driven supply chain join ventures
dismantling the traditional boundaries of organizations, the definition and bound-
aries of cost economies change dramatically. On the other hand, as data, infor-
mation and knowledge become the focal point for collaboration at all levels,
accounting for intellectual property rights, data, and intangible assets will likely to
become a focal point (Viscusi and Batini 2014).
More than 20 % of large companies already class data as an asset on their balance
sheets. Data valuation is likely to become a key accounting skill in the future.
Valuing data to date have been impossible as intangible assets tend to get hidden in
existing reporting and governance systems once developed for the industrial age.
Even when visible, data usually gets an arbitrary valuation. Calculating depreciation
is difficult, and information velocity makes such calculations even more difficult and
unreliable. In addition, like in most assets, the value of data depends on its utili-
zation. To move towards a costing model of data resources, new methodologies need
to be developed and agreed amongst accounting and finance professionals that
determine key assumptions about what data is of value (Chua 2013).
What matters from an accounting point of view is to assess the monetary value
or profit an organization creates out of the internet of everything and big data. To do
so, management accountants should combine internal financial data and perfor-
mance analysis with overall economic performance metrics. They should also be
able to compare these to some referent organizational groups within the sector. In
addition, these will be affect by political, economic social and technological
influences on their industry and on the business. Yet, nowadays all these
assumptions are open to negotiation and redefinition. Technology trends redefine

industry boundaries, companies’ strategic trajectory and capabilities against other
organizations. For example, Facebook compete with Google for advertising revenues,
even though what they offer to users are completely different value propositions
(Dias 2013).
In summary, in the same way defining new markets has been the key driver for
tech-driven business models; developments in the area of cost and management
accounting will likely follow. Big data driven business models are still in their
infancy, hence cost economics are still in embryonic stages.
4.10 Organizational Advantages and Opportunities
The big data era is an open minefield for all who are willing to unlearn their old ways
and experiment with new ones. According to McKinsey, some sectors or indeed
some players maybe more advantaged over others to making such moves (Gobble
2013). For example, companies those have invested deeply in IT and have large data
pools to exploit as well as those in information industries, who have the expertise to
perform sophisticated analytic techniques, are likely to be the first to create business
value. On the other side of the spectrum, public sector organizations are in need for
radical change and can more readily open up their information resources to leverage
the value of public services to citizens. They are more likely as well to find large
pools of volunteers willing to help with crowdsourcing projects.
Mature sectors characterized by a large number of competing organizations are
likely to stay behind, due to ego-driven leadership, strategic myopia, technophobia,
and conflicting views about the future of the industry and sector infighting (EY
2014). Are however such sectors that could most benefit from sharing company
information, identifying new market opportunities and adopting ‘collaborate-to-
compete’ networking partnerships. The extent to which sector leaders and associa-
tions are ready to invest in these opportunities will define the future of their industry.
Companies should be able to reflect on whether and how they can create new
business models based on big data. The Internet of Things (IoTs) will intensify
information abundance and raise more points for reflection, such as, e.g., how do
we convert the data generated or captured by IoTs into knowledge to provide a
more convenient environment to people. To be able to devise new business models,
companies should open up to new ideas and make sure they can quickly assimilate
them wherever these come from, (grassroots initiatives, open innovation competi-
tions). They should also be prepared to modify or fundamentally change their
internal structures to deploy new products and services. This requires a ‘u-turn’ in
mentality and culture for most organizations (Tsai et al. 2014). Tsai et al. (2014)
identify a number of IoTs applications that will change existing sectors and ser-
vices. These have been loosely categorized into: (1) smart cities, (2) smart envi-
ronment, (3) smart water, (4) smart metering, (5) security and emergencies, (6)
retail, (7) logistics, (8) industrial control, (10) smart agriculture, (11) smart animal
farming, (12) domotic and home automation, and (13) eHealth. IoTs also opens up
4.10 Organizational Advantages and Opportunities 77
the possibilities for businesses to think of services that utilize big data analytics to
(a) better inform ‘human’ processes, (b) automate ‘human’ processes, and to (c)
coordinate with other ‘things’, thus, creating end-to-end automated services.
4.11 Organizational Challenges and Threats
There are a number of challenges with respect to redesigning a big data driven
business model. Technology assessment capability, technology maturity levels,
evolving legal frameworks around privacy and intellectual property, data analyst
talent scarcity, technology investment capability, internal data availability and
quality, interoperability, security issues, limited business process understanding and
associated metrics, data hoarding and old fashioned hierarchical leadership: these
are just few of the barriers to big data driven approaches to business model rede-
sign. These barriers have been referred to also in previous sections and chapters. In
this section, it would be appropriate to refer to challenges with respect to adopting a
big data driven business model innovation (Osterwalder et al. 2005).
4.11.1 Creativity and Innovation Capability Deficit
The big data innovation era suggest that managers need to think “laterally” outside
the box. Many established organizations have promoted to director positions good
administrators who can maintain order and efficiency. For many small and medium
organizations the top management team is burdened with the day-to-day operation
of the business, leaving little time, “headspace” and energy to renew their skills,
network and keep up with developments outside their immediate field. Management
structures often lack both the personal creativity and innovation skills and the right
attitude to foster creativity and innovation required to envision new big-data based
business models.
4.11.2 Interrogating Big Data
Big data demands inductive reasoning, a skill much underused in modern society
and (Malle 2013) a fundamental element of the big data disruption. Knowledge
from big data is produced by “agnostic” statistics, lacking a causal framework to
explain it. Induction allows us to generalize a phenomenon, even if it is observed
only once. For managers who are used to seek the certainty of reliability, repeat-
ability, standardization, risk assessment, to base business decisions on non-
repeatable, inductive observations will be a fundamental paradigmatic change.
Inductive reasoning is developed through constant interaction with data examined
through various lenses to develop into interpretation patterns (i.e. reasoning heu-
ristics) (Malle 2013).
4.11.3 Plug and Play Architectures
Speed is a core factor for the creation of value through from big data. Big data is
going to speed up decision making, requiring operations to leverage to match the
rate of change. Can today’s established organizations respond with operational
agility to such fast-changing, data-driven decisions? Though modularizing the
business architecture into Lego-like, component-based, architecture seems ideal, in
reality one needs to ponder the feasibility. Consider, for example, human resources.
Does the new configuration require staff upskilling? Does it require putting new
processes or policies in place? Will people require training? What about the con-
tractual agreements or job descriptions? If a business is using contracting staff or
temporary staff, can it source the right people timely? Can the business maintains
commitment of getting the job done? In addition, contractual agreements and
relationships with suppliers and channels can complicate making changes to
operating models (Bürgi et al. 2004).
According to Bürgi et al. (2004) one mitigating strategy is to create a business
model portfolio based on market scenarios which organizations can switch as
conditions change. While this could be a solution, it does not take into consider-
ation the reality of a turbulent environment, characterised by ambiguity about the
factors that will affect the business model. Hence, it is impossible to foresee them
and factor them in drafting job descriptions or contractual agreements. Perhaps big
data will not only change how we develop business models but also our definition
of what business models are, what are the elements of their architecture and what is
necessary for strategy.
4.12 Summary
This chapter has discussed business models as the architectural logic that identify
how business elements (such as a business structure, business processes, infra-
structure, and systems) fit together to coordinate value creation (Osterwalder et al.
2005). It then went on to describe the impact of big data on each of the elements as
identified in the Business model canvas proposed by Osterwalder and Pigneur
(2010). In particular, it has discussed the new impetus that big data and IoTs
technologies give to mass customization and personalization of product and ser-
vices. Furthermore, the chapter has investigated big data as a value proposition in
its own right and how an industry may be created around the sales of big data and
big data analytics technology, analytics consultancy and data scientist recruitment.
The chapter has also touched upon big data solutions for B2B and B2C logistics as
well as for customer relationship management and customer service.
Furthermore, the chapter has analyzed the impact of big data on revenues, as it
has facilitated new forms of value creation from the emergence of new currencies,
in combination with social media and cryptocurrencies. Then, it has been also
described how big data has emphasized the value of ‘utility from’ (as opposed to
4.12 Summary 79
‘ownership of’) capital resources. It has also been explained how this shifts the
emphasis away from the organization as an entity towards an understanding of the
organization as an a dynamic process of value creation. Also, the chapter has
explored this a bit further by understanding the implications of key partnerships in
the big data era and the spanning of organizational boundaries. It has also discussed
the monetization implications; the opportunities and challenges it raises for
accounting, budgeting and performance metrics. With these in mind, the chapter
has highlighted some of the advantages, opportunities challenges and threats around
business model innovation, relating them more to the readiness of organizations to
assume such an undertaking, rather than describing the many possible big data
driven business models to evolve in the future.
Finally, the chapter concludes acknowledging that while so far big data has been
used to improve existing business models, far more futuristic scenarios will emerge
in combination with other emerging technologies.
References
Amit, R., Zott, C.: Value creation in E-business. Strateg. Manage. J. 22, 493–520 (2001)
Baden-Fuller, C., Morgan, M.S.: Business models as models. Long Range Plann. 43, 156–171
(2010)
Bharadwaj, A., Sawy, OA.El., Pavlou, PA., Venkatraman, N: Digital business strategy: toward a
next generation of insights. MIS Q., 37, 471–482 (2013)
BlaBlaCar: http://www.blablacar.com/ (2014). Accessed 16 Nov 2014
Bürgi, P., Victor, B., Lentz, J.: Modeling how their business really works prepares managers for
sudden change. Strateg. Leadersh. 32, 28–35 (2004)
Butler, S.: Ten things Christmas taught us about the UK retail revolution. http://www.theguardian.com/
business/2014/jan/12/10-things-christmas-taught-us-about-uk-retail-revolution (2014). Accessed
16 Nov 2014
Card, J.: Bitcoin: a beginner’s guide for entrepreneurs. http://www.theguardian.com/small-
business-network/2014/oct/17/-beginners-guide-for-entrepreneurs (2014). Accessed 16 Nov
2014
Chua, F.: Big data: its power and perils. IMA-ACCA. www.accaglobal.com/bigdata (2013).
Dias, C.: Industry insights: management accountants’ role with big data. http://www.cimaglobal.com/
Thought-leadership/Newsletters/Regional/The-CIMA-Edge-South-Asia-and-Middle-East/2013/
December-2013/Industry-insights-Management-Accountants-Role-in-Internet-of-Everything-
and-Big-Data/ (2013). Accessed 16 Nov 2014
Don, T., David T., Alex L.: Digital capital: harnessing the power of business webs (2000)
Duplat, M., Vercauteren, L.: Ban on Distribution of Certain Financial Products in Belgium. White
& Case LLP, Belgium (2014)
EY: Big data—changing the way businesses compete and operate. EYGM Limited, UK (2014)
Fakto, S.: Never sell data. http://www.forbes.com/sites/stevefaktor/2014/03/25/never-sell-data/
(2014). Accessed 16 Nov 2014
Feuilherade, P.: Personal systems herald “smart mobility”. http://www.iec.ch/etech/2014/etech_
0314/ind-1.htm, (2014). Accessed 16 Nov 2014
Fretty, P.: Merging crowdsourcing with big data analytics. http://www.big-dataforum.com/302/
merging-crowdsourcing-big-data-analytics#sthash.cby7c3Bk.dpu (2014). Accessed 16 Nov
2014
Gereports: Big data meets 3-D printing: big data to monitor laser-printed jet engine parts. http://
www.gereports.com/post/77209216443/big-data-meets-3-d-printing-big-data-to-monitor
(2013). Accessed 16 Nov 2014
Gobble, M.M.: Resources: big data: the next big thing in innovation. Res. Manage. 56, 64–67
(2013)
Grepper, R.: Coolest cooler: 21st century cooler that’s actually cooler. https://www.kickstarter.
com/projects/ryangrepper/coolest-cooler-21st-century-cooler-thats-actually (2014). Accessed
16 Nov 2014
Groves, W., Collins, J., Gini, M., Ketter, W.: Agent-assisted supply chain management: Analysis
and lessons learned. Decis. Support Syst. 57, 274–284 (2014)
Hagen, C., Ciobo, M., Wall, D., Yadav, A., Khan, K., Miller, J., Evans, H.: Big data and the
Creative Destruction of Today’s Business Models. AT Kearney Publishing, Chicago 1−18
(2013)
Introducing SILVIA. http://silvia4u.info/silvia/ (2014). Accessed 16 Nov 2014
Jeffries, A.: Kickstarter is not a store, except when it is. http://www.theverge.com/2013/4/17/
4230440/kickstarter-is-not-a-store-except-when-it-is (2013). Accessed 16 Nov 2014
Kaggle Inc.: The home of data science. http://www.kaggle.com (2014). Accessed 16 Nov 2014
Malle, J.-P.: Big data: farewell to cartesian thinking? http://www.paristechreview.com/2013/03/15/
big-data-cartesian-thinking/ (2013). Accessed 16 Nov 2014
Nakamoto, S.: Bitcoin: a peer-to-peer electronic cash system. bitcoin.org. October 2008. https://
bitcoin.org/bitcoin.pdf (2014). Accessed 16 Nov 2014
Offsey, S.: Micro-segmentation in the age of big data. http://marketbuildr.com/blog/segmentation-
in-the-age-of-big-data/ (2014). Accessed 16 Nov 2014
Osterwalder, A., Pigneur, Y.: Business Model Generation: A Handbook for Visionaries, Game
Changers, and Challengers. Wiley, New Jersey (2010)
Osterwalder, A., Pigneur, Y., Tucci, C.: Clarifying business models: origins, present, and future of
the concept. Commun. Assoc. Inf. Syst. 16, 1–25 (2005)
Riedy, C.: The sharing economy spooking big business. Conversat. Trust. http://theconversation.
com/the-sharing-economy-spooking-big-business-19541 (2014). Accessed 16 Nov 2014
Spigitengage. http://www.mindjet.com/spigitengage/ (2014). Accessed 16 Nov 2014
Tsai, C.-W., Lai, C.-F., Chiang, M.-C., Yang, L.T.: Data mining for Internet of things: a survey.
IEEE Commun. Surv. Tutor. 16, 77–97 (2014)
Turner, D.: CIOs must reinvent their businesses on three levels: Hung LeHong. http://www.
ibmbigdatahub.com/blog/cios-must-reinvent-their-businesses-three-levels-hung-lehong (2013).
Viscusi, G., Batini, C.: Digital information asset evaluation: characteristics and dimensions. In:
Caporarello, L., Di Martino, B., Martinez M. (eds.) Smart Organizations and Smart Artifacts
SE—9, vol. 7, pp. 77−86. Springer International Publishing (2014). doi:10.1007/978-3-319-
07040-7_9
Wagle, L.: How big data helps banks personalize customer service. http://www.forbes.com/sites/
ibm/2014/06/16/how-big-data-helps-banks-personalize-customer-service/ (2014). Accessed 16
Nov 2014
Yan, J., Xin, S., Liu, Q., Xu, W., Yang, L.: Intelligent supply chain integration and management
based on cloud of things. Int. J. Distrib. Sens. Netw. 624839, 1–15 (2014) doi: http://dx.doi.
org/10.1155/2014/624839
Zage, D., Glass, K., Colbaugh, R.: Improving supply chain security using Big data. In: 2013 IEEE
International Conference on Intelligence and Security Informatics. IEEE, pp. 254−259 (2013)
Part II
Organization
Big Data Governance
5
Abstract
The aim of this chapter is to help readers gain an overview of the topic of ‘Big
Data Governance’ to find easily the information it offers. Over the past few
years, many organizations have begun their adventure with big data. While many
are only experimenting with new technologies and innovations to improve their
competitive advantage, few are truly thinking ahead for long-term success and
even fewer will truly succeed in gaining such competitive advantage. The reason
is simple: you cannot compete on analytics alone. After all, what do analytics
analyze? Information needs to be trusted in order to be acted upon, but how that
could be accomplished? That is the traditional role of data and information
governance, which also the case of big data has to create trusted high quality
information to make big data analytics more effective, which brings us to the
subject of this chapter. Thus, this chapter reviews the common definitions of big
data governance and its core components. Then the chapter focuses on common
big data maturation models. The next section is the organizational benefits and
challenges of governing big data, followed by three case studies to establishing a
business case for big data governance.
5.1 Introduction to Big Data Governance
‘Big Data’ is less about size and more about quality. Indeed, data comes from a
variety of sources such as RFID tags, social media, GPS signals, traffic flow sensors,
satellite imagery, broadcast audio streams, banking transactions, digital pictures, the
content of web pages, online videos, telemetry from automobiles, financial market
data, the list goes on. From organizational perspectives, as organizations have
grown, the data associated with them also grew exponentially and today there are
many complexities related to their data resulting in a great challenge; to give them a
single consistent view of their data.

DOI 10.1007/978-3-319-10665-6_5
84 5 Big Data Governance
Most of the big organizations have data in multiple applications and in different
formats. The mobile revolution, which business firms are experimenting nowadays,
has completely changed how they capture the data and build intelligent systems.
Big data tends to be operational in nature and are characterized by ‘3 Vs’—large
volumes, high velocity and a variety of formats such as structured, unstructured and
semi-structured (Bertot and Choi 2013; Kruger and Foster 2013; Michael and
Miller 2013; Sathi 2012; Steele and Nash 2012; Zicari 2012). Figure 5.1 shows big
data’s three dimensions. These three dimensions and related characteristics are a
useful lens through which we can understand the nature of data and how it can be
interpreted. Also, Table 5.1 illustrates the measures of scale of data.
Big data challenges are therefore multifaceted and characterized by a combi-
nation of the so-called ‘3 Vs’ (shown in Fig. 5.1 and also mentioned in the Preface),
that make it difficult to extract information and business insight. Furthermore and in
addition to those three characteristics, when big data is processed and stored, other
dimensions such as governance, ethics, security, and policies come into play (Cline
2014; Franks 2012; Hurwitz et al. 2013; Jee and Kim 2013; Michael and Miller
2013; Raghupathi and Raghupathi 2014; Russom 2008; Warden 2011; Webb
2006). Considering governance, there are currently various definitions available for
data and information governance. For example, Data Governance Institute (DGI)
(2014) defines Data Governance as
Data Variety
Data Velocity Data Volume • Table
• Real time • Megabyte • Data base
• Near real time • GigaByte • Photo, Web,
• Periodic • Terabyte Audio
• Batch • Petabyte • Social, Video,
Mobile
Fig. 5.1 Big data’s growths dimensions and related characteristics
Table 5.1 Measuring big

data Measuring big data
1,000 Gigabytes = 1 Terabyte (TB)
1,000 Terabytes = 1 Petabyte (PB)
1,000 Petabytes = 1 Exabyte (EB)
1,000 Exabytes = 1 Zettabyte (ZB)
1,000 Zettabytes = 1 Yottabyte (YB)
Adapted from Mohanty et al. (2013)
5.1 Introduction to Big Data Governance 85
a system of decision rights and accountabilities for information-related processes, executed

according to agreed-upon models which describe who can take what actions with what
information, and when, under what circumstances, using what methods.
Furthermore, Debra Logan Vice President of Gartner defines Information

governance as
the specification of decision rights and an accountability framework to encourage desirable
behavior in the valuation, creation, storage, use, archival and deletion of information. It
includes the processes, roles, standards and metrics that ensure the effective and efficient
use of information in enabling an organization to achieve its goals (Debra Logan 2010).
As for big data governance, Soares (2013) defines it in a clear and compre-
hensive manner as follows (p. 4):
Big data governance is part of a broader information governance program that formulates
policy relating to the optimization, privacy, and monetization of big data by aligning the
objectives of multiple functions (Soares 2013)
He then explains the main parts of the definition in greater detail to illustrate
different aspects of ‘big data governance’ as described in Table 5.2. Also, summary
of the main parts of big data governance is shown in Fig. 5.2, where the role of
information governance is highlighted, likewise.
After explaining the key aspects of big data governance, we are now moving to
the big data framework. Effectively managing big data requires skill sets and
technologies different from those used to manage “regular” data. It may also require
a change in business processes.
Companies need to start with a big data governance program to ensure there is
clarity on data access, integration, usage, management and ownership of big data.
Therefore, it is important for organizations to expand the scope of their existing
data governance program to include big data or, if one does not exist, to establish a
data governance program to support your existing enterprise data and your big data.
A high-level framework for big data governance adapted from Soares (2012, 2013)
is shown in Fig. 5.3. This high-level framework includes three key dimensions: big
data types, information governance disciplines, and industries and functions, which
will be discussed in what follows.
5.1.1 Big Data Types
As data volumes and complexities continue to grow and as organizations continue

to acquire big data, this massive amount of data will be overwhelming without
strong data governance frameworks. Big data can be classified into five distinct
types: web and social media, machine-to-machine, big transaction data, biometrics,
and human generated (Soares 2012, 2013). In this classification, web and social
media includes clickstream and social media data such as Facebook, Twitter,
LinkedIn, and blogs. Machine-to-machine data (or M2M) refers to technologies that
allow wireless, wired or hybrid systems to communicate with other devices such as
Table 5.2 Definition of the key components of big data governance

Components Description
Big data is part of a broader Organizations should extend the current scope of
domain of information governance information governance to include big data, using
power users of big data such as data scientists as well as
supporting big data with other associated disciplines
such as metadata, privacy and master data
Big data governance is about A data governance policy should obey the
policymaking organization’s legal and regulatory requirements. A big
data governance policy might state that an organization
will not integrate a customer’s online social account into
his or her master data record without the customer’s
informed consent
Big data must be optimized Organizations need to optimize and improve the quality
of their big data in the following areas:
• Metadata: to build information about inventories of big
data
• Data quality management: to cleanse big data just as
companies conduct preventive maintenance on physical
assets
• Information lifecycle management: to archive and
retire big data when it no longer makes sense to retain
these massive volumes
Privacy and confidentiality of big Organizations also need to create and follow appropriate
data is essential policies and procedures to prevent the misuse of big
data, considering regulatory, and legal risks when
handling social media, geolocation, biometric, and other
forms of personally identifiable information
Big data must be profitable and Monetization is the process of converting an asset such
monetized as data into money by selling it to third parties or by
using it to develop new services
Big data represents natural tensions Big data governance needs to make a balance between
across multiple functions competing objectives across multiple functions, but
how can it be achieved? In this situation, big data
governance needs to bring all the parties together to
determine whether the potential revenue from the new
services outweighs the associated reputational and
regulatory risks
Adapted from Soares (2012, 2013)
utility smart meter readings, RFID readings and GPS signals. Biometric data refers
to the automatic identification of a person based on his or her anatomical or
behavioral characteristics or traits such as fingerprint, a face, an outline of a hand,
an ear shape, a voice pattern, DNA or even body odor. Big transaction data can be
referred to as healthcare claims, telecommunications call detail records, and utility
billing records. Finally, human generated data can be considered as call center
agents’ notes, voice recordings, email, paper documents, surveys, and electronic
medical records.
Optimization Privacy
Policy Monetization
Alignment of
Information Big Data
Governance Multiple
Governance
Functions
Fig. 5.2 Summary of the key components of big data governance. Adapted from Soares (2013)
Fig. 5.3 A high-level

framework for big data
Industries & Functions
governance. Adapted from

Soares (2012, 2013)
Big Data Governance
Big Data Types
5.1.2 Information Governance Disciplines
It can be said that information governance is the glue that drives value and
moderates risk for big data exploitation in organizations. It enables an organization
to leverage information as a competitive asset.
Information governance, in essence, is the set of principles, policies and pro-
cesses by which an organization ensures that information is protected and aligned
with its needs and objectives (Bonenfant et al. 2012; Eisenhauer and Young 2012;
Newman and Logan 2008; Santovena 2013; Steele and Nash 2012; Tsai et al. 2011;
Big Data Now 2012; Oracle 2011). Information is an organizational asset that
includes both data and the context by which that data has meaning (Soares 2013).
Eisenhauer, the chair of Data Governance Society (DGS) and Rick Young the
managing director of 3Sage consulting company consider the information gover-
nance framework as a pyramid with corporate strategy at the top and IT infra-
structure (i.e. data capture and storage) such as enterprise platforms (CRM, ERP,
Portal), data base management and network systems at the bottom (Fig. 5.4).
Corporate Strategy
What does the organization want to achive?
Corporate Information Strategy
What information do we need to achieve our goal and what do
we intend to do with such information?
Information Governance
How to protect and align the information in accordance with our
business requirements?
IT infrastructure
How can we create, manage and configure enterprise platform
to support our business?
Fig. 5.4 Information governance. Adapted from Eisenhauer (2012)
Taking the above issues into account, information governance encompasses

several disciplines, the core among them are illustrated in Fig. 5.5. The traditional
disciplines of information governance also apply to big data governance and require
several disciplines of “Data Management” and their respective solutions in order to
deliver value. These disciplines are organization, metadata management, security
and privacy, data quality, business process integration, master data integration,
and information lifecycle management (Ballard et al. 2014; Santovena 2013; Soares
2012, 2013). In the following, each of these areas will be explained briefly.
Fig. 5.5 Core disciplines of

information governance
Organization
Information
lifecycle Metadata
management
Information
Governance
Disciplines
Master data Privacy
integration
Business
process Data quality
integration
Organization The information governance organization needs to consider add-

ing big data to its overall framework such as organization structure, roles and
responsibilities. In doing so, the information governance council might seek new
members who can provide a unique viewpoint on big data such as data scientists. It
might also decide to appoint stewards for social media or add additional respon-
sibilities to the job description of existing stewards.
Metadata management The big data governance program needs to integrate big
data with the enterprise metadata repository for example by adding big data terms
within the business glossary. In this respect Ballard et al. (2014) explains:
“providing context around content through tagging, using the wealth of available
metadata, and implementing governance processes can rapidly accelerate progress
in this area” (p. 17).
Security and privacy Enhancing security and maintaining privacy is an
emerging concern for the majority of, if not all, big data projects. There are many
discussions as to how big data affects security- and privacy-related matters such as
legal, moral, safety, and political aspects in organizational settings. In many of these
areas, the consequences are only beginning to emerge. Big data governance needs
to identify sensitive data and establish policies about its acceptable use. Although
organizations must comply with security and privacy regulations, nevertheless,
there are few regulations governing the protection of new data categories, such as
geolocation.
Data quality Data quality management is a discipline that includes the methods
to measure, improve, and certify the quality and integrity of an organization’s data
(Batini and Scannapieco 2006; Huang et al. 1999). Indeed, quality is relative to the
business requirement, which is drawing from governance principals to provide data
that is fit to purpose of the organization.
Business process integration The program, initially, needs to identify key
business processes that require big data and then needs to define key policies to
support governance of big data (Soares 2012). For example, drilling and production
are key business processes within the oil and gas industry. The governance program
must establish policies around the retention period for sensor data such as tem-
perature flow, pressure, and salinity on an oilrig.
Master data integration Master data is the high-value information that is used to
support critical business processes across the enterprise such as information about
products, materials, employees, partners, and accounts to name a few. Many
organizations imagine that they want to analyze social media to find out customers’
expectations, likes and dislikes, but do they know who their customers are? Do they
know their best customers? (Ballard et al. 2014). Therefore, the big data governance
program needs to identify and design policies with regard to the integration of big
data into the master data management environment (Ballard et al. 2014; Bonenfant
et al. 2012; Dyche 2012; Mohanty et al. 2013; Morabito 2014c; Sathi 2012; Soares
2012, 2013).
Information lifecycle management The increase in big data volumes make the
situation challenging for organizations to understand the regulatory and business
requirement that determine what data to retain in operational and analytical systems,
Table 5.3 The description of supporting disciplines in information governance

Supporting disciplines Description
Classification and Since big data rely heavily upon large volumes of
metadata unstructured data, it is assumed that data classification takes
on a lesser role. The reality however is quite different. The
users of the information rely upon segmentation to create
order out of chaos and to enhance the insight that the data
provides. As a result, this situation requires improved
capability to separate the unstructured content, then rearrange
and compare them in a new and meaningful manner
Audit information logging Sadly, most organizations consider the importance of this
and reporting capability when the situation becomes problematic; when they
realize they are unable to trace back to the root cause of the
corrupted data. Often, there are issues with no audit
information logging for the data, or the time stamps are not
captured and retained at all. As a result, the slowly changing
nature of data cannot be tracked and reviewed
Data architecture Some of the factors that affect the design of your architecture
include integration of big data sources, consideration of
suitable data governing functions such as information security
and privacy, information quality validations as well as storage
and separation of big data to enhance its consequent use
Adapted from Ballard et al. (2014)
what data to archive and what data to delete. With IT budgets continuing to be under
pressure, over managing information is a gross waste of capital resources. Conse-
quently, the program needs to expand the retention schedule to include big data
based on regulations and business requirements (Morabito 2014a). Therefore, the
big data governance program needs to influence compression and archiving policies,
tools and best practices to reduce storage costs and to improve performance.
However, while Soares (2012) considers all the above areas as core disciplines
of information governance, Ballard et al. (2014) divides the key areas into two
categories of ‘Core Disciplines’ and ‘Supporting Disciplines’. In this classification,
the core disciplines include areas such as: information security and privacy,
information lifecycle Management and data quality management, which highly
overlaps with Soares’s definition of these key aspects. The supporting disciplines
(shown in Table 5.3), however, include domains such as classification and meta-
data, audit information logging and reporting, and data architecture.
5.1.3 Industries and Functions
Considering now big data analytics, they are determined by use cases that are
specific to a given industry or function. Such industries can be customer service,
insurance, retail, healthcare, marketing, risk management, information security,
information technology, and human resources.
Healthcare Industry
• Solution: Sentiment analysis

• Big data type: Web and social media
• Related disciplines: Privacy
• If someone posts a complaint on Twitter, the health plan might want to
first post a limited response and then move the conversation offline.
Telecommunications Industry
• Solution: Location-based services

• Big data type: M2M data
• Related disciplines: Privacy
• The marketing team wants to sell geo-location data to third parties to
offer coupons to subscribers. However, the privacy department is
concerned about the regulatory risks associated with sharing such data.
Fig. 5.6 Different business use cases of big data analytics. Adapted from Soares (2012)
Figure 5.6 illustrates two different use cases in the areas of healthcare and
telecommunication. The figure shows how big data analytics differ from one
industry to another.
5.2 Big Data Maturity Models
Many data governance maturity models are in use, but they all share one key
characteristic (Krishnan 2013; NASCIO 2009; Newman and Logan 2008; Sathi
2012; Oracle 2011): the organization’s climb to the top level is challenging, and
sustaining that level will be even more challenging in the case of big data. This
section focuses on some of these models in order to introduce the reader with
some of the key concepts underlying the big data evolution topic. Reviewing and
evaluating maturity models should occur early in the process in order to form an
understanding of the end state. Furthermore, it must be noted that the delivery
process is an ongoing enterprise operating discipline, which fits under the greater
umbrella of enterprise architecture (Hurwitz et al. 2013; Russom 2008, 2013).
5.2.1 TDWI Maturity Model
The first maturity model is The Data Warehousing Institute (TDWI) maturity
model. TDWI explains that the many goals and tasks related to data governance can
be refined down to four main imperatives or requirements, which in turn group into
a pair of organizational imperatives and a pair of technical ones, namely (Mohanty

et al. 2013; Russom 2008), as shown in Fig. 5.7.
Data governance imperatives in this classification have two key characteristics.
First, each data governance imperative has lifecycle stages, which unfold over time,
and progression through them means a form of development. Second, the imper-
atives as a group imply a time sequence. For instance, it is apparent that imperative
1 must create a cross-functional team before imperative 2 can actually harmonize
team goals with business initiatives.
As to the imperatives 3 and 4 Russom (2013) explains: “less obvious, however,
is that imperative 3 should be governing IT systems before imperative 4 starts using
IT systems to automate governance processes. Although dependencies like these
determine an order for commencing the imperatives, the imperatives must even-
tually coexist and interact” (p. 3). Taking these issues into account. The TDWI
Maturity Model is illustrated in Fig. 5.8 and described in Table 5.4.
The idea behind the TDWI Maturity Model is to show a course of development/
evolution for organizations that need to know where to start and where to go with a
particular type of initiative. The Gulf and Chasm stages represent the difficulties
and challenges that early- and later-stage programs encounter during their BI
journey, respectively. Organizations in the Gulf struggle with sponsorship, funding,
Imperative 1/ Organizational
• Maintain a Cross-Functional Team and Process
Imperative 2/ Organizational
• Align with Data-Intense Business Initiatives
Imperative 3/ Technical
• Govern Data Usage via Technical Implementations
Imperative 4/ Technical
• Automate DG Process via Technical Implementations
Fig. 5.7 Four imperatives of data governance. Adapted from Russom (2008, 2013)
Prenatal & The Child & The Adult & Sage

Infant Stages Gulf Teenager Stages Chasm Stages
Fig. 5.8 TDWI maturity model. Adapted from Russom (2013)

5.2 Big Data Maturity Models 93
Table 5.4 Description of stages in TDWI maturity model

Stages of maturity Description
Prenatal and infant In this phase, organizations rely heavily on manual means applied in an
stages ad hoc manner as a low-end solution to their business or technological
concerns. By moving towards the infant stage and over time, a study of
requirements has led the organization to a specific technology or
practice
The Gulf Jumping this obstacle depends on the organization institutionalizing the
solution concepts it has already verified in the previous step
Child and teenager At this stage, the organization expands its technological initiatives or
stages practices it recently committed to. However, growth could emerge in a
limited context, such as a handful of divisions. As a result, growth
slows down in the teenager stage
The Chasm Crossing the chasm successfully requires dramatic global changes, like
enterprise adoption or solution re-architecture
Adult and sage In the adult stage best practices and technology implementation details
stages established and developed in the teenager stage continue to mature. The
silo deployments of the child stage are gone, replaced in the sage stage
by centralized organizational control and technology integration
Adapted from Russom (2013)
data quality and project scope (TDWI Blog Contributors 2014). Companies in the
Chasm, on the other hand, struggle with the politics, logistics, and dynamics of
delivering an enterprise BI environment (TDWI Blog Contributors 2014). Although
each organization must customize it according to its own situation and needs, the
maturity model can ultimately help with preparation and provides an objective
yardstick for evaluating the current state of an initiative.
5.2.2 Analytics Business Maturity Model
Sathi (2012), by taking a different approach with regard to the TDWI one, intro-
duces the analytics business maturity model (with five levels of maturity) and
explains that this model allows organizations to specify current and target levels of
maturity and what can be achieved in each phase. The stages are as follows: ad hoc,
foundational, competitive, differentiating and breakaway (Fig. 5.9).
At the ad hoc stage the company is generally falling behind other companies in
the corresponding sector information. This stage is an initial phase in which the
company starts to develop the capability to gather consistent information from its
key functional areas. Beyond basic reporting is not available or possible and time-
consuming, manual efforts are generally required to gather the information needed
for day-to-day business decisions.
Considering the foundational stage, the company is still unable to gather key
information and falling behind the majority of its competitors. Indeed, information
Breakaway
Differentiating
Competitive
Foundational
Ad hoc
Fig. 5.9 Analytics business maturity model. Adapted from Sathi (2012)
is not consistently available or utilized to make enterprise-wide business decisions

and a degree of manual efforts to gather information is still required. On the
contrary, competitive stage represents companies whose capabilities in gathering
key information generally are in line with the majority of similar companies.
This maturity level is also the starting point to establish some consistency in key
business metrics across the enterprise.
Once at a differentiating stage, a company whose execution of key business
strategies through utilization of information is considered as better than most other
companies in its sector. Management has the ability to adapt to business changes to
a degree and business leaders as well as users have visibility to key information and
metrics for effective decision-making.
Finally, when at the breakaway stage, a company is generally considered to be
the best in the class in its execution of key business strategies, where information is
utilized enterprise-wide for multidimensional decision making and key predictive
performance indicators are used in modeling for outcomes.
5.2.3 DataFlux Data Governance Maturity Model
DataFlux is another comprehensive framework for data governance maturity models.

According to this model, as the organization moves through the sequence from stage
one to stage four, the value harvested increases and simultaneously the risk asso-
ciated with “bad data” decreases (Table 5.5). Tony Fisher, President and General
Manager of DataFlux explains that the term ‘data governance’ can get gray in
different conversations such as in big data governance, data management, knowledge
management, data assets and information assets (Data et al. 2013; Morabito 2014c;
NASCIO 2009; Soares 2012, 2013; Tallon 2013). DataFlux has improved its
maturity model to highlight a business perspective that drives the need for managing
data as an enterprise asset and to achieve the necessary levels of data quality.
Table 5.5 DataFlux data governance maturity model

Level of maturity Characteristics
Undisciplined There is redundant and duplicate data, divergent sources and records at
(Stage 1) this stage. The main problem is that bad data and information will lead
to bad decisions, and lost opportunities
Reactive (Stage 2) This is the beginning of data governance. Improvements are
experienced at the department level
Proactive (Stage 3) The enterprise recognizes the value of a unified view of information and
therefore begins thinking about Master Data Management (MDM). IT
and business groups collaborate at this stage. At this phase the culture is
also preparing to change
Governed (Stage 4) Information is unified across the enterprise. Business needs drive IT
projects. The enterprise has a sophisticated data strategy and framework.
Employees have realized that information is a key enterprise asset
Adapted from NASCIO (2009)
5.2.4 Gartner Maturity Model
Gartner maturity model (Fig. 5.10) makes the point that enterprise information
management (EIM) is not a single project per se. Rather, it is a long-term program
that evolves over time. Gartner developed the maturity model to give insight and
provide guidance to organizations that are serious about managing information as a
strategic asset (NASCIO 2009; Newman and Logan 2008).
While this maturity model shows the steps in the maturity process and appli-
cation of big data governance, it also presents action items for each level of maturity
(Table 5.6). Gartner’s EIM concept presents an integrated approach to managing
information assets and has five major goals including: unified content, integrated
master data domains, seamless information flows, metadata management and
semantic reconciliation, and data integration across the IT portfolio.
Unaware Aware Reactive Proactive Managed Effective
Fig. 5.10 Gartner maturity model. Adapted from Morabito (2014c) and NASCIO (2008, 2009)
Table 5.6 Action items related to each stage of maturity in Gartner maturity model
Level of maturity Features (Action items)
Unaware • Strategic choices are made without adequate information
• Lack of information architecture, principles, rules and procedures for
sharing information
• Lack of information governance, privacy, security and accountability
Aware • Recognizing the value of information
• Starting to realize the risks associated with not accurately managing
information and information sharing
Reactive • Business understands the value of information as an organizational asset
• Information is shared on cross-functional projects
• Taking steps towards cross-departmental data sharing
Proactive • Governance roles and structure becomes formalized
• Data governance is incorporated with systems development
methodology
Managed • Policies are developed for achieving consistency and accepted
throughout the organization
• Governance organization is in place to address concerns related to cross-
functional information management and sharing
Effective • Top management sees competitive advantage to be gained by properly
exploiting information assets
• EIM strategies link to risk management and productivity targets
5.2.5 IBM Data Governance Maturity Model
Data governance has gained such importance that IBM has formed a Data Governance
Council to define and issue master data quality policies (Mohanty et al. 2013). One of
the initiatives from this council is the ‘data governance maturity model’ based on the
Software Engineering Institute (SEI)) Capability Maturity Model (CMM) (Fig. 5.11).
IBM data governance council explains that there are currently eleven data
governance domains, which reside within four major categories: Outcomes, Ena-
blers, Core Disciplines, and Supporting Disciplines (Table 5.7 (NASCIO 2009)).
Accordingly, business outcomes require enablers and enablers that are supported
through core and supporting disciplines (see also the above discussion on infor-
mation governance core disciplines).
Quantitatively
Initial Managed Defined Optimizing
Managed
• Process Poorly • Process • Process • Proces • Focus on
controlled and characterized characterized quantitatively continuous
reactive for projects for the measured and process
and is organization controlled improvement
managable and is
proactive
Fig. 5.11 IBM Data Governance Council maturity model. Adapted from NASCIO (2008, 2009)
Table 5.7 The four major groups in IBM data governance council maturity model
Outcomes Core disciplines
• Data risk management and compliance • Data quality management
• Value creation • Information life-cycle management
• Information security and privacy
Enablers Supporting disciplines
• Organizational structures and awareness • Data architecture
• Policy • Classification and metadata
• Stewardship • Audit information logging and reporting
As described earlier, maturity models and frameworks are necessary tools for the
data governance program. The maturity models describe the relevant milestones in the
process of big data governance. While such frameworks present concepts and the
relationships among the concepts, a methodology is needed by which organizations can
move from one level to another and travel up the maturity model. Having introduced
some of these models here, it should be said that there are several other maturity models
such MDM Institute Data Governance Maturity Levels, Oracle Data Governance
Maturity Model (NASCIO 2008, 2009) and so on that can also be looked at. This sub-
section attempted to introduce the readers with some of the key concepts of maturity
models in the area of big data/ information governance. Maturity models provide a
measure for enterprises to measure their success in managing data and recognizing
information as an enterprise asset. As for these issues, in what follows, a discussion is
provided on the organizational challenges and benefit related to big data governance.
5.3 Organizational Challenges Inherent with Governing

Big Data
While big data can yield exceptionally useful and valuable information, they also
present new challenges with regard to how much data to store, how much this will cost,
whether the data will be secure, and how long it must be maintained (Ballard et al.
2014; Michael and Miller 2013; Russom 2013; Santovena 2013; Soares 2013; Zicari
2012). For example, inaccurate, incomplete or fraudulently manipulated data pose
increasing risk as enterprises become more dependent on the data to drive decision-
making and assess results (Craig 2011; Fischer 2013; Hurwitz et al. 2013; Jee and Kim
2013). Whether it is a SAP ERP system or a big data database, top management must
think about how bad data can and will get into their system. Richard Neale who is a
business intelligence industry veteran points to the importance of attention to data
quality and explains (Interview: The Need For Big Data Governance 2014):
There’s no silver bullet, no shortcut to fantastic business insight just because you’re using a
new technology. It may sound boring, but you have to pay attention to data quality, and cut
bad data off at the source using a “data quality firewall” approach to your big data
repository. (Interview: The Need For Big Data Governance 2014)
He then refers to the newer, less-structured data such as social media feeds that
complicate the notion of “data quality”. Other commonly cited barriers organiza-
tions must overcome to implement data governance policies and procedures
(NASCIO 2008) are as follows:
• Defining requirements and interpreting and understanding the strategies and

procedures concerning data sources
• To encourage agreement of all parties regarding appropriate rules and policies
• Developing new tools to help governing big data
• The cost of implementing policies for sustainable data governance
• Incompatible systems and IT infrastructures
• Divergent and opposing priorities within the organization
• Getting management to understand what is necessary
Data quality (see dimensions in Table 5.8) is just one of the key components of
the information governance program. Data quality is about how good is the data?
How broad is the coverage? How fine is the sampling resolution? How well
understood are the sampling biases? A good process will, naturally, make bad
decisions if based upon bad data.
Another issue is compliance and security, as also noted in the previous sections.
Big data governance helps organizations to protect their sensitive and strategic
information such as intellectual property, business plans and product designs, key
performance indicators, sales figures, financial metrics and production metrics used
to make critical decisions. The following list was also presented in
SearchDataManagement.com as the key reasons that data governance projects fail
(NASCIO 2008; Oklahoma 2013):
• Cultural barriers
• Lack of top management support
• Underestimating the amount of work involved in data governance projects
• Lack of understanding that business definitions vary
• Trying to move too fast from no-data governance to enterprise-wide data-governance
Table 5.8 Definition of data quality dimensions

Dimensions Characteristics
Accuracy Data precisely reflects the object or transaction it describes
Reliability Data is consistent across multiple transactions
Credibility The degree to which decision makers trust both the accuracy and reliability
of data
Timeliness Data is available to the information consumer when it is needed
Appropriateness The degree to which the data itself is relevant to the needs of an
organization
Completeness All of the relevant or required data is readily available for use when
required
5.3 Organizational Challenges Inherent with Governing Big Data 99
Data governance in essence can be viewed as a formal orchestration of people,

process, and technology to enable an organization to leverage data as an enterprise
asset. Without a proper data governance process, big data projects can lead to a lot of
trouble such as misleading data and unexpected costs (Jee and Kim 2013; Morabito
2014c; Mouthaan 2012; Steele and Nash 2012; Warden 2011). By controlling how
information is created, shared, cleansed, consolidated, protected, maintained, retired
and integrated within your enterprise, information integration and governance
strategies turn uncertain data into trusted and insightful information (Bonenfant et al.
2012; Dyche 2012; NASCIO 2008; Sathi 2012; Soares 2012, 2013; Best Practice
Guidelines 2013). Analytic and operational applications can overcome the dual
challenge of rising uncertainty in data combined with overwhelming growth in the
volume and variety of information. Data governance programs provide a framework
for setting data-usage rules as well as implementing controls designed to ensure that
information remains accurate consistent and accessible (Fischer 2013; Santovena
2013; Soares 2013; Zicari 2012).
5.4 Organizational Benefits of Governing Big Data
Why do data governance? Basically, organizations do data governance for three

main reasons (Kruger and Foster 2013):
(1) to increase their revenue and/or value by improving the quality of data,
(2) to manage data complexity by addressing data integration issues, reducing data
redundancy and reducing organizational and/or systems complexity, and
(3) to ensure survival through attention to risks by supporting privacy, formalizing
internal checks-and-balances and avoiding data disasters.
Based on 2,583 responses from 461 respondents with respect to ‘big data
management’, Russom (2013) reports that the vast majority of them consider big
data management (BDM) an opportunity (89 %). Conventional wisdom today says
that big data enable data exploration and predictive analytics to discover new facts
about customers, markets, partners, costs, and operations (Russom 2013). A tiny
minority of the respondents to the survey reported by Russom, however, consider
BDM a problem (11 %). Undoubtedly, big data offers technical challenges due to
their size, speed, and diversity. According to the survey, for example, data volume
alone is a showstopper for a few organizations. To determine the most compelling
reasons for big data management the survey asked respondents: “If your organi-
zation were to successfully manage and leverage big data, which business and
technology tasks would improve?”. At the top of the list, survey respondents
selected data analytics (61 %) more than any other answer. According to survey
respondents, common analytic applications can benefit from BDM, including fraud
detection (21 %) and risk quantification (16 %). Respondents also mentioned that
BDM delivers value to the business. This is borne out in the ranking of survey
responses, which place near the top of the list business value from big data (33 %)
and numerous and accurate business insights (34 %). Similar benefits of BDM
include business optimization (28 %), addressing new business requirements
(22 %), and understanding business change (22 %).
According to the survey reported by Russom (2013), big data’s large data
samples and diverse range of data sources can lead to broader data sourcing for
analytics (32 %), more data for data warehousing (24 %), and improved data
staging for data warehousing (23 %). Finally, according to the survey sales and
marketing activities improve with BDM. These include the recognition of sales and
market opportunities (28 %), definitions of churn and other customer behaviors
(18 %), better targeted social influencer marketing (16 %), and understanding
consumer behavior via clickstreams (16 %), as well as related analytic applications
such as customer-base segmentation (27 %) and sentiment analytics and trending
(24 %) (Russom 2013).
5.5 Case Studies
This section discusses three sample examples to establishing a business case for big
data governance: (1) quantify the financial impact of big data governance on
patients’ privacy, (2) the value of big data quality, and (3) the impact of data quality
and master data on big data initiatives (Soares 2012).
Case 1: Quantify the financial impact of big data governance on patients’ privacy
A major health insurer had over 50 unencrypted hard drives stolen from a call
centre facility, which had been closed. The stolen drives contained big transaction
data including sensitive information such as patient names, social security numbers,
date of births and diagnosis codes.
Point of Attention: One advantage that a big data governance program gives
is that of adhering to the privacy, security, financial standards and legal
requirements.
The insurer reached a settlement for $1.5 million with the United States
Department of Health and Human Services. Furthermore, the insurer spent more
than $17 million to notify the more than one million members who were affected. In
order to prevent recurrent of a similar event the health insurer implemented data
encryption on all types of big and small data, including 1,000 servers, 6,000
workstations and 136,000 volumes of backup tapes (Soares 2013).
Case 2: The value of big data quality

The big data governance program needs to establish an operational control to ensure
the accuracy of big data to underpin and support analyses. For example, geospatial
data are critical to exploration and production companies in the oil and gas industry.
These data might be about land-based drilling and offshore drilling as well as wells
5.5 Case Studies 101
Table 5.9 The business benefits from using sentiment analysis and sound product master data
A. New product line value at list price $10,000,000
B. Discount at the end of the season 70 %
C. Start of the season discount 30 %
D. Difference in discounting levels (B − C) 40 %
E. Gross profit benefit as a result of better market trends analytics A × D $4,000,000
Adapted from Soares (2013)
that might be abandoned. There are several examples of poor geospatial data
governance that have had an intense business impact. An oil company used
incorrect geospatial coordinates to drill a well at the exact location of an abandoned
well. Because of this inaccuracy, the crew had to reposition the rig and start work
all over again which resulted in losses of millions of euros.
Point of Attention: Big data governance is essential for maintaining the

quality of big data analysis, allowing managers to make decisions and take
action with confidence.
In another instance, the Exploration and Production (E&P) teams used incorrect
navigational coordinates to drill a hole in an adjacent field that belonged to another
company.
Case 3: the impact of data quality and master data on big data initiatives
A popular global retailer was experiencing reduction in product profit margins due to
increased promotional activity (Soares 2013). To address this business concern, the
retailer decided to collect and analyze product feedback from customers in social media
websites such as, e.g., twitter to determine the pricing strategy for new products.
If the sentiment analysis (Morabito 2014b) was not very promising during the
product launch, the retailer would decide to update its pricing in the master product
catalog and offer discount of 30 %.
This would replace its usual strategy of selling merchandise at the end of season
at the discount of 70 %. As a result, the retailer was able to improve its profit
margin (see Table 5.9).
Organizations need to establish a big data governance program to ensure clarity on

various aspects of data access, integration, usage, management and ownership of
big data management (Ballard et al. 2014; Craig 2011; Fischer 2013; Michael and
Miller 2013; Mohanty et al. 2013; Morabito 2014c; Mouthaan 2012; Santovena
2013; Soares 2012; Big Data Now 2012). Maturity models, as mentioned
earlier, are one of the widespread areas in the field of improving organizational
performance as well as process management capabilities.
As with many other organizational initiatives, successful implementation of big
data governance requires an enterprise perspective (NASCIO 2008, 2009). A sys-
temic perspective on information governance is therefore needed to fully exploit the
benefits coming from big data, in order to assure that organizations can answer
these key questions (Big Data Governance 2014):
• What are the data sources?

• Which are the most valuable among the data sources?
• What are the requirements by line of business or division?
• What is the purpose, definition as well as usage of each data attribute?
• Do you need data stewards? What can they do for the organization?
• Is it possible to integrate big data into existing infrastructure?
• How accurate does the data need to be?
• What rules, policies and strategies are required for managing, monitoring,
stewardship, privacy and security?
• What are the permissions for sharing data?
Other key questions could be: How do we verify the authenticity of the informa-
tion? Can we verify how the information will be used? What options do we have as for
the decisions to be taken? What is the context for each decision? How will organi-
zations protect all of their sources, processes, and decisions from theft and corruption?
Without a clear framework for big data governance and use, none of the above
questions can be properly answered. Furthermore, businesses run the risk of becoming
paralyzed under a chaotic mixture of data, much of which has become obsolete and
past its expiration date. Indeed, big data governance helps the business to get a clear
understanding of what the organization wishes to achieve from its data sources. Data
governance frameworks helps organizations govern appropriate use of and access to
critical information such as customer information, financial details and unstructured
content, measuring and reporting information quality and risk to enhance value and
mitigate exposures. Without data governance, it becomes difficult for organizations to
maintain their competitive edge. The implemented technology may function exactly
as designed, but the information deriving from the technology may be suspect.
5.7 Summary
This chapter attempted to introduce the readers with several guiding principles for data
governance in a big data environment. Indeed, organizations need to take the right step
in time to get the best out of enterprise data and big data. Without data governance, the
data will be inconsistent, unreliable and unrepeatable. Data governance helps ensure
that metrics are defined consistently within the organization. Clearly documented
standards and definitions mean everybody can understand precisely what everybody
5.7 Summary 103
else is talking about. Data governance provides confidence in key decisions, limit
organizational costs and prevent analysis and reporting issues. Data governance
encourages the measurement of successes and failures.
Furthermore, maturity models provide a framework for organizations to measure
their success in managing data and information as an enterprise asset. Thus, data
governance maturity models can be used and looked at as references in commu-
nication, awareness building, and the marketing of data governance. Big data
governance is more than standards, reporting, and prioritization of projects; it is a
business function providing structure for maintaining high data standards and
securing against the risks of data theft or loss. Within ‘Big Data’ projects, privacy
and regulatory controls play a pivotal role.
Finally, big data governance is not static; it must evolve over time to meet the
changing objectives of the organization. Indeed, with the inclusion of big data, e.g.,
in decision-making or operations, changes may be bigger and, hence, governance
has to be more comprehensive.
References
Ballard, C., Compert, C., Jesionowski, T., Milman, I., Plants, B., Rosen, B., Smith, H.:
Information Governance Principles and Practices for a Big Data Landscape. IBM (2014)
Batini, C., Scannapieco, M.: Data Quality: Concepts, Methodologies and Techniques. Springer,
Heidelberg (2006)
Bertot, J.C., Choi, H.: Big data and e-government: issues, policies, and recommendations. In: The
Proceedings of the 14th Annual International Conference on Digital Government Research,
pp. 1–10 (2013)
Best Practice Guideline: Big Data. A guide to maximising customer engagement opportunities
through the development of responsible big data strategies (2013)
Big Data Governance is Key to Harnessing Insight from Big Data. http://www.firstsanfranciscopartners.
com/practices/big-data/big-data-governance/ (2014). Accessed 16 Nov 2014
Big Data Now: 2012 Edition. O’Reilly (2012)
Bonenfant, M., Ménard, M., Mondoux, A., Ouellet, M.: Big data and governance. Research Group
on Information and Surveillance in Daily Life, GRICIS Research Centre of the University of
Québec, Montréal (2012)
Cline, J.S.: The promise of data-driven care. N. C. Med. J. 75(3), 178–182 (2014)
Craig, F.C.T., Ludloff, M.: Privacy and Big Data. O’Reilly Media, Sebastopol, CA (2011)
Data, B., et al.: Australian Public Service Better Practice Guide for Big Data and Big Data
Analytics (2013)
Data Governance Institute (DGI). http://www.datagovernance.com/adg_data_governance_
definition/ (2014). Accessed 16 Nov 2014
Debra Logan: What is information governance? And why is it so hard? http://blogs.gartner.
com/debra_logan/2010/01/11/what-is-information-governance-and-why-is-it-so-hard/, (2010).
Dyche, J.: The Intersection of Big Data, Data Governance and MDM (2012)
Eisenhauer, J.A., Young, R.: Big Data. 3Sage Consulting Company, Atlanta (2012)
Fischer, U.: Big Data: Impact, Benefits, Risk and Governance. Fischer IT GRC Consulting &
Training Group, Wettingen, CH (2013)
Franks, B.: Taming The Big Data Tidal Wave: Finding Opportunities in Huge Data Streams with
Advanced Analytics. Wiley and SAS Business Series. Wiley, Hoboken, New Jersey (2012)
Huang, K., Lee, Y., Wang, R.Y.T.: Quality, Information and Knowledge. Prentice-Hall, Inc.,
Upper Saddle River (1999)
Hurwitz, J., et al.: Big Data for Dummies. Wiley, Hoboken (2013)
Interview: The Need For Big Data Governance. http://timoelliott.com/blog/2014/01/interview-the-
need-for-big-data-governance.html, (2014). Accessed 16 Nov 2014
Jee, K., Kim, G.-H.: Potentiality of big data in the medical sector: focus on how to reshape the
healthcare system. Healthc. Inform. Res. 19(2), 79–85 (2013)
Krishnan, K.: Data Warehousing in the Age of Big Data. Elsevier, Amsterdam (2013)
Kruger, K., Foster, J.: Big Data Governance. ISACA SA 2013 Annual Conference, Johannesburg,
SA (2013)
Michael, K., Miller, K.W.: Big data: new opportunities and new challenges. Computer 46(6), 22–24
(2013)
Mohanty, S., Jagadeesh, M., Srivatsa, H.: Big Data Imperative: Enterprise ‘Big Data’ Warehouse,
‘BI’ Implementations and Analytics. Apress, New York (2013)
Morabito, V.: Big Data. Trends and Challenges in Digital Business Innovation, pp. 3–21. Springer,
Heidelberg (2014a)
Morabito, V.: Social Listening. Trends and Challenges in Digital Business Innovation SE, pp. 67–87.
Springer International Publishing, Heidelberg (2014b)
Publishing, Heidelberg (2014c)
Mouthaan, N.: Effects of big data analytics on organizations’ value creation. University of
Amsterdam (2012)
NASCIO: Data Governance: Managing Information as an Enterprise Asset Part I—An
Introduction, 859 (2008)
NASCIO: Data Governance Part II: Maturity Models—A Path to Progress, USA (2009)
Newman, D., Logan, D.: Gartner Introduces the EIM Maturity Model (2008)
Oklahoma Office of Management and Enterprise Services: Oklahoma Interoperability Grant Data
Roadmap (2013)
Oracle: Enterprise information Management: Best Practices in Data Governance, Oracle Inc.
(2011)
Raghupathi, W., Raghupathi, V.: Big data analytics in healthcare: promise and potential. Health
Inf. Sci. Syst. 2, 1, 3 (2014)
Russom, P.: The Four Imperatives of Data Governance Maturity. TDWI Monograph (2008)
Russom, P.: Managing big data. TDWI Research. TDWI Best Practices Report (2013)
Santovena, A.Z.: Big data: evolution, components, challenges and opportunities. http://hdl.handle.
net/1721.1/80667, (2013). Accessed 16 Nov 2014
Sathi, A.: Big Data Analytics: Disruptive Technologies for Changing the Game. MC Press, Boise,
ID, USA (2012)
Soares, S.: Big Data Governance. Information Asset, LLC (2012)
Soares, S.: A Platform for Big Data Governance and Process Data Governance. MC Press Online,
LLC, Boise, ID, USA (2013)
Steele, J., Nash, C.: Ethics of Big Data. O’Reilly, Sebastopol, CA (2012)
Tallon, P.P.: Corporate governance of big data: perspectives on value, risk, and cost. Computer 46
(6):32–38 (2013)
TDWI Blog Contributors, Crossing The Chasm Part 1: Delivering Strategic Value. http://tdwi.org/
blogs/wayne-eckerson/2009/10/chasm-1.aspx (2014). Accessed 16 Nov 2014
Tsai, J.Y. et al.: The effect of online privacy information on purchasing behavior: an experimental
study. Inf. Syst. Res. 22(2), 254–268 (2011)
Warden, P.: Big Data Glossary. O’Reilly Media Inc., Sebastopol (2011)
Webb, P., et al.: Attempting to define IT governance: wisdom or folly? In: HICSS’06, pp. 1–10
(2006)
Zicari, R.V: Big Data: Challenges and Opportunities. This is Big Data (2012)
Big Data and Digital Business
Evaluation 6
Abstract
The advances in data and processing technology has radically changed the way
today’s digital business operates. Big data technology provides analytics tools
that allow digital businesses to identify useful insights about their customers and
create products to meet their customer’s requirements. However, digital
businesses working with big data technology face various challenges including
the data acquisition, the storage space and the evaluation of the data as well as
the right choices of the big data applications and the infrastructures. This review
will explore digital business evaluation using big data and the advantages as well
as the challenges of big data technology in digital business.
6.1 Introduction
Innovations in technology and greater affordability of digital devices have allowed

digital business to collect a vast amount of data in digital form. These big data are
often generated from social networking interactions, videos, emails, images, audios,
logs, mobile phones, apps postings and many other sources (Sagiroglu and Sinanc
2013). Several research in the big data topic have proved that, in years to come the
concept of big data in digital businesses will be intensified and diversified as the
amount of data produced increases (Zhang 2013; Zheng et al. 2013; Katal and
Wazid 2013). For example in 2003, 5 exabytes of data information were created
around the world but today this number of data is created in just 2 days (Sagiroglu
and Sinanc 2013). Only in 2012, the digital data produced was extended to 2.72
zettabytes, this made 1021 bytes, which is expected to double every 2 years, making
approximately 8 zettabytes in 2015.
Results from recent studies indicated that, mobile phones have become one of the
main generators of big data, with 6 billion of mobile subscriptions worldwide gener-
ating alone around 10 billion of text messages every day. Digital services such as, e.g.,
Google controls around 7.2 billion pages per day, processes approximately 20 peta-
bytes of data every day, and it has over one millions servers globally (Sagiroglu and

DOI 10.1007/978-3-319-10665-6_6
106 6 Big Data and Digital Business Evaluation
Sinanc 2013; Muhtaroglu et al. 2013). Facebook with 1.3 billion active accounts, every
month received about 140 billion photos with 125 billion friend connections in 70
languages (Wikipedia 2014). Twitter with 271 million monthly active users processes
around 500 million tweets every day (Wikipedia 2014).
These digital businesses have radically changed the way data are used in a digital
world. Over the past years we have seen a great increase in the number of digital
business, around 571 new websites are deployed every minute. However, a big
question remains on how digital businesses use bid data to evaluate and enhance
their productivity. To understand this problem, we will briefly explore the concept
of digital business evaluation using big data. The advantages and challenges of big
data technology in digital business and finally we will provide some case studies
and recommendations for institutions.
6.2 Digital Business Evaluation Using Big Data
Some of the challenges in today digital services are the economic uncertainty, and
the increase on innovation in services or products and new business models. Most
digital business, regardless of the market type, begins with some elements of
research including customer acquisition, products and services development, part-
nerships and merging. In current big data driven competitive market, it is critical to
proactively restructure the business model to push revenue by evaluating all the key
structures of the digital business (Rajpurohit 2013).
Most efficient digital businesses have technology and research procedures in place
that use big data analytics to evaluate their business models by identifying the best
opportunities and excluding the wrong ones easily and quickly. Nevertheless, to
achieve this type of intelligent business innovative technique, it is necessary to
develop a research process that uses concept based factors to allow big data analysis
find opportunities that are relevant to the business’s priorities. For example, a digital
business may want to develop a new service that will target a set of customers in the
market. With real time big data, researchers are capable of identifying patterns that
can estimate whether or not a new service or product are financially viable, which will
make the decision-making efficient and effective (Rajpurohit 2013; Gopalkrishnan
and Steier 2012).
Taking the above issues into account, the analytics searches to identify oppor-
tunities and drive revenue are often iterative and interactive as shown in Fig. 6.1.
This process generally includes six steps (Rajpurohit 2013), that are known and
applied in traditional intelligence activities, but require different application policies
and protocols due to the characteristics of big data:
• Analytics goals: identifying key priorities is a very important step to understand

the business context and efficiently set up analytics goals. For example some
digital businesses may look for new insights, while others are evaluating their
businesses models against competitors.
6.2 Digital Business Evaluation Using Big Data 107
Analytics Selection Processing of

Goals of Data Data
Visualization Evaluation and

& Patterns Data Mining
Feedback Interpretation
Fig. 6.1 Big data analytics process. Adapted from Rajpurohit (2013)
• Selection of data: After defining the goals, it is very important to select the right
data required for the analysis. Since big data are made up of references to a
variety of information, selecting the relevant data allow the analytics processes
to be focused and efficient.
• Processing data: this step is designed to enhance data reliability by defining
missing data, identifying and removing no relevant data or outliers. Since big data
come from many different sources, this step also include the data transformation,
which compiles all the data to one consistent format that can be easily analyzed
by later steps. The complexity of this step often depends on the size and the
variety of the data and the nature of the research. Thus, a good processing step
will simplify later processes of the analysis.
• Data Mining: at this stage, the data is processed to extract meaningful signals
based on the goals of the analysis. Consequently, the complexity of this step
depends on the nature of the business and the goals of the analysis. Most data
mining operations include computational algorithms such as classification,
clustering or regression, to mention a few, which are capable of extracting
patterns that would otherwise remain undiscovered using traditional assessment
methods. A suitable data mining method is the selection based on the business
goals. The performance of the data mining methods generally improves as they
process more data.
• Evaluation and patters interpretation: this consists of evaluating the results of
the analysis. Depending on the goals originated at the first step, the resulting
patterns are evaluated to extract meaningful insights that can enhance the
business performance.
• Visualization and feedback: This step uses advanced visualization techniques

such as tables, lists or graphs to represent the results in a more intuitive way for
non technical staffs or management teams for better understanding of the analysis
outcome. To improve the analysis results, a direct feedback process can be set up
to refine the prediction process.
Business evaluation with big data gives a powerful analysis tool to digital
businesses managers allowing them to understand the market and customer’s needs
in terms of products or services, thus, providing them better decision-making tools.
In the next sections we will discuss the organizational advantages and challenges of
big data application in the digital businesses.
6.3 Organizational Advantages and Opportunities
Most digital companies are dealing with a variety of new data, which arrives from
many different sources (see Fig. 6.2). Companies, which can adapt the big data
technology into their day to day operations have significant advantages over those
without any big data technology capability (Muhtaroglu et al. 2013). Big data
resources present a great opportunity for digital business models, as we have seen
with Google, eBay, Amazon, Facebook and Netflix, Borders and many other
Big Data
Data Analytics $$$
Fig. 6.2 Big data and analytics configuration for business value (Mithas and Lucas 2010)
businesses (Mithas and Lucas 2010). Therefore, the big data technology is often
used in digital business for two main reasons, as follows:
(1) The first reason for big data technology is to provide analytics tools (as seen in
Sect. 6.2) that allow a company to gain useful insights about their business and
guide the management in high-level decision making process.
(2) The second purpose of this technology is to provide tools for customer’s
satisfaction. Big data allow the development of different application platforms
and real-time services and products that leverage a variety of data in order to
satisfy consumer needs and provided value to the business such as efficient,
intelligent service and many others.
The aforementioned argument can be supported by looking at the way organi-

zations invest on Big data. Over the past years global spending on digital business
IT resources (big data) have exceeded 3 trillion US dollar per year (Mithas and
Lucas 2010). However the main questions remain: what is the return from all this
investments? How does big data technology provide tools in order to gain com-
pletive advantages in digital business? To address these questions, it is vital to
highlight and analyze the digital business models in terms of fundamental concepts
such as customer value of proposition, consumer segmentation and key revenue
channels (Muhtaroglu et al. 2013; Osterwalder et al. 2010; Morabito 2014).
6.3.1 Customer Value Proposition
The customer value proposition focuses on how products and services offered by a
company satisfy customer’s needs with regard to competitors’ offering. With the
application of big data technology in digital business, a variety of innovative
products and services can be delivered to add more value or solve customer’s
challenging problems. Recently many innovative applications have been built off
the back of smart use of big data to fulfill customer’s needs, replacing the use of
traditional data models. Examples of those applications are the price comparison
applications, which enable customers to view pricing information on products from
different sources. Many studies have shown that, with these applications customers
can save around 10 % on every shopping they made on digital market (Manyika
et al. 2011). RedLaser is one of the applications that give customers the power to
shop around by scanning the bar code of a product with their smart phone devices
and receive price comparisons for the products and other related products. Other
customer value proposition applications are provided by digital businesses
including eBay and Amazon. They provide tools that allow customers to search a
product from a large number of sellers. These services provide transparency,
producing “buy” or “sale” opportunities for customers and businesses (Manyika
et al. 2011). The application also offers more information about the products
including customer’s reviews and quality of the products.
Also, the big data technology has added great value to digital entertainments
companies such as, e.g., Netflix, whose digital business model is to provide films,
TV entertainment. To establish itself in the digital market, Netflix was able to
design an application that offers consumers searchable films from a large set of
movies and TV series. The service also provided more information about the films
including customer’s reviews and suggestion on related films. The big data appli-
cations have, thus, helped Netflix to provide innovative and efficient services to its
clients and increased its revenue to 4.3 billion US dollars with more than 50 million
subscribers in 2014 (Wikipedia 2014). Considering all these applications of the big
data technology, the customer value propositions of digital business are expected to
improve further in near future.
6.3.2 Customer Segmentation
Delivering the right products or services to the right customer groups can signifi-
cantly increase the profitability. Equally, delivering products or services to wrong
customer groups will turn them away from an organization to its competitors.
With the success of modern technology, the Internet usage has dramatically
increased; digital market is becoming more demanding and sophisticated. In this
complex market customers are demanding more from their services providers and
with an excessive competition, digital businesses that are unable to fulfill theses
expectations are swiftly left behind by the competition. Understanding your
customers, products and services have never been more important in digital business.
Customer segmentation based on big data applications allows digital businesses
understand key characteristics and behaviors of their customers, enabling them to
design and deliver products and services, which are more likely to be relevant to a
target customer segments. Complete relevance can lead to greater conversion rates,
which eventually will lead to more profitability (Experian Hitwise 2010).
In a large group of customers there are different type of customers and each type
can have many characteristics that define them. Using segmentation based on the
big data information analytics allows a digital service to group these individuals by
similar characteristics. Simple examples of this are dividing your customer group
into female and male customers or young and old customers; or more complex such
as developing a search engine to locate a group of individuals who have express
some interests in your new products. With the customer segmentation tools based on
big data technology, digital businesses can have a greater flexibility in developing
products or services depending on a specific demand of the target segment or the
most profitable customers groups.
The big data applications in customer segmentation are very wide and depending
on how niche a digital business want each segment to be. However, there are four
main types of data in the big data including demographics, lifestyle, behavior and
value, that can be used to perform the customer segmentation (Experian Hitwise
2010).
• Demographics: customer segmentation according to demography is mainly

based on characteristics such age, gender, ethnicity, race. These characteristics
cannot depend on an individual willingness or action.
• Lifestyle: consists of characteristics that can be influenced by an individual.
These range from household income, locations, entertainments, attitudes, beliefs
and so on.
• Behavior: defines the decision making processes of customers, it also help to
understand how emotions affect customer’s buying behavior as well as their
usage rate and response with regard to a given product. For examples, how often
a customer visits websites or social networks to look for the products or brands
she or he like.
• Value: this process uses past transactional information of a customer to identify
how profitable a customer is to the company. The process defines the buying
power of a customer.
In general the segmentation process help digital business to target groups of

customers who are likely to respond to given products or services offerings. These
groups must have similar attributes and each group should have a characteristic that
makes it different from others. Each group must be large enough to generate sig-
nificant additional value to the business.
6.3.3 Channels
Channels constitute different ways on how a digital business gets in contact with its
clienteles and provides them value propositions. The big data driven applications of
digital businesses collect information and reach customers through many different
channels. These channels can be used as mediums to deliver products as well as
collecting information from customers. Furthermore, the applications are generally
delivered through two main channels, web based applications and/or mobile
applications. The web based applications for desktop and notebook users can be
delivered through the product website such as, e.g., eBay and Amazon, which offer
services that allow customers to search products from a large number of merchants
and provide customers reviews for more transparency. Whereas the mobile appli-
cations can be delivered using stores like Google play (2014) and Apple Store
(2014). The increase in mobile applications will allow more innovations and
efficient delivering of big data driven applications through marketing channels
(Manyika et al. 2011).
6.3.4 Customer Relationship
The customer relationship is defined as the relationship a digital business builds and
preserves with its clienteles (Manyika et al. 2011). With the growth of competitions
and technology such as Internet and mobile applications, a digital business lives or
dies based on how it build relationship with the customers. In the past, digital
business often relied on broadcast mediums to build and maintain relationship with
customers, while today’s technology allows customers to connect and interact with
each other around the world, which has dramatically changed the customer’s
relationship in digital market. To succeed in today digital market, companies need
network systems beyond traditional marketing and sales advertising messages.
The big data applications provide a network tools that take advantage of the
customers interaction with the technology to create and manage relationships with
customers or clients that goes beyond simple marketing and sales promotional
offers. As discussed above, these applications may include automated systems that
can recognize some characteristics of a group of customers and provide information
associated to transactions. For example, a tool in an online retail business may
recommend or suggest products an individual customer may be looking for based
on information of the customer’s previous purchases (Manyika et al. 2011).
These applications may also offer products information and customer’s feedback
on individual products, which enable businesses to become a part of their cus-
tomer’s conversations (Rogers 2011). Figure 6.3 shows a summary of few
advantages of big data in digital business.
Better
Products and
Services
Tailor Products Understand

or Services Market and
Customers
Advantages of Big
Personalize Data in Digital Improve
Customers Business Capabilities
Improve
Minimize Risks Decision-
Making
Faster
Decisions
Fig. 6.3 Big data advantages. Adapted from Rogers (2011)

The increased diffusion and adoption of smart phones technology and social net-
works have changed the way digital businesses operate. To be a major player in the
digital market, many digital businesses have adopted big data technologies. Indeed,
taking advantage of the customer’s interaction with the technology and with each
other on a global scale, digital businesses have managed to target customers with
the right products and services. This has provided them with a better understanding
of their customers, products, hence made them more efficient and increased their
profitability (Experian Hitwise 2010; Rogers 2011).
However, digital businesses working with big data technology face various
challenges, including the acquisition of the data, the storage space, the right choices
of the big data applications and the infrastructures. Businesses using big data
technology need to understand both the needs of the technology and customer’s,
because choosing a big data application without knowing its advantages may result
in a waste of resources (Rogers 2011). Digital businesses need to understand their
clienteles and identify their businesses goals before adopting any big data appli-
cations. Whether you are an online retails or provide digital technology, having
knowledge about your customer’s behavior and coming up with new ways to add
values in your products or services are the key drivers to succeed in today’s digital
market (Rogers 2011).
Addressing challenges presented by big data technology will be difficult con-
sidering the volume of the ever-increasing data and its variety. Currently available
big data tools, analysis methods, technology architecture and management are not
fully capable in dealing with the complexity of the big data (Katal and Wazid
2013). In what follows, some of the main challenges in the big data application in
digital business are briefly discussed.
6.4.1 Key Resources
Key resources refer to data, infrastructures, analytics and skilled people a digital
business needs in order to provide value to its clienteles through big data appli-
cations (Morabito 2014). The data access and management can be a great obstacle
for many digital businesses. Due to the high volume of data and privacy laws, a
majority of digital businesses do not have the right platforms for data collection and
management across the enterprise. Moreover, the success of big data technology
relies on its capability of assessing data in a timely manner. However, designing
tools for storing and managing real time data is very complex and require time and
money. Therefore, digital companies might need to develop partnerships with other
companies to share the costs associated with data acquisition. This form of data
collection, however, may threaten the culture of competitiveness and secrecy (Katal
and Wazid 2013). Other strategies to mitigate costs related to the increase in volume
of data, is to use cloud technology. However, uploading terabytes of data in real
time will require a large amount of time (Katal and Wazid 2013). Accordingly,
storing and accessing large data in timely manner as well as reducing the processing
time remain some of the biggest challenges in big data technology (see again
Fig. 6.1, showing the processing involved in big data analysis).
Furthermore, a company might also face challenges in identifying the right data
and defining advantages the data have to offer. Because data related business models
are often different to the traditional business revenue models, as discussed above.
This type of analysis will require more sophisticated analytics tools and highly skilled
people with statistics and computing knowledge and capabilities, which in turn, is one
of the biggest challenges of big data projects. There is a shortage of qualified skilled
people capable of understanding and interpreting raw data using technology to find
some meaningful business insights (Davenport and Patil 2012).
The big data applications also require a more sophisticated technology landscape
due the growing nature of different data sources (Morabito 2014). Therefore,
leveraging big data requires strong collaboration with more innovative technology
services that can help to develop the right IT infrastructures capable of adapting in
efficient manner to the changes occurring in the technology landscape. Consequently,
many digital businesses actually require assistance from companies offering plat-
forms and infrastructures solutions. Because big data technology needs continuous
maintenance and development of different infrastructures and platforms, especially
when the number of customer’s increases and the volumes of the data get even
bigger. In addition to companies offering platforms and infrastructures, several
middle class servers have developed and employed together with existing computing
platforms like Hadoop (Katal and Wazid 2013), an open source project by Apache
Software Foundation. The aim of this project is to provide a reliable and efficient
share storage capacities with robust analytics systems. This will significantly
improve the processing of data, which is considered to be one of the big issues in big
data analysis (Katal and Wazid 2013).
6.4.2 Privacy and Security
Privacy and security have become some of the biggest challenges in the big data
technology (Morabito 2014). Digital businesses dealing with big data applications
are required by the law to protect the privacy and the security of its customer’s data
as well as setting clear boundaries of some personal information. With the increased
in the data volume and the consequent risk of cyber attacks, it is becoming pro-
gressively difficult for digital businesses to protect data privacy (Katal and Wazid
2013). For example the United States (US) based digital company eBay, which has
128 million active customers had its customers database hacked between late
February and March 2014 (Kelion 2014). The database included users name, email
address, date of birth and phone number. This personal information can be used to
steal customer’s identities and even access other systems. Thus privacy and security
are major concerns when dealing with sensitive information.
6.4.3 Cost Structure
The cost structure aims to define to the costs of all the projects incurred by a digital
business, including the investment in customer value propositions, and other
business activities such as marketing, infrastructures and building partnerships with
others companies (Muhtaroglu et al. 2013). The biggest cost of the big data tech-
nology comes from the infrastructure and data management tools. Companies such
as Microsoft, Oracle, IBM have spent over $15 billion on software companies
specializing in data analytics and management (Cukier 2010). In 2010, big data
analytics and management firms were reported to be worth more than $100 billion
and this value is estimated to grow by 10 % every year due to the increased demand
of more efficient data management and analytics tools (Cukier 2010).
The availability of the cloud based infrastructures and application stores have
significantly decreased the financial resources requirement to enter the big data
technology (Muhtaroglu et al. 2013; Morabito 2014). Some businesses, like, e.g.,
eBay and Amazon, may develop in-house infrastructures and applications. These
applications need be developed with real time processing features. With these
applications and infrastructures, the businesses may also need to continually invest
and improve their big data services with innovations to create new solutions to
customers growing problems and maintain their place in the vast and competitive
Storage
space
Skilled
Privacy
People
Challenges for
Security Digital Data
Business complexity
Complex
Costs
Analytics
Complex IT
Infrastructures
Fig. 6.4 Big data challenges in digital business. Adapted from Muhtaroglu et al. (2013), Morabito
(2014)
market. Figure 6.4 summarizes some challenges of the big data application in
digital business explored in previous sections.
To illustrate how big data are being used in industry, two examples of digital
businesses using big data are discussed in the following sections. The examples aim
at highlighting the key challenges facing the organizations and to discuss the way
they are using big data solutions to overcome these challenges. The two organi-
zations considered within the case studies are Facebook and Netflix, both highly
successful digital businesses and both facing complex data challenges.
6.5 Cases Studies
This section will analyze two organizations that are pioneers in the digital business
domain: Facebook and Netflix. Particular attention will be paid to the way they use
big data technologies to evaluate business performance and adjust their strategies
accordingly.
Facebook is one the most relevant and diffused social media platform, which
allows users to connect to friends and family using a webpage called a profile. The
majority of the content on Facebook is user generated, from user comments, likes,
photos and videos. The profile acts as a medium from which to interact with other
members’ profiles. New content is generated and members further add to this new
content with comments likes and other interactions. For Facebook this in effect
leads to a virtuous circle of new content by users.
The platform is a social platform, where other social media sites offer specific
activities, i.e. Twitter for status updates or news, Instagram for photos and MySpace
for music. Facebook has content from all these media on the same site, which is
generated by both members and third party distributors such as music companies
and retailers. The website is the largest social media network on the planet primarily
because of this wide ranging appeal (note: anyone over the age of 13 can join
Facebook).
Looking a little deeper at the company, one can see it is also one of the largest
data hosts on the planet. From pictures to video, albums to chat messages the
website host users most treasured data. The website is a maze of data generated
daily by users; thus, these data are the companies greatest asset but also its greatest
challenge. Indeed, Facebook is dependent on the data generated by its users to
generate income; thus, Facebook is a company, which generates income through
selling advertising to its customers while they are using the site. The reason why
this model is different with regard to other digital businesses is because Facebook
can direct adverts, which are specific to the users preferences and demographic
information. This information is a key advantage that Facebook has over its
competitors, there is a value to the information it has on its users which competitors
do not have and is valued by advertisers. The challenge is analyzing this data using
big data methods to effectively target users and provide them with relevant
advertising.
6.5 Cases Studies 117
Facebook has 1.3 billion monthly users worldwide, some 830 million of them
using the site on a daily bases. Also, the users generate 4.5 billion activities
including updating photos, video and sharing comments and messages.
In recent estimates Facebook has been shown to produce over 4.5 billion pieces
of content and over 750 Terabytes each day. This large volume of data makes the
process of analysis very cumbersome. Further Facebook has indicated its users use
7 petabytes of photo storage on its website each month with data centers that have a
capacity of 100 petabytes. This excess capacity is a necessity as content on the site
is always increasing.
Point of Attention: The problem with having big data is that the vast
majority of data never gets used. A recent article surveying big data issues
noted that only about 0.5 % of all data is ever analyzed (Barrenchea et al.
2013). Thus, the vast majority or 99.5 % of data is not used as part of the
analytics processing. This means storage space is allocated for this data but
may not be required for viewing by the users or as part of customer analytics.
As there is a vast amount of information, which the company process each day,
Facebook has developed tools to manage how the data are processed and analyzed
to bring users more relevant information. These tools also provide users with more
relevant social media content, including articles and statuses; for example, the
algorithms implemented provide users with targeted advertising content.
Thus, Facebook uses big data analytics to direct advertising most relevant to its
user base. The advertising is based on geo locations as well as age, sex and other
demographic factors. This means that there is a unique set of adverts presented to
each user of the Facebook system. The advertising on Facebook helped the com-
pany bring in $7.8 billion in revenue during 2013 (Wikipedia 2014). Furthermore,
Facebook tweaks its algorithms on an on-going basis and one example is the tweak
of its advertising delivery formula. This has allowed it to move from generating
only a small proportion of its income from mobile advertising in 2012 to now
generating over 60 % of its revenue from mobile devices. As the site grows,
predictions say that more and more content will be generated from mobile devices.
Thus, mobile devices have helped increase revenue growth at Facebook but this
was only after changing the advertising algorithms and making content more
location specific, this in turn allowed Facebook to charge more for the advertising.
Considering now the infrastructure, Facebook opened a Prineville data center,
which is 62 k square feet in size. It houses over 500 cold storage racks, which can
hold 2 Terabyte of data. The standard data storage rack in the data center facility
uses about 8 kW of power. Facebook also has more than 100 petabytes data
capacity in this facility. However, Facebook has also admitted that it only uses
about 7 petabytes for storage photo in its facilities each month. This means there is
a lot of redundancy built for future growth. To empower this capacity, Facebook
switched on more secure offline storage facilities (cold storage) at its Prineville data
center, for storing pictures and other user media on hard disks. As well as using
existing data storage technologies Facebook has also worked with vendors to create
cheap storage vaults that keeps files on Blu-ray discs. Each vault contains about
10,000 discs with each containing around 100 Gigabyte storage space, a single one
of these cabinets has room for one petabyte, which is 1,000 Terabyte of data.
Though this technology is still in development it has the potential to change how
data is stored and further reduce storage costs.
Facebook has data in two forms. Flash data are stored on flash drives and are
quickly accessible (for example the login data for user or the most recently
uploaded articles). These data are easier to analyze because the upload times for
performing analytics is relatively quick. The second type of data is long standing
‘cold’ data which are rarely accessed (for example a photo album from a few year
earlier which has not been access since upload). The majority of data will fall into
the ‘cold’ form meaning that whilst they are valuable customer data, they may not
be accessed, but need to be retained in the likelihood the customer attempts to
access them in the future.
Finally, Facebook has a number of analytics tools to analyze and manage data.
Facebook manages its server data using the hadoop 2.0 platform and it has
developed a Hive tool which is a software fork sits on top of the Apach Hadoop
infrastructure to support the processing of data. Figure 6.5 shows a sample diagram
illustrating how the Facebook hive platform fits within the Hadoop software stack.
Hive provides an SQL-like interface to query data stored on Apache hadoop
platform. The Apache Hive is considered the standard for SQL queries over a
petabyte. It is a comprehensive and compliant engine that offers a powerful set of
tools for analysts and developers to access Hadoop data. Facebook initially
developed the Hive platform but other vendors now use it and have further
developed the tool. For example, Amazon web services maintains a software fork of
Hive that is included in Amazon MapReduce on Amazons Web Services, which in
turn are used by amazons web services and cloud client (including Netflix, whose
case is discussed in what follows).
HADOOP 1.0 HADOOP 2.0

Facebook
MapReduce Others
(cdata processing) (data processing) Hive
(data processing)
MapReduce YARN
(cluster resource management &
(cluster resource management)
data processing)
HDFS HDFS
(Redundant, reliable storage) (Redundant, reliable storage)
Fig. 6.5 Big data Hadoop structure. Adapted from Mayer (2013)
Point of Attention: The selection of hardware and software infrastructure to

support Big data analytics is a key factor that can help to evaluate the novelty
of a digital business.
The second case study of this Section considers Netflix one of the largest
streaming sites on the planet, working in 40 countries, and providing a streaming
service for movies, TV shows and factual content. The service offers access to
premium content based on a monthly subscription fee (Wikipedia 2014).
Netflix faces similar big data issues to the ones seen above for the case of
Facebook; however, the company has adopted slightly different approaches to how
it deals with these challenges. The volumes of data and complexity of analytics are
both equally challenging but the main difference is that Netflix data are not user
generated but rather videos licensed by the digital rights owners. Netflix then
provides a service in which allows users to stream the videos in exchange for a
monthly fee. The Netflix business model is based upon customers using the service
and paying the monthly subscription fee for a continued use of the service.
Point of Attention: Big data and analytics driven business models require a
careful attention to sourcing, computing, and data storage strategies for their
effective implementation and valuation in digital business competitive
environment.
The primary motivation for Netflix is to ensure customer have access to content
which would ensure they continue to be customers of Netflix. It has often been
compared to a pay per view TV station; however, unlike, television stations Netflix
customers are a part of a continuous loop of feedback with users using their
membership fees, actual viewing as well as content ratings to provide feedback to
Netflix. This instantaneous feedback allows the organization to always be in step
with its customer’s needs and demands. Big data analytics have helped in this
analysis process and supported the creation of new content such as the “House of
Cards” and “Orange is the new black” drama television series, content which is
only available on Netflix, which in turn has got more people using the service.
Furthermore, Netflix estimates that 75 % of viewers are from recommendations
from its customer recommendation engine (Vanderbilt 2013). The recommendation
engine is pivotal to how Netflix retains customers. As the business model is based
on customer continuing to use the service and continuing to pay the monthly
subscription fees. The recommendation engine has been created with using a
number of mathematical techniques to improve the performance of the recom-
mendation system. The goal is to primarily increase the likelihood someone
watches content provided by the recommendation engine.
Thus, the key data problem for Netflix is twofold: first, using the customers
viewing information and rating to provide users with appropriate recommendations
for viewing; then, ensuring the quality of the customer experience when the cus-
tomer is viewing content. In order to meet the requirements for streaming such a
large data capacity Netflix had two options to stream content to customers:
1. Build out the own data center to service clients streaming needs or
2. Use a cloud service from a third party to meet client requirements.
Netflix chose to meet client requirements using a cloud service provided by

Amazon web services. Using this service has allowed for Netflix to have on demand
access to hosting services from the providers. Reducing costs when demand is
lower but this cost would be increased costs when demand was higher as the extra
capacity is charged at a premium. Therefore, the hardware for Netflix data is cloud
data center run by Amazon’s web service. Storing data in a cloud environment can
be much cheaper than hosting it locally. While using the cloud solves the problem
of data cost and has the benefits of scalability. The problem with cloud computing is
the geographic proximity of data across potentially multiple location. Netflix has a
number of server centers it uses across the globe under the Amazon web service.
This allows for it to have geographical presence of its content in multiple regions
Box 1
• Data Size: There are 80 clusters with over 2,500 nodes (May 2014).
• Challenge: To maintain service and high quality of videos and
customer data, while also creating agility for development of new
tools for Netflix’s technology teams.
• Solution: An open source Apache data platform called Cassandra,
this allows Netflix to create and manage large data and minimize the
chance of failures and outages to customers.
without having employees in that region. An example is the Amazon web service
server Netflix had planned to use in Brazil for its South American and Central
American customers. Indeed, Netflix found there was no performance advantage to
using servers in Brazil. One advantage of the cloud service is that it could be
switched off relatively quickly and moved to a new location. And this is exactly
what happened in the case of Brazil. Because there was not advantage in using the
server in Brazil to serve those customers, Netflix reverted to serving them from the
United States (US) East Coast as the traffic would have been via Miami in any case.
For its European customer Netflix began serving them through Ireland, in the
Amazon web service server centers in Dublin. There were slightly higher costs to
perform this operation in Dublin as compared to the east coast US but the company
chose this approach as it reduced latency for streaming because of geographic
proximity. Another key component in cloud computing is disaster recovery in the
event a server is knocked out but act of nature. Having multiple data centers aids to
have contingency plans in the event a single data center is knocked out.
Furthermore, as the cloud brings a number of challenges to managing data in a
geographically distributed environment Netflix has developed new tools to manage
this highly complex environment. In particular, Netflix uses a tool called Cassandra,
which is an open source SQL database; this tool was initially created by Facebook
and now is under the open source Apache license. It scales across multiple data
centers all over the world. This tool is optimized for random write/reads, which
makes real-time queries efficient. The example in the box 1 shows how Netflix
manages the large data and it holds how the analytic tools help they have developed
help to manage complex challenges inherent in big data.
Big data and big data analytics require not a single methodology or approach for
their use in digital business evaluation. What seen in both the above case studies is
the way in which the approaches change to meet the needs of each company. In
particular, there is no perfect fit in terms of the hardware or software. The challenge
is identifying the most appropriate technique for the organization and this would
vary from one organization to another. In what follows recommendations are
provided for these issues.
6.6.1 Hardware
When we speak about big data, the first suggestion for holding this vast influx of
data is cloud computing. Though in the case study this is not necessarily the right
approach for companies such as Facebook whose intrinsic value comes from having
direct access to user data. The two options open to digital businesses is to decide on
whether to use the traditional data center model in which the organization owns and
runs data centers or to use cloud based services. The choice for organizations is
usually determined by the organizational strategy and cost.
Another challenge when dealing with big data is finding the right balance
between processing power and data storage. Getting the right balance is important
as it ensures the right mix of hardware components can be identified and acquired.
Key to this process is benchmarking the activities that would be commonly
expected from the data center.
6.6.2 Software
The main framework used for data management is currently Hadoop. This is an
Apache software infrastructure used by Facebook, Amazon, Netflix and many other
digital businesses. The key advantage of this tool is that many of its components are
open source. Key Hadoop users such as Facebook (Hive) and Netflix (Cassandra)
have developed add-ons to Hadoop framework, which adds further value to Hadoop
ecosystem.
Key to the evaluation of how digital businesses should use big data is under-
standing how its customers utilize the data and then creating solutions, which fit
these specific requirements. In the case studies we see how hardware and software
vary though the data are seemingly similar. This is because key to the big data
agenda is flexibility in how digital businesses organize and utilize their data for
maximum business gain. This gain is only possible when companies understand
their data and implement strategies to effectively monetize it without alienating
customers. In conclusion data can be used to create information and information to
create knowledge, and knowledge for a digital business is money.
6.7 Summary
This chapter discussed how big data and analytics can be used to evaluate business
performance. It described six steps that summarize this process: Goals, Selection of
Data, Processing Data, Data Mining, Evaluation and Visualization & Feedback. It
then presented an analysis of the advantages and opportunities of using big data and
analytics, identifying customer value proposition, customer segmentation, channel
diversity, and better customer relationship as the most important ones. On the other
hand, it also analyzed the challenges that organizations are facing when they want
to adopt these technologies and create organizational advantage and highlights the
importance of having skilled people in this area that is relatively new and thus the
talent supply is scarce. Privacy and Security and the relatively high costs were also
identified as challenges. Finally, this chapter closes with the analysis of two case
studies: Facebook and Netflix, to demonstrate how these organizations have used
big data and analytics to evaluate and shape their business models.
References
Apple Store. http://store.apple.com. Accessed 16 Nov 2014
Barrenchea, M.: Big data: big hype? http://www.forbes.com/sites/ciocentral/2013/02/04/big-data-
big-hype/. Accessed 16 Nov 2014
Cukier, K.: A special report on managing information: data, data everywhere. http://www.
economist.com/node/15557443 (2010). Accessed 16 Nov 2014
Davenport, T.H., Patil, D.J.: Data scientist: the sexiest job of the 21st century. Harvard Bus. Rev.
90, 70–76 (2012)
References 123
Experian Hitwise. Getting to grips with social media—An Experian Insight Report, Experian
Limited (2010)
Google Play. http://play.google.com/store. Accessed 16 Nov 2014
Gopalkrishnan, V., Steier, D.: Big data, big business: bridging the gap. In: BigMine ’12:
Proceedings of the 1st International Workshop on Big Data, Streams and heterogeneous Source
mining, pp. 7–11 (2012)
Katal, A., Wazid, M., Goudar, R.H.: Big data: issues, challenges, tools and good practices. In:
2013 6th International Conference on Contemporary Computing (IC3 2013), pp. 404–409
(2013)
Kelion, L.: eBay makes users change their passwords after hack. http://www.bbc.co.uk/news/
technology-27503290. Accessed 16 Nov 2014
Manyika, J., Chui, M., Brown, B., Bughin, J., Dobbs, R., Roxburgh, C., Byers, A.H.: Big data: the
next frontier for innovation, competition, and productivity. McKinsey Global Institute, San
Francisco (2011)
Mayer, C.: Hortonworks show YARN future Hadoop with HDP 2.0 preview. http://jaxenter.com/
hortonworks-show-yarn-future-hadoop-with-hdp-2-0-preview-106254.html (2013). Accessed
16 Nov 2014
Mithas, S., Lucas, H.C.: What is your digital business strategy? IEEE IT Prof. 12, 4–6 (2010)
Morabito, V.: Trends and challenges in digital business innovation. Springer International
Muhtaroglu, F.C.P., Demir, S., Obali, M., Girgin, C.: Business model canvas perspective on big
data applications. In: Proceedings—2013 IEEE International Conference on Big Data, Big
Data 2013, pp. 32–37 (2013)
Osterwalder, A., Pigneur, Y., Smith, A., Movement, T.: Business model generation: a handbook
for visionaries, game changers, and challengers. Wiley, New York (2010)
Rajpurohit, A.: Big data for business managers—bridging the gap between potential and value. In:
Proceedings—2013 IEEE International Conference on Big Data, Big Data 2013, pp. 29–31
(2013)
Rogers, D.: The network is your customer: 5 strategies do thrive in a digital age. Yale University
Press, UK (2011)
Sagiroglu, S., Sinanc, D.: Big data: a review. In: International Conference on Collaboration
Technologies and Systems (CTS), pp. 42–47. IEEE, San Diego, CA (2013)
Vanderbilt, T.: The science behind the Netflix algorithms that decide what you’ll watch next.
http://www.wired.com/2013/08/qq_netflix-algorithm/ (2013). Accessed 16 Nov 2014
Wikipedia: Facebook. http://en.wikipedia.org/wiki/Facebook?uselang=ja?iframe=true&width=
90%&height=90%. Accessed 16 Nov 2014
Wikipedia: Netflix. http://en.wikipedia.org/wiki/Netflix. Accessed 16 Nov 2014
Wikipedia: Twitter. http://en.wikipedia.org/wiki/Twitter. Accessed 16 Nov 2014
Zhang, D.: Inconsistencies in big data. In: Proceedings of the 12th IEEE International Conference
on Cognitive Informatics and Cognitive Computing (ICCI*CC), pp. 61–67 (2013)
Zheng, Z., Zhu, J., Lyu, M.R.: Service-generated big data and big data-as-a-service: an overview.
In: IEEE (ed.) IEEE International Congress on Big Data, pp. 403–410. IEEE, Santa Clara, CA
(2013)
Managing Change for Big Data Driven
Innovation 7
Abstract
Big data is now becoming a key organizational asset, which represents a strategic
basis for business competition. This development is making organizations to
consider new innovative techniques on maximizing the potentials of big data as
well as the challenges it creates. Yet, the success of many organizations demands
new skills as well as new perspectives on how the epoch of big data could
advance the speed of business processes. With the growth of big data, new
analytics tools have evolved together with new progressive business models. In
this chapter, we explore the innovative capabilities of the growing big data
phenomenon by discussing issues concerning its methodologies, its impact on
organizations business models, novel tools for analytics including challenges
encountered by many business organizations. Our findings are substantiated by
describing the real-life cases of Adobe and Hewlett Packard organizations.
7.1 Introduction: Big Data—The Innovation Driver
Numerous organizations are hunting down better approaches to always innovate and
optimize their practices, products, and services in today’s competitive business
environment (Piatetsky-Shapiro 2013). Traditionally, innovation is seen as a factor
of consistent research and improvement processes using manual conventional pro-
cedures (Gobble 2013). Due to the rise of big data, organizations are gradually
depending on accumulated computerized data acquired from different sources such
as their suppliers, customers and shareholders for recognizing innovative products
and service systems (Sagiroglu and Sinanc 2013). Although access to information for
many business organizations seems like a ‘walk in the park’, there exist enormous
challenges of identifying valuable information, analyzing, adapting, or applying it
for the development of productive business performance (Yan 2013).

DOI 10.1007/978-3-319-10665-6_7
126 7 Managing Change for Big Data Driven Innovation
Today, numerous organizations are progressively investing their resources on big

data as a business asset including mechanisms for its analytics as one of their core
profitable capabilities (Louridas and Ebert 2013). However, big data also encompass
organized in-house commercial data produced from daily trade transactions, such as
sales, suppliers, customers, and inventories. Moreover, unstructured peripheral
and non-transactional data are also in the fore, which includes online behaviors of
customers’, interactions on social media and ecommerce dealings indexed by big data
giants e.g. Google and Facebook. Nevertheless, many experts share similar opinion
about the innovative strategies of big data, although, a few admit that the innovative
capabilities of big data are still relatively unclear and unexploited. Some have offered
useful suggestions and guidelines for business organizations to generate innovative
services and products from big data applicable to various fields (Groves et al. 2013;
Wielki 2013). In the next section, we present some well-researched guidelines for
deriving innovation from this huge asset called ‘big data’. Remarkably, some tech-
nology business giants are already implementing some of these innovative guidelines
for improving their products and services (Davenport and Dyché 2013).
7.2 Big Data—The Key Innovative Techniques
The big data age has produced innovative principles for business management. In
previous eras of technological innovation, business pioneers learnt that the smallest
well-organized scale was a crucial factor of competitive success (Cui et al. 2014).
Similarly, competitive profits are possible to be accumulated by organization’s
collecting more and enhanced data but also utilize such data at an efficient scale
(Bjelland and Wood 2008).
Be it as it may, in what follows we critically reflect on the five techniques shown
in Fig. 7.1 (i.e., integration of data platforms, testing through experimentation,
real-time customization, generating data-driven models, and algorithmic and
Integration of Data Testing through

Platforms Experimentation
Big Data
Innovation
Real-time Generating Algorithmic and

Customization Data-Driven Automated-
Models Controlled Analysis
Fig. 7.1 The big data innovation techniques. Adapted from Davenport and Dyché (2013)
7.2 Big Data—The Key Innovative Techniques 127
automated-controlled analysis), in order to identify how big data may further assist
business organizations in developing innovative strategies, which are adaptable to
the rapidly changing models occurring in today’s highly extrinsic technological era.
7.2.1 Integration of Data Platforms
One of the biggest challenges encountered by many business organizations is

accrual of datasets lurked departmentally, e.g. sales, marketing, service operations
etc. (Dandawate 2013). Unfortunately, this business model impedes the timely
utilization of the datasets for complex synthesized analytics (Dean and Ghemawat
2008). In addition, the amassing of information within business units brings a huge
challenge for innovation. For example, numerous financial organizations stay
limited from the failure to share their huge data between assorted business areas,
such as money management, lending and financial markets (Economics 2013).
Consequently, business organizations are being prevented from developing a
coherent outlook of understanding inter-relationships amongst financial markets as
well as individual customers (Davenport and Dyché 2013).
Due to the growth of technological devices, information is now readily available
to almost anyone (Mithas et al. 2013; Russom 2013; EY 2014). This development
poses a threat to the innovative potentials of organizations solely depending on
exclusiveness of their data as the core resource of their competitive advantage
(Davenport and Dyché 2013). Considering the real-estate sector, which relies on
trades of asymmetric information such as, e.g., access to transaction data and
awareness of the ‘bid and ask’ behavioral pattern of purchasers. Such information
entails a momentous financial cost as well as energy. Conversely, real-estate online
experts in big data and analytics are now bypassing agents; thus creating online
forums to allow purchasers and sellers to share their views on the value of properties
as well as generating parallel bases for real-estate data (Dandawate 2013).
A further put-down at exclusive or proprietary data is the assemblage of freely
available satellite imagery that, when processed and analyzed, comprises indica-
tions about physical amenities of an organization’s competitors (Stonebraker et al.
2013). These satellite detectives collect insights into business checks as exposed by
the capacity of the physical amenities e.g. goods movements etc.
In order to develop innovative products from big data, organizations are now
attempting to uncover their departmental reserves of exclusive data: the integration
of datasets from multiple systems and the search of external suppliers’ and
customers information to create new products is now a crucial determinant of their
manufacturing costs. Extra integrated data platforms now permit organizations and
their suppliers to pool insightful resources during the design phase (Cui et al. 2014).
7.2.2 Testing Through Experimentation
With the advent of big data, there arise a novel potential for a different sort of
decision making using controlled experiments. Currently, organizations are testing
hypotheses and analyzing results to inform their business operational processes.
Testing through experimentation are helping administrators differentiate causality
from correlation, therefore decreasing the inconsistency of results and improving
product performance (Mithas et al. 2013).
For example, many online business organizations are continuous testers; by
apportioning a quota of their web page views to undertake experiments that iden-
tifies the elements that motivates the active engagement of end-users (Cui et al.
2014; Davenport and Dyché 2013; Parmar et al. 2014). In addition, businesses
marketing their products physically also make use of experiments in order to
influence their marketing strategies. For example, fast food merchants McDonalds,
has equipped some of their outlets with devices that collect data about customer
interactions, ordering patterns as well as in-store consumer traffic. The importance
of these huge datasets is that experts are able to model the effect of variations on
restaurant designs, menus sales and productivity (Kelly et al. 2014; Qubole 2014).
7.2.3 Real-Time Customization
In this section, we describe how organizations are innovatively using big data by
making real-time customization conceivable. Organizations are now able to monitor
the behavior of customers from click streams on the Internet, modify their inclina-
tions, and most importantly model their possible actions in real time. Furthermore,
they are learning to ascertain when customers are close to making a buying decision
and then pre-empt the transaction process to a finishing point by bundling desired
products combined with a reward offer. Real-time custom analysis can also increase
procurements of products by the most valuable customers of a business organization
(Drake et al. 2014).
As an innovative driver, big data real-time customization is relatively obvious in
retail sectors due to the size and value of data available from online procurements,
interactions on online social networks (OSNs) and more lately, location-specific
smart phone communications. Nevertheless, other sectors can also gain from
this novel system of Big-Data application, owing to the growth of sophisticated
analytical tools for distributing customers into more insightful micro-segments
(De Fortuny et al. 2013).
7.2.4 Generating Data-Driven Models
The growth of big data has enhanced the creation of innovative groups of businesses
that have accepted data-driven business models. Several of these businesses perform
inter-dependent roles in value chains generating valuable consumable data formed
7.2 Big Data—The Key Innovative Techniques 129
by business transactions (Economics 2013; Eisenhardt and Martin 2000; Fan and
Bifet 2012). For example, a transport business organization recognized that during
their daily business operations, enormous volumes of information on product
deliveries are being collected. In order to maximize this opportunity, the organiza-
tion created a division that trades the data to complement business and economic
predictions (Kelly et al. 2014).
Another good example of the development of new models is an organization that got
so many insights from analyzing data as part of a manufacturing improvement, which
prompted a decision to build a business to do similar work for other organizations
(Hartmann et al. 2014). Currently the business combines supply-chain data for many
manufacturing customers and sells software tools to improve their performance. Out-
standingly, the newly developed service business now overtakes the income generated
by the organizations parent manufacturing business (Provost and Fawcett 2013).
7.2.5 Algorithmic and Automated-Controlled Analysis
As also discussed in previous chapters (see, e.g., Chap. 6), big data increase the
development and implementation for algorithmic and automated-controlled analysis.
In some manufacturing organizations, algorithms are used to analyze sensor data from
lines of production, in this manner, the wastage of raw materials are limited, production
outcomes are increased and then costly human involvements are avoided (Mithas et al.
2013). This type of data is usually analyzed by groups of computers to enhance
production and diminish downtimes through the process of imputing their results to
real-time tasks. Products such as photocopying machines used to aircrafts can now
create streams of data capable of tracking their usage. In some situations, manufac-
turers are able to analyze the inward data and then automatically repair software
malfunctions. Besides, computer hardware merchants are collecting and analyzing
such data to plan pre-emptive maintenance before customers’ operations are inter-
rupted by system failures. As a means of innovation, such automatic-controlled data is
also utilized to make necessary changes in product design (Elmqvist and Irani 2013).
The key point is value-added performance, improved risk assessment, as well as
the capability to uncover useful insights that would otherwise remain unknown.
Recently, more business organizations are participating in this decision-making
industrial revolution due to the massive reduction in price of sensors, communi-
cations devices and analytic software (Özcan et al. 2014).
7.3 Big Data: Influence on C-Level Innovative Decision

Process
A lot has been stated about the emergent significance of big data as a tool for
business innovation. But how are these types of data utilized to drive innovation?
How does it affect the in-house tactical initiatives of business managers and
executives? Is big data beginning to influence C-Level decisions?
As earlier mentioned, big data offer enormous benefits to senior executives of

a business organization. Big data have the capacity to improve insights into
customers buying patterns, forecasts as well as increasing productivity and efficiency
for a sales executive (Taylor and Labarre 2006).
Furthermore, big data through analytics can assist a Chief Marketing Officer
(CMO) in developing a vibrant marketing strategy, social media analysis and
customer subdivision. They can also be utilized to understand impending design
needs to make logical business decisions. For Chief Information Officers (CIOs),
big data can establish the link between IT and business goals during board meetings
with the Chief Executive Officer (CEO) (Lohr 2012; Piatetsky-Shapiro 2013;
Taylor and Labarre 2006).
7.3.1 Stimulating Competitive Edge
In order to be a source of business advantage, big data systems need to build new
types of processing to data-sets targeted at stirring innovative impact through the
formation of ground-breaking business models, goods or services (see, e.g., the
discussion in Chap. 4).
As of the time of putting this book together, big data driving income generation,
optimization and business benefit which is getting the attention of C-Level executives.
It covers many areas from direct advertising, to scam detection as well as transaction
optimization. For example, in the financial sector, banks emphasize on performing
analyses on in-house and customer transactions aimed at detecting scams; likewise
retail outlets are modifying their inventories based on previously aggregated insights
made possible by big data. A key strategy in getting measurable value from big data is
actually analytics; this usually attracts the curiosity of C-Level executives who seek to
drive a competitive edge in the global economy (Economics 2013).
In a survey lead by a technology research organization, Economist Intelligence
Unit discovered that senior executives require big data for their workforces due to
the perceived belief that the use of analytics tools will enhancement productivity
within the organization (Manyika et al. 2013). Thus, business organization CIOs are
now playing the role of software evangelists for big data analytics solutions. They
are also making sure that the CEOs and other C-Level executives realize that
insights derived from big data can be the key edge to drive a competitive advantage
which results to the targeting of innovative markets, fraud reduction, better-
informed business decisions etc. (Lohr 2012; Taylor and Labarre 2006).
7.3.2 Predictive Analytics: Data Used to Drive Innovation
The explosion of big data implies that predictive analytics is gaining wider
acceptance for operational use. For example, SAP’s vice-president of product
7.3 Big Data: Influence on C-Level Innovative Decision Process 131
marketing analytics, James Fisher, predicts that advanced analytics market will be
worth about $3bn by the year 2016. He highlighted that many SAP customers make
use of the organization’s predictive analytics platform—Hana, to improve profit and
reduce expenses (Bell 2013; Fan and Bifet 2012). Although in the environment of
customer relationship management (CRM), predictive analytics is at an experi-
mental phase which explains the reason organizations whose business model are
technologically driven, are in the forefront with applying predictive analytics to
CRM (Huo et al. 2014). In what follows examples of businesses using big data
analytics are discussed.
Bigpoint, a games development organization, are utilizing predictive analytics to
monetize players and upsurge their revenue by an estimated 10 and 30 % a year
(Bell 2013). The predictive model of the organization allows them to make clever
real-time decisions about a player’s actions. For instance, it acts as trigger for the
predictive system to analyze prior gaming behavior when a player’s ‘ship’ is
damaged. If applicable, a custom-made context-related note offers the player a new
‘ship’ for a minimal fee (Yan 2013).
Her Majesty’s Revenue and Customs (HMRC) has framed a predictive analytics
platform called ‘Adept’ keeping in mind the end goal to build debt collection and
risk estimation. Adept incorporates analytics into the debt administration process
with a specific end goal to development debt gathering interruptions for deferred tax
payments on a yearly premise. HMRC’s systems are integrated with ‘Adept’
utilizing predictive modelling techniques to inform a comprehensive risk and
behavior-driven collection tactics. Remarkably the technology also recognizes
diverse types of debtor categories and then targets its communications to these
categories through particular attributes. By March 2015, it is estimated that HMRC
will have the capacity to collect an extra £3bn of debt (HM Revenue and Customs
2013).
In the financial services sector, predictive analytics is used to improve the
methods of debt collection. Various organizations are attempting to identify that a
certain type of customers responds undesirably to payment prompts, while some
will make timely payments without reminders; others tend delay payment when
prompted. By means of recognizing these behavioral tendencies, business organi-
zations are able to enhance their processes as well as effective customer relations.
In the retail segment, Tesco is building on its reputation for new technologies to
enhance CRM. Tesco has started a scheme, which offers suggestions to customers
about products related to their purchases in real time. A ‘loyalty card’ is injected
into a device attached to a shopping trolley, through which personalized sugges-
tions are presented to the customer during the final buying process. For instance, a
customer may place an electronic shaving clipper in the shopping trolley: as a result
he receives a 2-for-1 offer related to the item, such as batteries for clippers. With the
availability of data related to the customer’s identity as well as their purchases,
Tesco is making an effort to forecast their next actions and then provide inter-
vention mechanisms showing related offers.
7.4 The Impact of Big Data on Organizational Change
In order to fully explore the existent and prospective capabilities of the big data
boom, business organizations need to enforce certain modifications to their in-house
operational procedures by considering the three kinds of stakes described in what
follows (see also what discussed in Chap. 6).
• Data: Business organizations need systems to collect large magnitudes of

information in specific industrial formats allowing for stress-free accessibility
and analysis. However, due to the enormous proliferation of online social
networks and allied technologies, most large organizations now have this—in
fact; they normally have more than they are able to utilize (Bughin et al. 2010;
Groves et al. 2013).
• Analytical tools: The second stake required is cutting-edge analytical tools, such
as, e.g., NoSQL and Hadoop (see also the Preface and Chap. 6 in this book).
NoSQL and Hadoop are proprietary open-source tools and platforms which are
universally accessible (Morabito 2014); but the biggest challenge is getting the
right people proficient enough to make good use of them (Kelly et al. 2014;
Olson 2010).
• Expertise: This is usually the most challenging category of the stakes: advanced
analytics necessitates a workforce with high-tech skills in data science to global
privacy laws, as well as a proficient knowledge of the business including relevant
sources of value (Lohr 2012).
Nevertheless, the mere acquisition of the aforementioned stakes is insufficient in

helping organizations because big data is not a standalone additional technology
initiative. Frankly, big data is not a technology initiative at all; it is actually a
business system that needs technical savvy. Therefore, the sheer addition of
capability and expertise alone cannot enable the information technology functions
to start engendering database insights. The big leaders of analytics have discovered
that to succeed with big data, organizations need to practice the consistent process
of deeply embedding big data into their business models (Manyika et al. 2013).
This is the most efficient technique to guarantee that derived insights are pooled and
shared across all units within the organization. It also guarantees that the entire
organization identifies the combined effects and scale benefits that well-conceived
analytics proficiency can deliver (Ji et al. 2012). Taking the above issues into
account, big data analytics can be relevant to organizations in four areas listed
below:
• Refining current products and services

• Refining in-house processes
• Developing innovative product offerings
• Changing business models.
7.4 The Impact of Big Data on Organizational Change 133
To get a clearer meaning of the concepts mentioned above, the following sec-
tions presents a guideline for business organizations to follow if they are keen on
modifying their operational procedures in order to develop innovative products
from the insights derived from big data’s analytics.
Firstly, however, let us consider a basic aspect of starting any business
organization: goals. This is the best place for organizations to start the big data
embedding process; clarifying out their goals. Examples of some useful goal-oriented
statements are; “we will integrate cutting-edge analytics and insights as the critical
component of all major decisions” or “we will adopt big data systems as a
new-fangled way of undertaking our marketing strategies”. A well-defined goal by
C-Level executives is a vital prerequisite for the kind of organizational change big
data offers. Additionally, C-Level executives also need to answer these sorts of
questions e.g., “How far are we ready to go”? “How will our business performance
improve through big data”? “What do we need to focus on”?
7.4.1 An Incentivized Approach
Many organizations are possess opportunities for complex analytics, still, very little
of them have the capabilities to further improve their trajectory by accurately
defining priorities and selecting the right angle of entry. With the organization’s
goals well defined, executives can work on developing a horizontal analytics
capability. They learn how to overcome internal struggle and create the skill to
utilize big data throughout the organization (Yan 2013).
Organizations do not easily change and the value of analytics may not be
obvious to everybody. Therefore, a continuous work in helping employees and
customers change their daily behaviors as well as continuing in the new innovative
path without lapsing is necessary. Business organizations need to clearly define the
owners and sponsors for analytics initiatives. By providing adequate inducements
for behaviors that are analytics-driven, they will ensure that big data is integrated
into processes for making key decisions (Russom 2013).
A good example of an incentivized approach for adopting analytics is Nordstrom;
Nordstrom, a leader in the retail space when it comes to enhancing the online
shopping experience raised the obligation for analytics to an upper managerial level
in its organization in order to make analytical tools and insights extensively
accessible including integrated analytics-driven goals into its vital tactical initiatives.
7.4.2 Creating a Centralized Organizational ‘Home’
Leaders of big data need to create an organizational ‘home’ for their progressive
analytics proficiency, which can be supervised by a principal analytics officer. The
process of creating an organizational ‘home’ embroils some important design
choices (Demchenko 2013). An organization has to define its approach for the
deployment of big data. Furthermore, the collection and ownership of data across
corporate functions has to be assigned and a well-structured plan to generate
insights must be implemented (Yan 2013). The technological infrastructure, privacy
policy and access rights must be hosted and maintained by the organization.
However these tasks seem quite burdensome to achieve, hence we present four
models organizations need to adopt in order to comply with the needed changes for
big data optimization.
• Organizational division: If each business organization’s division have their

distinct data sets they become capable of making their specific big data decisions
with reduced supervision or monitoring. An example of organizations that uses
this business model are AT&T and Zynga (Pearson and Wegener 2013).
• Organizational division with central support: This is when organizational
divisions decide their own choices then co-operate to work on designated
initiatives. An example of organizations that uses this business model are Google
and Progressive Insurance. Progressive, Insurance depends on it to capture
driving behaviors and define customer risk profiles (Bain Insights 2013; Pearson
and Wegener 2013).
• Center of Excellence (CoE): In this model, a self-governing center supervises the
organization’s big data. Various organizational divisions follow initiatives
guided and coordinated by CoE. Also, Amazon and LinkedIn are examples of
organization’s that depend on this model (Demchenko 2013).
• Fully centralized: In this model, the corporate center is responsible for priori-
tizing business creativities. Netflix is an example of a business organization that
adopts this approach (see also Chap. 6). The business goals and operating model
of an organization greatly affects the type of model they adopt for big data
change (EMC 2013). An organization with in-depth analytics capabilities as well
as a focus on experimentation, testing and innovation, such as, e.g., Google,
normally depends on a decentralized method (Couchbase 2010). However, CoE
is found to be the mostly adopted model due to its enormous advantages and the
fewer restrictions. An operational CoE allows cross-business-division access and
data sharing capabilities. In addition, CoE assists in providing organizations
adequate analytics strategies thereby setting the roadmap for the maintenance
of privacy policies (Pearson and Wegener 2013). For example, a foremost
European telecommunications organization, is attempting to change its internal
procedures by deploying big data to analyze customer data in order deliver
improved services and offers as well as utilizing network traffic data to enhance
network supervision and investments. These capabilities will all be coordinated
by a CoE.
7.4 The Impact of Big Data on Organizational Change 135
7.4.3 Implementing the Changes—First Steps
As business organizations are already delving into the world of big data, the com-
plexities we have discussed necessitate the need for analytics capabilities to be
anchored in the organization in order to create substantial insights (Meeker and Hong
2014). Currently, many organizations executives are carting beyond competitors,
while others are making intense efforts to keep up with pace. Nevertheless, the first
step needed is the benchmarking of the organization and determination of the
organization’s present situation in big data analytics, compared with that of rivals.
This process will further assist the organization in deciding the essential investment
needed to advance their relative position. In circumstances were an organization is
considerably behind the rivalry, a rapid innovative platform is often compulsory to
create and withstand change. This begins by testing hypotheses in order to study
where and how cutting-edge analytics is most expected to benefit the organization
(Computer Research Association 2011).
7.5 Methodologies for Big Data Innovation
7.5.1 Extending Products to Generate Data
Data generated by products can stimulate the development of better business

models, product innovations and business work effectiveness (Bughin et al. 2013).
Extending products to generate data is highly influential in conditions where data
can stimulate enhancement in the characteristics of products. Moreover, another
benefit of this methodology is supporting product manufacturers to maintain
interaction with the products regardless of their physical locations. Connectivity can
offer the platform for developing existing connections or forming completely novel
types of consumer connections (McAfee and Brynjolfsson 2012). For example, the
Engine Health Management (EHM) proficiency of Rolls-Royce uses embedded
sensors to observe and store parameters of aircraft engines such as temperature,
pressure, altitude, and vibration.
7.5.2 Digitizing Assets
The ability to produce a physical illustration of a digital design using a 3 Dimen-

sional (3D) printer external to old-fashioned supply chains has the prospective to be
totally disruptive. Recently, business organizations are considering the vast busi-
ness opportunities of these types of technologies. This methodology will also
facilitate the expansion of measurement capabilities and additional backup capacity
for other systems of value enhancement. Peking University People’s Hospital are
digitizing healthcare records as well as integrating telemedicine that leverages real-
time alerts and mobile communications. The goal however, is to observe the vital
signs of patients in real time for the sustenance of individualization and respon-
siveness (Groves et al. 2013). In Korea, HomePlus, a retail chain business, has
made computer-generated stores in physical surroundings such as subway stations.
Today, consumers can scan items on supermarket shelves with their smartphones
and then get the products brought to their homes within a short period of time
(Lavalle et al. 2011).
7.5.3 Trading Data
Data as a business asset has the capacity of motivating new business models
through commerce, distributing acquired knowledge and engendering creativeness
(Wielki 2013). Through this means, the process of trading data can encourage
bigger union between businesses. The Marine Institute of Ireland is an example of
an organization making accessible data it collects on ecological situations, con-
tamination levels and marine life open to external bodies. Their collected data is
utilized to advance the safety of the public and increase economic activity by
refining the accuracy of flood forecast as well as allowing well-organized seafood
and shipping procedures (Bollier and Firestone 2010; Chen et al. 2012). Likewise, a
huge community-based traffic and navigation application called Waze facilitates
traffic and information about maps to be swapped among linked users to develop
their daily commute in real-time (Obeng-Odoom 2009). The volume of data rises
proportionately to the participation of its users, hence, refining the quality of data as
users outwit traffic and boost routes.
7.5.4 Forming a Distinctive Service Capability
The internal data within business organizations can be converted into self-standing
businesses themselves (Computer Research Association 2011; Ji et al. 2012; Meeker
and Hong 2014). Transforming such capabilities into distinct novel businesses can
be fruitful when the capabilities of the particular data can be explored across various
market situations, especially where new developing economic environments are
evolving to exploit new prospects. For example, IBM computerized its in-house
transportation booking system and expenditure reporting methods, which reduced
their managerial budgets up to 75 %. With this success, they further advanced the
system into a distinct internal business known as the Global Expense Reporting
Solutions for IBM customers (Parmar et al. 2014). Citigroup did a comparable
demonstration as they created models to reveal market inadequacies that stalled their
customers’ ability to make the best utilization of payment systems. It refined and
extended its models into a set of particular customer benefits under the standard of
Citidirect BeMobile (Parmar et al. 2014).
7.6 New Big Data Tools to Drive Innovation 137
7.6 New Big Data Tools to Drive Innovation
As earlier discussed in previous chapters, the growth of technological platforms

such as online social media (OSNs) has resulted in the massive growth of big data.
More recently, every time users attempt to monitor the distance covered in the
course of driving a car, surfing the Web and interacting with their smart phone
applications to find the closest café, data are being created at an alarming volume,
velocity and variety. However, retrieving the right information in time to answer
quick inquiries remains a huge test for people, as well as for organizations.
Having the right devices or tools to execute a relatively accurate analytics is
highly significant and can lead to the most creative business systems. In the era of big
data, tools such as Hadoop, 1010data, Actian etc. are all extremely useful in
analyzing and gaining insights from data architectures by applying hybrid data
modeling concepts (Jinbao and Wang 2012). In this section, we extend the
discussion in Chap. 6, briefly describing the Hadoop big data robust platform as well
as four top tools, which can be integrated with it in helping business organizations
maximize the opportunity provided by big data for their productivity.
7.6.1 The Hadoop Platform
Hadoop is an open source software platform that facilitates the distributed

processing of huge datasets through collections of product servers. The platform is
developed with scalable capabilities from a sole server to thousands of machines.
Hadoop platform has an extraordinary level of fault tolerance and its resiliency
stems from the software’s ability to detect and deal with failures at the application
layer rather than depending on high-end hardware systems (Yan 2013).
7.6.2 1010DATA Cloud Analytics
1010data’s database provides support for parallel processing with its own query
language that supports a subsection of SQL functions comprising graph and time-
series analyses. 1010data also possesses the capabilities to deal with semi-structured
data and machine data. Additionally, the platform owners offers a whole stack
which comprises of data-visualization tools, data integration, monitoring and
reporting and many other innovative analytic functions e.g. machine learning and
statistical analysis (Saleem 2014).
The private-cloud method of 1010data helps their customers in managing and
scaling infrastructure easily. Other features include the capability for a central data
management and application programming interfaces (APIs) which are integrated
with back-end systems access controls. As a custom services provider 1010data
creates private-cloud applications aimed at meeting customer needs, which is highly
in contrast to cloud provider—Amazon, which conveys low-cost services to tens of
thousands of customers (Batty 2013; Bughin et al. 2010).
7.6.3 Actian Analytics
Actian analytics is an open-source transactional database with a high-speed

analytical database management system (DBMS). In April 2013, Actian analytics
acquired another big data analytics company, ParAccel, which introduced a better
move into big data analytics with an enormously parallel processing database
system currently referred to as Actian Matrix (Bughin and Chui 2010; Groves et al.
2013; Russom 2011).
Actian analytics depends on the synthesis of cloud services, data-integration and
analytical DBMS options in which Hadoop is a fixture of the architecture of data
management. SQL ETL (ETL which means Extract, Transform and Load, is a
technique used to collect data from various sources, transform the data depending
on business needs and load the data into a destination database) is included in
Actian dataflow and data cleansing on Hadoop alternatives that work with distri-
butions from Cloudera, Apache and others.
7.6.4 Cloudera
The big data tool—Cloudera, provides support for Hadoop in order to ensure
enterprise-grade performance, data-access control, security and reliability. In
addition, Cloudera offers exclusive software components such as Cloudera Navi-
gator, Cloudera Manager as well as mechanisms for backup and recovery (Batty
2013; Groves et al. 2013; Russom 2011). This tool is massively rich with open-
source components such as, Cloudera Search, best controlled by Cloudera Manager
to deliver and monitor workloads. Another software component, Cloudera Navi-
gator enables the provision of access control and auditing. Cloudera big data
software is gradually growing to become a “heavy weight” for data analytics.
(Wanderman-milne and Li 2014). Furthermore, it provides reference hardware
configurations for leading Hadoop software distributors. It is also a hugely parallel
processing columnar analytical database management system aimed for swift
analysis of large structured datasets. HP Vertica Analytics competes with the
similar tools such as IBM PureData and Pivotal Greenplum; it is designed to match
legacy enterprise data warehouse settings such as Teradata (Özcan et al. 2014).
Accompanied with the development and release of the Vertica 7, HP integrated
“FlexZone” intended allow their users search data in huge datasets beforehand and
then the database scheme is defined with related analyses.
Based on the above highlighted new tools of big data, we can state that the
progression of big data is relatively proportionate to the innovative development of
the tools required for its analysis. The growth rate of big data is at an alarming rate
which motivates business organizations to start innovating how to collect, store,
organize, and analyze it. For a rapid progressive change to occur within the business
processes of organizations, innovation must be at the fore. Frankly, there can be no
innovation without adequate information (big data), combined with the right ana-
lytical tools and accurate interpretation from analyzed datasets.
7.7 Models of Big Data Change 139
7.7 Models of Big Data Change
As the size and velocity of data grows from gigabytes to terabytes and petabytes,
the sorts of data generated by various software applications (Web and Mobile
Based) becomes more dynamic to deal with than traditional data structures.
Therefore, traditional database structures are now faced with the challenge to
gather, store, explore, share, analyze, and visualize data. As a result many business
organizations are now adopting new data models to manage big data challenges
such as NoSQL (Not Only SQL) database, e.g. Cassandra, HBase (Couchbase
2013; Morabito 2014). In addition, they employ a distributed computing infra-
structure such as Hadoop. NoSQL databases are referred as key-value stores that are
distributed, highly scalable and most importantly non-relational.
In this section, we present rational descriptions of the data models of collecting
and storing big data from a technological perspective. Thereafter, we focus on
explaining how business organization models are changing due to the technological
advances in the data models of big data (Jinbao and Wang 2012).
7.7.1 Big Data Business Model
In this section we continue and extend the discussion in Chap. 4, examining the
model that guides business organizations in adopting and utilizing the innovative
data processing model of big data as well as innovative analytics to build a sus-
tainable competitive edge. Although within the context of the big data business
models, many business organizations are changing their processes quite carefully
due to few uncertainties about the starting process. In addition, some concerns are
also about the multitude of new technological innovations needed to be deployed in
order start a successful journey with big data (Parmar et al. 2014).
Conversely, other organizations are innovatively changing their processes at a
more rapid pace for the integration of big data analytical systems into their current
business models with the sole aim of improving their decision-making competences.
In order to get a good grasp at big data business model, the next sections
describes what is popularly known as the “Big Data Business Model Maturity” (see
Fig. 7.2) (El-Darwicheet al. 2014; Schmarzo 2012). This model provides a clear
insight on how business organizations can take advantage of the changing data
storage model in order to push the big data opportunity in a global business eco-
system (EY 2014).
7.7.2 The Maturity Phases of Big Data Business Model
1. Business Monitoring: In the business-monitoring phase organizations are

positioning Business Intelligence (BI) systems to keep track of their continuing
market performance. Business monitoring utilizes fundamental analytics to
Fig. 7.2 The maturity phases

of big data business model. Business
Adapted from El-Darwiche Metamorphosis
et al. (2014) and Schmarzo
(2012)
Data Monetization
Business Optimization
Business Insights
Business Monitoring
highlight market performance areas and sends automated alerts with relevant
information to concerned entities. One of the benefits of the business monitoring
phase is the leveraging of benchmarks retrieved from the innovative big data
storage model, which also includes indices such as product performance and
customer satisfaction for the accurate identification of less and more-performing
market areas or segments (Drake et al. 2014; De Fortuny et al. 2013; Serenko et al.
2014).
2. Business Insights: In the business insights phase, the storage model of big data)
is fully utilized. This is executed by way of making logical meaning of the “sche-
malessly” gathered data using statistics, predictive analytics and data mining in order
to discover substantial and executable market insights (Fayyad and Uthurusamy
2002). Consequently, the results of the insights analyzed are then integrated into their
existing business processes. This situation is quite similar to the concept of an
“intelligent” dashboard such as Facebook’s insights and Google Analytics but in this
case, instead of only displaying tables and graphs of data, materials and relevant
insights are uncovered. Additionally, the insights provide specific and executable
recommendations. At the later part of this chapter, a real life case study on the
utilization of big data is presented which demonstrates how the Business Model of big
data has brought innovative changes to business organization with the aid of business
insights. Nevertheless, below are some few instances of the impact of the business
insight phase on businesses:
• In marketing, the business insights could help in identifying, e.g., that certain
in-flight promotion activities are more effective than others, combined with
detailed recommendations on how much is expended on marketing in order to
shift to the more efficient activities (Eisenhardt and Martin 2000).
• In manufacturing, the business or market insights could, e.g., also help in finding
observations that some machines used for manufacturing are operating beyond
the bounds of their control coupled with a well-selected maintenance timetable
which may include part replacement for each problematic machine (Economics
2013).
• In the area of customer sales support the business insights phase has the
capabilities, e.g., of detecting observations that select “Special Card holders”
procurement activities have reduced below a particular threshold of regular
activity and then an automated alert can be sent to customers through the
monitoring phase with email information about a reduction coupon.
Thus, the business insights phase is arguably the core of big data’s business
model due to its direct connection with the data itself combined with its extensible
integration capabilities with other phases (Miluzzo 2014).
3. Business Optimization: In the business optimization phase, organizations
normally utilize embedded analytics systems in order to improve various compo-
nents of their business processes in an automated manner. For most organizations
involved with big data modeling, this is the phase where some components of
business operations can experience a massive turn-over (Gobble 2013).
Some good examples of business optimization includes:
• Expenditure budgets built upon in-promotion performance.

• Scheduling of resources founded upon customer purchase logs, and buying
behaviors.
• Inventory optimization using existing and forecasted buying patterns, alongside
local demographic and events data.
• Product valuing founded on existing buying patterns and brand interest insights
collected from social media sites.
4. Data Monetization: This phase involves the process where business orga-
nizations attempt to package their data with analytics insights for sales transactions
with other organizations. They achieve this by integrating analytics directly into
their products in order to produce “smart” products to upscale their customer
relationships and radically re-develop customer satisfaction (Kelly et al. 2014).
An example of this phase is a smartphone application whereby insights about
product performance, customer purchasing behaviors and trends of markets are sold
to marketers and promoters. For instance, the application “MapMyRun” (see
Fig. 7.3) could collate data from their smartphone application with users and
product insights in order to sell to sports wears manufacturers, retailers of sporting
goods, healthcare providers and insurance organizations.
Fig. 7.3 The “MapMyRun”

application source: captured
by the author
5. Business Metamorphosis: In the business metamorphosis phase, the critical

objective for business organizations is to leverage the captured insights about their
customers’ behavior patterns, product performance and general market trends in
order change their business models into innovative services in novel markets. In the
sections below, we present three unique examples of business organizations oper-
ating in the business-changing phases of big data business model.
7.7.3 Examples of the Business Metamorphosis Phase
Globally, many energy organizations are migrating into the “Home Energy Opti-
mization” business by providing recommendations on replacing appliances using
predictive maintenance data gathered from accurate insights (Lavalle et al. 2011;
Morabito 2014). In addition, they recommend the particular brands to buy through
decisions centered on the performance of dissimilar appliances as compared to
consumer usage patterns, weather data and ecological circumstances e.g. energy
costs and situations of local water supply.
Secondly, retailers are also migrating into the “Shopping Optimization” occu-
pation by endorsing some particular products given a customer’s existing pur-
chasing patterns as compared with other customers alike (Manyika et al. 2013).
Thirdly, major airlines migrating their business processes into the “Travel
Delight” business which exceeds the offering of rebates on air travel based on the
travelling behaviors, frequencies and preferences of customers; but also
re-searching on economically viable deals on rental cars, hotels, local tourist sites,
shows, and shopping in the areas of interest.
7.8 Big Data Change Key Issues
We suggest there are three fundamental issue areas that need to be addressed in
dealing with big data: storage issues, management issues, and processing issues. It
is worth noting that each of these represents a large set of technical research
challenges in its own right, likewise.
7.8.1 Storage Issues
The amount of information has blasted each time a new storage medium that has
been invented. What is different the most recent explosion—mainly because of
online social networking—is that current storage medium is not suitable to face it.
Besides, information is continuously made by everybody and everything (e.g.,
gadgets, and so on.), and not just, as to this point, by experts, for example,
researcher, columnists, scholars, etc. (Wielki 2013).
Current disk technology cut-off points are around 4 TB for every disk. In this
way, 1 EB would oblige 25,000 disks. Regardless of the possibility that an Exabyte
of information could be transformed on a single computer system, it would be not
able to directly attach the vital number of disks. Access to that information would
overpower current communication networks. Expecting that a 1 GB for every
second system has a compelling feasible exchange rate of 80 %, the acceptable data
transmission is around 100 MB (Russom 2013). Thus, exchanging an Exabyte
would take around 2,800 h, in the event that we accept that a maintained exchange
could be maintained. It would take more time to transmit the information from an
accumulation or capacity point to a processing point than it would to transform it.
Two results show themselves in this situation. First and foremost, process the
information “set up” and transmit just the resultant data. Consequently, “bring
the code to the information”, in contrast to the conventional technique for “bring the
information to the code” (Kaisler et al. 2013). Secondly, perform triage on the
information and transmit just that information which is critical to downstream
analysis. Notwithstanding, integrity and source metadata ought to be transmitted
alongside the actual data (Computer Research Association 2011). Also, with the
growing challenges of transmitting and storing the massive volume of data,
the reliability of results derived from downstream analysis needs to undergo
effective validation processes in order to produce accurate insights.
7.8.2 Management Issues
Management is the most challenging issue in dealing with big data. This issue initially
surfaced 10 years prior in the UK eScience activities where data was circulated
geologically by numerous elements. Determining issues of access, metadata, usage,
overhauling, and influence have ended up being significant hindrances. Dissimilar
to the accumulation of data by manual systems, where thorough conventions are
regularly followed with a specific end goal to guarantee correctness and validity,
computerized data gathering is significantly looser. The wealth of advanced data
representation precludes a tailored technique for data gathering (Kwon et al. 2014).
Data capability frequently centers more on missing data than attempting to vali-
date every item. Information is regularly fine-grained, for example, clickstream or
metering information. Given the volume, it is unrealistic to accept each data item:
new methodologies to data validation are required. The wellsprings of this data vary
—both transiently and spatially, by organization, as well as system for accumulation.
People help advanced data in mediums convenient to them: such as reports, drawings,
pictures, sound and multimedia recordings, models, programming practices, client
interface outlines, and so on—with or without sufficient metadata depicting what,
when, where, who, why and how it was gathered and its provenance. Still, these
categories of data are promptly accessible for assessment and analysis (Kwon et al.
2014).
7.8.3 Processing and Analytics Issues
The effective processing of Exabyte’s of data necessitates wide-ranging parallel

processing and novel analytics algorithms for the purpose of well-timed and
Table 7.1 Summary of differences between big data analytics and traditional data analytics
(Dandawate 2013; Yan 2013)
Big data analytics Traditional data analytics
Data structures An ecosystem appropriate for semi- An ecosystem appropriate
structured and un-structured data for structured data only
The volume unit is usually in terabyte or The volume unit is usually
petabyte in megabyte/gigabyte
Population used Data analysis of unknown population Data analysis of known
for analysis samples population samples
Technologies NoSQL approach to data Relational database models
adopted Uses Hadoop framework SQL approach
Parallel data processing Batch processing of static
data
Research and Collaborative working ecosystem Individuals can work
development independently
Traditional data are normally collected in a structured manner while big data takes an unstructured
form
7.8 Big Data Change Key Issues 145
actionable information (Kaisler et al. 2013). Traditional data processing systems are
unfit for absorbing the vast volumes of big data and cannot be executed in seconds
or a couple of minutes (see Table 7.1 for a comparison between big data analytics
and Traditional Data Analytics). There are relatively few available guidelines to
achieve expansive amounts of insights from semi- or unstructured data (Jacobs
2009). Unfortunately, this issue cannot be tackled by dimensional modeling and
online analytical processing (OLAP). To enhance a faster processing time, some
technical considerations which are calculated into the systems design, e.g. the
degree of the ratio of the speed of sequential disk reads to the speed of random
memory access. However, we aim to elucidate more on this issue in the following
section as we delve into the organizational challenges and opportunities of big data.
Big Data comes from various data generating sources such as, e.g., scientific
experiments and simulations, which are capable of producing petabytes of data on a
daily basis. A huge volume of these data are relatively irrelevant and therefore can
be filtered by orders of scale. These challenges requires urgent research in the
science of data reduction that can logically process raw data into a size that users
can deal with without losing useful insights. Likewise, online analysis methods
capable of processing such streaming data at real time are also needed to solve these
challenges. The organizational challenges discussed are concerning data acquisi-
tion, information extraction, Data Integration, Aggregation, and Representation.
7.9.1 Data Acquisition
One of the major organizational challenges is associated with the process of data
collection and processing. For instance, if a sensor analysis results vary consider-
ably from others: it is probably because the sensor is damaged, nevertheless there is
also an uncertainty if the artifact that deserves attention (Madden 2012). Addi-
tionally, the data collected by these sensors are mostly spatially and temporally
correlated (El Abbadi 2011).
Another enormous challenge is to automatically create the appropriate metadata
to portray the type of data recorded as well as how it is recorded and measured. For
instance, in experimental analysis, extensive insight in regards to particular
exploratory conditions and methods may be required to have the capacity to
translate the results accurately, and it is vital that such metadata be recorded with
observational data (Hampton et al. 2013). Metadata procurement frameworks can
minimize the human load in recording metadata. An alternate imperative issue here
is information provenance. Recording data about the information at its introduction
to the world is not valuable unless this data might be deciphered and helped along
through the data analysis pipeline. For example, a processing blunder at one stage
can render ensuing analysis futile; with suitable provenance, we can undoubtedly
distinguish all subsequent data processing that relies on this stage. Hence, research
is needed into producing suitable metadata and into data frameworks that convey
the provenance of data and its metadata through data analysis systems.
7.9.2 Information Extraction
Recurrently, the data collected exists in a format, which may be unready for
analysis. Let us consider the gathering of health records electronically, which
involves transcribed notations from numerous doctors, organized data from sensors
and measurements and image data such as X-rays. Such data cannot be left in this
form and efficiently analyzed. Reasonably, a data extraction process is needed to
draw out the essential information from the underlying sources and then present it
appropriate for analysis (Sagiroglu and Sinanc 2013).
The process of achieving this goal is an on-going technical challenge. The myth
that “Big data is always accurate” is essentially far from the truth. For instance,
hospital patients may decide to protect unsafe behavior and then medical personnel
may occasionally misdiagnose an illness. In addition, patients may also incorrectly
remember the name of a drug which could lead to misplaced information about their
medical records (Yan 2013). Current work on data cleaning accepts well-recog-
nized restrictions on well-understood error models; for many evolving big data
fields these do not exist (Blohm 2011; Williams et al. 2014).
7.9.3 Data Integration, Aggregation, and Representation
With the surge of big data, capturing and tossing it into an archive is insufficient.
Let us consider for instance the data acquired from a scope of scientific investi-
gations. In the event that we simply have a bundle of data sets in a storehouse, it is
unlikely that anyone will ever have the capacity to discover, let alone reuse the data.
For many organizations, the availability of satisfactory metadata offers hope on
better analytics, although, a few reliability challenges may be faced due to incon-
sistencies with results derived from repeated analytics (Russom 2013; Wielki
2013).
Moreover, the process of data analysis is significantly challenging, considering
that the greater part of data integration, aggregation and representation needs to
occur in a totally automated way in order to enable an efficient and extensive scale
analysis. This process requires variations in the structure of data and semantics to be
communicated in structures that are machine justifiable, and afterward mechanically
resolvable. In this regard, an extra research work is needed to accomplish a com-
puterized-fault free uniqueness determination in data structuring (Dandawate 2013).
7.10 Case Studies
In this section we discuss fact-sheets of case studies, which illustrate at a glance

how the issues presented in previous sections have been addressed by real
businesses.
The first case study considers Adobe, the multinational computer software
company headquartered in San Jose, California, United States, focused upon the
creation of multimedia and creativity software products. However, the business
model of Adobe has profoundly improved over the last few years. Alongside the
sales of software manually, Adobe has integrated cloud-based subscriptions to its
business model (Cui et al. 2014). Consequently, products now quickly released and
upgrades are incessant as compared to their previous business model of providing
upgrades every 18 months (Mullich 2014). A significant part of this innovation is
the customer interactions, which are now incessant and happening through multiple
channels, e.g. social media, e-mail, the call center, direct sales and display ads.
Also, there has been a radical transformation on the types of customer, transactions
and product data now accessible by Adobe. Although before the evolution of
concept of big data, the Adobe customer data were restricted to name, location and
billing details. The explosion of big data has pushed Adobe into collecting infor-
mation on how customers are using their products. With these sorts of data, they are
able to gain better understanding about the purchasing behavior of customers and
probably meet their future expectations (Davenport and Dyché 2013). Though
customer insights were provided through their multiple interactive channels, Adobe
encountered the problem of fractured and channel-specific views. The organization
desired to advance an all-inclusive assessment of customers to convey an extremely
custom-made, appealing user-experience. In addition to this challenge, the orga-
nization needed a real-time data on business performance information to enable
employees find the precise data they desired rather than sorting through dumps of
data.
Adobe implemented a cross-functional cooperation, integrating large sets of
data, providing a single form of performance metrics, reacting to customer indi-
cations in milliseconds and enabling data-driven exploits in order to solve the
challenges highlighted. One of the huge breakthroughs of Adobe regarding big data
is the provision of analytics and other promotion tools by offering Creative Cloud to
enable access to Adobe applications. Even though the organization was able to
monitor how customers reacted to a definite marketing inventiveness the combi-
nation of such information with customers e-mail messages and posts on social
media was quite cumbersome (Cui et al. 2014).
Point of Attention: The creation of a comprehensive strategy driven by

analytics is a core factor for managing effectively change for big data driven
innovation, due to its direct connection with the data itself combined with its
extensible integration capabilities with other business processes.
To solve the data correlational issues, Adobe executed a comprehensive data

strategy called the Adobe Dash (Mullich 2014). Adobe Dash is a commercial
intelligence platform that performs real time connections of data across numerous
sources. In-memory computing was also adopted due to the weakness of traditional
databases in aggregating data at a proportionate speed. Therefore, with the ability to
rapidly aggregate data from various channels through in-memory computing,
Adobe now provides more tailored customer interactions and a reliable user-
experience. For example, once customers click on online classified ads, Adobe
provides them a matching identity and a custom-made message built on that
information. Such personalized marketing often conforms to the preferences of their
customers.
The second case study analyses Hewlett-Packard (HP) as a leading advocate of
big data. As at the time of putting this book together, HP had an intensive supply
chain which delivers 120 PCs, 100 printers, 1,200 inks and toner cartridges, seven
network devices, seven terabytes of storage and five servers every 60 s (Prophet
2013). The greatest factor that has ensured the successes of their shipments is the
result of limitless actions and timely decisions made during the course of the
worldwide supply chain. One of the biggest challenges technology giants like HP
encounters is the means of coordinating their decisions to enhance pricing, decrease
out-of-stocks and move manufacturing to support the ever-increasing demands of
their customers. Similar to Adobe, HP has adopted in-memory computing, which
allows real-time resolutions through rapid analysis of big data. With In-memory
computing, HP is able to take speedy actions based on insights into its financial
scheduling, inventory processes and Enterprise Resource Planning (ERP).
Point of Attention: The impact of Storage and Processing hugely affects the
reliability of insights derived from big data, consequently impacting on real
time decisions made by business organizations. Therefore, organizations have
to improve the storage and processing capacity of their systems to prevent
making flawed critical business decisions.
Using the testing and experimentation big data innovative technique previously
discussed, the data query speeds of HP has been enhanced from a 2 h average with
traditional relational databases to only 88.23 s using in-memory computing
(De Fortuny et al. 2013; Jain 2011; Prophet 2013). The implication of these out-
standing results is that the employees of HP will never encounter the normal
fluctuations in processing times, which differs depending on the storage location of
the data. This has a positive effect on employee satisfaction as well as motivation
towards meeting market demands and customers’ varying desires. Employees’
motivation to fulfill market demands could be related to the efficient performance of
the system enabling them to meet various organizational goals are targets. Fur-
thermore, data can now be processed and accessible automatically due to in-built
analytics, dashboards and reporting tools, empowering employees to rapidly take
remedial actions and grab prospects. By means of improved insights into economic,
inventory and supply-chain data, HP is now able to respond to technical and
business concerns more swiftly while effectively managing risk and predicting
future customer demands just like Adobe. In today’s era of big data, the goal for an
insightful-driven analytics is to develop transforming data processes and enabling
novel capabilities for business innovation (Prophet 2013).
7.11 Recommendation for Business Organizations
As earlier mentioned, organizations are now integrating the dataset of each in-house
business units in order to pool resources that provides useful insights from large
data-sets. Some of the innovative techniques we mentioned such as, integration of
data, testing through experimentation, real-time customization, algorithmic and
automated-controlled analysis and generating data-driven models need to work a
relatively inter-connected fashion for a revolutionary innovation to occur. Never-
theless, there are gaps for improvement especially in the extraction and interpre-
tation process of big data. A more efficient method of accurate big data analytics
needs to be developed because major decisions are now being made from this
process.
Consequently, managing effectively change for big data driven innovation,
C-Level employees are now being influenced by insights gathered from big data.
This was not the case before; Big Data is now making non-technical employees
becoming relatively technical (Davenport and Patil 2012). In addition to these
discoveries, we have learnt that the distinctive advantage of big data as compared to
many other technological advances is the predictive analytical insights it offers.
With the right tools and skilled persons, the impact of big data is far-reaching
beyond what we can ever imagine. However, business organizations have a lot of
technical and non-technical modifications to make on their existing models if they
are keen to optimizing the gains of big data. Thus, adopting an incentivized
approach which will ensure that employees are encouraged to integrate big data to
their daily operations is a strategic change absolutely imperative.
From this we can understand, that the adaptation to big data and analytics must be
a culture for every employee and not only the ones that are technically savvy. Moving
along, we identified certain methodologies, which existing business organizations are
using to actively introduce new products and services. From the examples provided
on IBM, a startling discovery on how a unique business system can be developed
from the key business process of an organization was identified. This means that a
business organization has the capability of reducing its in-house managerial budgets
if the innovative methodologies discussed are implemented. The actual and potential
benefits provided by big data and analytics come with severe challenges for a
business. We explained the complex issues surrounding data acquisition, information
extraction and data aggregation as well as possible remediation techniques. So far the
capabilities of big data by far outweigh the obvious challenges encountered by many
organizations. The world of business is now experiencing a radical change not only
with scale but most importantly speed. Therefore, we highly recommend that
on-going research work and action at organizational level in this field should focus on
novel approaches for reducing errors during big data extraction and analytics pro-
cesses. This would increase the reliability of gathered insights and then lessen the
risks associated with devastating decisions made due to a faulted computerized data
structuring system.
7.12 Summary
This chapter has discussed how big data has become a key organizational asset,
which represents a strategic basis for business competition. The radical develop-
ment illustrated in previous sections is now influencing business organizations to
consider new innovative techniques on maximizing the potentials of big data. We
highlighted that the success of many organizations demands new skills as well as
new perspectives on how the epoch of big data could advance the speed of business
processes. One important factor of this technological process is the new analytics
tools that have evolved together with new progressive business models. In this
chapter, we explored the innovative capabilities of the growing big data phenom-
enon and addressed various issues concerning its methodologies for changes. Our
findings are substantiated by describing the real-life cases of Adobe and Hewlett
Packard organizations, which are globally considered as one of the biggest drivers
of big data analytics.
We have addressed various issues concerning big data’s innovation techniques
and modeling strategies adopted by many large business organizations. Without a
doubt, it is clearly seen that big data is now integrated into the business processes of
most organizations not because of the buzz it attracts but for its innovative capa-
bilities to completely transform any business landscape. Although there the inno-
vative techniques of big data are ever evolving, we were able to cover five
significant ones, which are paving way for the development of new products and
services for many organizations. We gave illustrations on Adobe and HP organi-
zations that have increased the velocity of their decision making processes as well
as improving consumer-buying behaviors through the integrated dashboard pro-
cessing analytics capabilities of big data.
References
Bain Insights: The big data boom is your company ready. http://www.forbes.com/sites/
baininsights/2013/10/30/the-big-data-boom-is-your-company-ready/ (2013)
Batty, M.: Big data, smart cities and city planning. Dialog. Hum. Geogr. 3(3), 274–279 (2013)
Bell, S.: Data-driven innovation. http://www.computerweekly.com/feature/Data-driven-innovation
(2013). Accessed 18 Nov 2014
Bjelland, O.M., Wood, R.C.: An inside view of IBM’ s “innovation jam”. MIT Sloan Manage.
Rev. 50101 (2008)
References 151
Blohm, S.: Large-Scale Pattern-Based Information Extraction from the World Wide Web. KIT
Scientific Publishing (2011)
Bollier, D., Firestone, C.M.: The Promise and Peril of Big Data. The Aspen Institute (2010)
Bughin, J., Chui, M., Manyika, J.: Clouds, big data, and smart assets: ten tech-enabled business
trends to watch. McKinsey Q. 56, 1–14 (2010)
Bughin, J., Chui, M., Manyika, J.: Ten IT-enabled business trends for the decade ahead. McKinsey
Q. (2013)
Chen, H., Chiang, R.H.L, Storey, V.C: Business intelligence and analytics: from big data to big
impact. MIS Q. 36(4), 1165–1188 (2012)
Computer Research Association: Challenges and opportunities with big data. Community white
paper. http://cra.org/ccc/docs/init/ (2011). Accessed 18 Nov 2014
Couchbase: Making the shift from relational to NoSQL. Couchbase (2010)
Couchbase: Why NoSQL? Couchbase (2013)
Cui, B., et al.: Big data: the driver for innovation in databases. Natl. Sci. Rev. 1(1), 27–30 (2014)
Dandawate, Y. (ed.): Big data: challenges and opportunities—big data: countering tomorrow’s
challenges. Infosys Labs Brief. 11(1) (2013)
Davenport, T.H., Dyché, J.: Big data in big companies. Int. Inst. Anal. (2013)
Davenport, T.H., Patil, D.J.: Data scientist: the sexiest job of the 21st century. Harv. Bus. Rev. 90,
70–76 (2012)
De Fortuny, E.J., Martens, D., Provost, F.: Predictive modeling with big data: is bigger really
better? Big Data 1(4), 215–226 (2013)
Dean, J., Ghemawat, S.: MapReduce: simplified data processing on large clusters. Commun. ACM
51(1), 1–13 (2008)
Demchenko, Y.: Defining the Big Data Architecture Framework (BDAF). SNE Group, University
of Amsterdam, Amsterdam (2013)
Drake, J.L., et al. (eds.): New Trends in Earth-Science Outreach and Engagement. Springer
International Publishing, Cham (2014)
Economics, P.: Big data, big brother, big money. Q. J. Econ. 11, 85–89 (2013)
Eisenhardt, K.M., Martin, J.A.: Dynamic capabilities: what are they? Strateg. Manag. J. 21(10–11),
1105–1121 (2000)
El Abbadi, A: Big data and cloud computing: current state and future opportunities. In:
Proceedings of the 14th International Conference on Extending Database Technology. ACM,
pp. 530–533 (2011)
El-Darwiche, B., Koch, V., Meer, D., Shehadi, R.T., Tohme, W.: Big data maturity: an action plan
for policymakers and executives. In: Bilbao-Osorio, B., Dutta, S., Lanvin, B. (eds.) The Global
Information Technology Report 2014—Rewards and Risks of Big Data. World Economic
Forum, pp. 43–51 (2014)
Elmqvist, N., Irani, P.: Ubiquitous analytics: interacting with big data anywhere, anytime.
Computer (Long. Beach. Calif.) 46(4), 86–89 (2013)
EMC: Big data transforms. The life science commercial model. EMC (2013)
EY: Big data—changing the way businesses compete and operate. EYGM Limited. (2014)
Fan, W., Bifet, A.: Mining big data: current status, and forecast to the future. ACM SIGKDD
Explor. News Lett. 14(2), 1–5 (2012)
Fayyad, U., Uthurusamy, R. (eds.): Evolving data into mining solutions for insights. Commun.
ACM 45(8), 28–31 (2002)
Gobble, M.M.: Resources: big data: the next big thing in innovation. Res. Manag. 56(1), 64–67
(2013)
Groves, P., Kayyali, B., Knott, D., Van Kuiken, S.: The “Big Data” Revolution in Healthcare-
Accelerating Value and Innovation. McKinsey & Company, Chicago (2013)
Hampton, S.E., et al.: Big data and the future of ecology. Front. Ecol. Environ. 11(3), 156–162
(2013)
Hartmann, P.M., Zaki, M., Feldmann, N., Neely, A.: Big Data for Big Business? A Taxonomy of
Data-Driven Business Models Used by Start-up Firms. University of Cambridge, Cambridge
Service Alliance (2014)
HM Revenue and Customs: Annual report and accounts 2012–13. HM Revenue and Customs
(2013)
Huo, J. et al.: BESIU physical analysis on Hadoop platform. J. Phys. Conf. Ser. 513(3), 032044
(2014)
Jacobs, A.: The pathologies of big data. Commun. ACM 52(8), 36–44 (2009)
Jain, P.: New trends and future applications/directions of institutional repositories in academic
institutions. Libr. Rev. 60(2), 125–141 (2011)
Ji, C. et al.: Big data processing: big challenges and opportunities. J. Interconnect. Netw. 13(3–4),
1250009 (2012)
Jinbao, Z., Wang, A.: Data modeling for big data. CA Technologies (2012)
Kaisler, S., et al.: Big data: issues and challenges moving forward. In: 2013 46th Hawaii
International Conference on System Sciences, pp. 995–1004 (2013)
Kelly, J., Vellante, D., Floyer, D.: Big data market size and vendor revenues. http://wikibon.org/
wiki/v/Big_Data_Market_Size_and_Vendor_Revenues (2014). Accessed 18 Nov 2014
Kwon, O., et al.: Data quality management, data usage experience and acquisition intention of big
data analytics. Int. J. Inf. Manage. 34(3), 387–394 (2014)
Lavalle, S., Lesser, E., Shock-ley, R., Hopkins, M.S., Kruschwitz, N.: Big data, analytics and the
path from insights to value. MIT Sloan Manage. Rev. 52(2) (2011)
Lohr, S.: The age of big data. http://www.nytimes.com/2012/02/12/sunday-review/big-datas-
impact-in-the-world.html?pagewanted=all&_r=0 (2012). Accessed 18 Nov 2014
Louridas, P., Ebert, C.: Embedded analytics and statistics for big data. IEEE Softw. 30(6), 33–39
(2013)
Madden, S.: From databases to big data. IEEE Internet Comput. 16, 1–2 (2012)
Manyika, J., Chui, M., Bughin, J., Dobbs, R., Bisson, P., Marrs, A.: Disruptive technologies:
advances that will transform life, business, and the global economy. McKinsey Global Institute
(2013)
McAfee, A., Brynjolfsson, E.: Big data: the management revolution. Harv. Bus. Rev. 90(10), 60–68
(2012)
Meeker, W.Q., Hong, Y.: Reliability meets big data: opportunities and challenges. Qual. Eng.
26(1), 102–116 (2014)
Miluzzo, E.: I’m Cloud 2.0, and I’m not just a data center. IEEE Internet Comput. 18(3), 73–77
(2014). doi:10.1109/MIC.2014.53
Mithas, S., Lee, M.R., Earley, S., Murugesan, S., Djavanshir, R.: Leveraging big data and business
analytics. IEEE IT Prof. 15(6), 18–20 (2013)
Publishing, Cham (2014)
Mullich, J.: Adobe gets personal with customers. Adobe Systems (2014)
Obeng-Odoom, F.: The future of our cities. Cities 26(1), 49–53 (2009)
Olson, B.M.: HADOOP: scalable, flexible data storage and analysis. IQT Q. 1(3), 14–18 (2010)
Özcan, F. et al.: Are we experiencing a big data bubble? In: Proceedings of the 2014 ACM
SIGMOD International Conference on Management of Data—SIGMOD’14, pp. 1407–1408
(2014)
Parmar, R., Cohn, D.L., Marshall, A.: Driving innovation through data. IBM Global Business
Services (2014)
Parmar, R., Mackenzie, I., Cohn, D., Gann, D.: The new patterns of innovation. Harv. Bus. Rev.
https://hbr.org/2014/01/the-new-patterns-of-innovation (2014). Accessed 18 Nov 2014
Pearson, B.T., Wegener, R.: Big data: the organizational challenge. Bain Co. (2013)
Piatetsky-Shapiro, G.: Comment on “A revolution that will transform how we live, work, and
think: an interview with the authors of big data”. Big Data 1(4), 193–193 (2013)
References 153
Prophet, T.: Investing in transportation to improve supply chain efficiency at HP. http://www8.hp.
com/hpnext/posts/investing-transportation-improve-supply-chain-efficiency-hp#.VGvO6IdJOjk
(2013). Accessed 18 Nov 2014
Provost, F., Fawcett, T.: Data science and its relationship to big data and data-driven decision
making. Big Data 1(1), 51–59 (2013)
Qubole, A.T.: How big data is revolutionizing the food industry. WIRED innovation insights.
http://www.wired.com/2014/02/big-data-revolutionizing-food-industry/ (2014). Accessed 19
Nov 2014
Russom, P.: Big data analytics. TDWI best practices report (2011)
Russom, P.: Managing big data. TDWI best practices report (2013)
Sagiroglu, S., Sinanc, D.: Big data: a review. In: 2013 International Conference on Collaboration
Technologies and Systems, pp. 42–47 (2013)
Saleem, M.A.: Lightweight stack for big data analytics. Head of Department: Dr. Steve Linton,
Supervisor: Dr. Adam Barker, Department of Computer Science, University of St Andrews,
June 2014 Declaration (2014)
Schmarzo, B.: Big data business model maturity chart—EMC2 InFocus. https://infocus.emc.com/
william_schmarzo/big-data-business-model-maturity-chart/ (2012). Accessed 18 Nov 2014
Serenko, A., Bontis, N., Hull, E.: An application of the knowledge management maturity model:
the case of credit unions. In: Proceedings of the Twentieth Americas Conference on
Information Systems, Savannah (ICIS2014), pp. 1–10 (2014)
Stonebraker, M., Madden, S., Dubey, P.: Intel “big data” science and technology center vision and
execution plan. SIGMOD Rec. 42(1), 44–49 (2013)
Taylor, W.C., Labarre, P.: Mavericks at Work: Why the Most Original Minds in Business Win.
HarperCollins e-books, New York (2006)
Wanderman-milne, S., Li, N.: Runtime code generation in Cloudera Impala. IEEE Data Eng. Bull.
37(1), 31–37 (2014)
Wielki, J.: Implementation of the big data concept in organizations—possibilities, impediments
and challenges. In: FEDCSIS, pp. 985–989 (2013)
Williams, K., Wu, J., Choudhury, S.R., Khabsa, M., Giles, C.L.: Scholarly big data information
extraction and integration in the CiteSeerX digital library. In: 2014 IEEE 30th International
Conference on Data Engineering Work (ICDEW) (2014)
Yan, J.: Big data, bigger opportunities collaborate in the era of big data—Data.gov’s roles:
promote, lead, contribute, and collaborate in the era of big data. Data.gov, U.S. General
Services Administration (2013)
Part III
Innovation Practices
Big Data and Analytics Innovation
Practices 8
Abstract
This chapter focuses on examples of data intensive innovations in practice,
providing fact-sheets of 10 interesting ideas in the field of big data and analytics
worldwide in 2014. The genesis of the selected ideas lies mostly in innovative
research projects that have developed successfully, becoming start-ups and spin-
offs and reaching the market, where they are currently applied. For each
innovation we provide an introduction to the main characteristics of the solution,
information about its developer coupled with technology and market readiness
indicators that can help in assessing the time-to-market, as well as some
indicators of user value in terms of perception, such as the user experience aspect
and the so called “Wow” effect.
8.1 Introduction
The combination of technological advances in data analytics, the increasing

availability of large data sources, the attention the topic is getting and genuine
innovative efforts have contributed to generating, in the last years, a number of
start-ups focusing on the topic. The need for connecting physical actions, experi-
ences, and characteristics with their digital counterpart is ever more urgent, as it
allows to integrate the two worlds and act on them in a multi-dimensional level.
Understanding the opportunities arising from this very dynamic area of innovation
requires to observe directly the projects and topics research centers are working on.
When the idea is valuable, the underlying research project often becomes a spin-off
or a start-up, reaching to the market in short time and developing through experi-
ence and collaboration the execution capability to make it possible. The ideas
selected and presented in this chapter originate mostly in effective research projects
launched first by excellence centers all over the World, that have been developed
and brought to the market, and are now being applied. The continuous scouting
activity conducted to prepare this part of research gives us tools to evaluate and
select interesting topics and ideas to keep in the innovation radar. This approach,

DOI 10.1007/978-3-319-10665-6_8
158 8 Big Data and Analytics Innovation Practices
together with the drivers of competitiveness relative to time-to-market and user

value, has been developed and presented in Morabito (2014).
8.2 Sociometric Solution
The key concept at the core of Sociometric Solution is the “human cue”: each
person spreads around many cues that can be captured through voice and position in
space, offering behavioral signals ready to be interpreted. The company that has
turned into practice an MIT research stream, adopts the approach of a three-stepped
process:
i. human metric identification,

ii. data sensing, by means of sensor-carrier badges,
iii. analysis and insights elaboration.
The Sociometric Badges are small sensor-rich tools that employees wear for a
certain time, allowing the collection and tracking of a vast amount of data related to
a set of human cues. Tracking includes tone of voice, time spent speaking versus
listening, connections and exchanges among individuals, movement and routes
around the offices. All data collected in this way is analyzed in order to understand
the human dynamics in the organization, through quantitative and objective data
representing, as much as possible, the reality. The kind of insights offered by
Sociometric enable the testing of hypotheses on layout, organization and team
structure that analytics intensive companies usually perform on their websites and
digital platforms, allowing to modify organizations in order to increase effective
communication and interaction between employees. As technology improves and
the data accumulate, it becomes progressively easier to make those experiments.
8.2.1 Developer
Sociometric Solutions® was founded by Ben Waber, Daniel Olguin-Olguin, Taemie

Kim, and Alex Pentland from the Human Dynamics Laboratory at MIT and
Tuomas Jaanu from Aalto University in Finland (Sociometric Solutions 2014). The
founders share research interests in the field of organizational dynamics, social
sensing technologies, “human analytics” and “weak communication signals”. While
it is widely acknowledged that effective communication and knowledge transfer are
crucial to an organization’s success, these behaviors are very difficult to measure.
According to Sociometric Solutions (2014) surveys and human observation provide
often biased and limited views into communication behaviors. Thus, their results
are actually of little practical usefulness for organizations or at least require for
additional effort in terms of generalization. Consequently, developers were strongly
committed to find a way to deploy in practice the “network intelligence” theory of
8.2 Sociometric Solution 159
Table 8.1 Company

Solution Sociometric Solutions
competitiveness indicators for
time-to-market Founded 2009
No. of products 1
Clients Enterprises
Partners Different companies
Market dimension Growing
Competitors No
Enabling infrastructure Not relevant
social signaling (Pentland 2008), describing how we can harness the intelligence of
our social network to become better managers, workers, and communicators.
In Table 8.1 there is a synthetic representation of some of the drivers of
competitiveness with a focus on a time-to-market relevant characteristics. The
measures are intended to describe the readiness of the technology, of the company,
of the market (both demand and offer side) and of the complementary infrastructure.
In particular, year of foundation and number of products are included as a proxy of
the stability of the company. Furthermore, the presence of clients substitutes a
measure of the product readiness; while the presence of strong incumbent com-
panies as partners in the development phase is another indicator of solidity.
The dimension of the market and its growth, combined with the presence of
competitors, indicate a certain level of readiness of the demand, which signals a
higher likeliness of a fast rate of diffusion of the product. Finally, the presence of
the enabling infrastructure is a complement and, thus, dependent on the readiness of
the technology itself, since whenever it is ready it can be effectively deployed only
if the infrastructure is ready as well. As appears from the table, Sociometric
Solutions is increasing its presence through different projects in a growing number
of companies.
8.2.2 Applications
The value behind such kinds of innovation lies mainly in an increased understanding
of companies’ employees behavioral patterns. The company’s research has supported
the importance of face-to-face interactions and communication, and building larger
networks among peers. One finding from the firm’s research was that simple physical
changes such as having larger tables in a cafeteria facilitated more interaction and
collaboration among colleagues than smaller tables. Another finding emphasized the
importance of group breaks and office spaces, which facilitate such breaks as having
an important effect on employee morale and efficacy. While Sociometric Solutions
collects full data through its sensors, it anonymizes the data in its customer reports,
only plotting broader patterns and connections to productivity rather than individual
tracking and performance (Sociometric Solution 2014).
Table 8.2 User value

indicators Fast learning Yes
User interface Good
User experience Positive
Process impact Medium
User feedback Positive
“Wow” effect Medium
Table 8.2 shows the set of drivers used to capture the concept of User Value. In
this case the potential user is the employee in the company conducting the
assessment, therefore a good user interface and user experience is crucial for
increasing engagement and adoption.
The application is simple and does not require much learning. The impact is
intended as a measure of how much the use of this innovation changes the estab-
lished processes for the user, therefore impacting on their willingness to change it.
The “Wow” effect is intended as the level of attraction and positive impact on the
user, which in this case is “Medium” due also to the sensitive characteristics of the
solution.
8.3 Invenio
Invenio is a process mining software that is able to design the organization processes
by reading the structured and unstructured information of a given company (Invenio
2014). Invenio processes application log files, documents, email messages, and
social signals to optimize an organization business processes. Its force resides in the
quality and efficacy of the graphic representation of the processes, even the hidden
ones, underlying companies operations. Through the interactive visualization it is
possible to dive deep into specific processes, test changes and modify directly
issuing new maps and workflows.
8.3.1 Developer
Invenio is a Malta based company, founded in 2012 with the goal of empowering
companies to better understand and map operative processes. The founding team
includes an Italian entrepreneur, who was fascinated by the prospect to understand
in detail working flows. Thanks to his 20 years experience in IT consulting he could
apply precious expertise to enhance working processes understanding.
In Table 8.3 the representation of the drivers of competitiveness on a time-to-
market basis shows, in synthesis, an advanced development of the technology
behind Invenio and a growing demand of the market.
8.3 Invenio 161
Table 8.3 Company

Solution Invenio
No. of products 1
Clients Enterprises
Partners Distribution partner
Market dimension Growing interest
Competitors Few
Enabling infrastructure Ready

User interface Positive
User experience Positive
Process impact Low
8.3.2 Applications
The application matches and confronts event data (i.e. observed behavior) and
process models (specific input or discovered automatically). This technology can be
applied to any type of operational processes. Also, potential applications can be
found where dynamic behaviors need to be related to process models, returning a
decision tree, a collection of clusters, or frequent patterns.
In Table 8.4 the drivers show a very high level of User Value. The design of the
application is simple and easy to grasp. The impact on established processes is low,
as the analysis does not require specific data collection and happens automatically.
Feedback from users has been very positive and is likely to foster a snowball effect
in adoption.
8.4 Evolv
Evolv, named one of the world’s top 10 most innovative companies in big data, is a
workforce science platform that analyzes employee performance data and provides
information for HR decision-making (Evolv 2014). Leveraging big data, it helps
solve workforce performance issues for the C-suite by using a configurable cloud
services platform. Evolv’s patented technology platform unifies and supplements
existing data from current systems, then using them to identify fact-based work-
force insights that can drive measurable ROI (Evolv 2014). Thus, Evolv exploits
data-driven methodology to help companies discover the core reasons behind
workforce performance, enabling executives to make better operational business
decisions.
Evolv was acquired in 2014 by Cornerstone OnDemand, a leader in cloud-based

applications for talent management. The company’s solutions help organizations
recruit, train, manage and engage their employees, empowering their people and
increasing workforce productivity. Thus, the acquisition of Evolv is expected to
expand Cornerstone by providing a platform enabling organizations to use relevant
internal and external data to factually evaluate the skills, work experience and
personalities of their employees and job candidates. Furthermore, Cornerstone
expects that same technology to serve as the engine to develop insights on the
massive data set accumulated by the company over the past decade from its more
than 1,800 global clients.
8.4.1 Developer
Max Symkoff co-founded Evolv in 2006 because he recognized the future

development in aligning employer needs with employee talents, while today he
supervises all aspects of Evolv strategy and operations as its CEO (Crunchbase
2014). Moreover, Jim Meyerle, Symkoff’s co-founder, thought of Evolv as a mean
to redefine how employers and job seekers go about finding each other. At first, the
challenge was making sure they had the right team to build and develop the vision
they had for using data to improve the match between employers and employees.
Neither of them had a deep technological background, while the problem they
wanted to solve required highly technical knowledge, so they hired other people
who were technically competent to do it.
Because their biggest initial risk to growth was hiring a great technical team,
they focused on making sure they had the right hiring model; thus, they established
a competency model for success within their specific company environment,
enlisting the help of industrial organizational psychologists to identify which
technical and behavioral competencies would be critical for success in a company
like theirs and they assigned each person on the hiring team to focus on evaluating
candidates for each one of these competency areas (Fast Company 2014a).
In Table 8.5 the time-to-market drivers of competitiveness display an advanced
development of the company and the market, with a ready technology and no issues
Table 8.5 Company

Solution Evolv
No. of products 1
Clients Enterprises
Partners Cornerstone performance cloud
Competitors Few
8.4 Evolv 163

indicators Fast learning Medium
User interface Very good
User experience Very good
Process impact High
of integration with established processes and enabling infrastructures, also because

it is provided as a cloud solution, while there is still a lack of competitors.
8.4.2 Applications
Evolv helps employers better understand employees and job candidates by com-
paring their skills, work experience, and personalities. However, Evolv takes it to a
deeper level, crunching more than 500 million data points on gas prices, unem-
ployment rates, and social media usage to help clients such as, e.g., like Xerox
predict, for example, when an employee is most likely to leave his job (Fast
Company 2014b). Some insights Evolv’s data scientists have discovered are, for
example, people with two social media accounts perform much higher than those
with either more or less of two, and in many careers, such as call-center work,
employees with criminal backgrounds perform better than those with clean records
(Lawrence 2014).
Table 8.6 shows an extremely high level of User Value. The application design
is simple and well-conceived, and the functionalities are straightforward and easy to
use.
The impact on established processes is high as it requires a change of mindset
and culture, although in the end the result is a simplification of an existing process.
Feedback from users has been very positive, and the reception of Evolv has been
enthusiastic at all the conventions and competitions it has taken part in, winning the
company a widespread support and a number of awards.
8.5 Essentia Analytics
Essentia Analytics’ decision support software uses latest behavioral finance

research to provide professional investors with an enhanced understanding of where
their skills lie when evaluating their own performance, so that “they can do more of
what they’re good at and less of what they’re not” (Essentia Analytics 2014).
According to Essentia Analytics (2014) all professional investors share the
problem that they are selling skill, but getting paid for performance. However,
performance is a measure of outcome and not a measure of skill, which is a function
of the decisions we make. Conscious decisions can be improved by implementing
process in a disciplined way, while unconscious decisions are more challenging,

requiring first self-awareness for optimizing their influence. Essentia’s decision
support software enables fund managers to capture richer data about their own
behavior, while improving the productivity and referenceability of their investment
processes (Essentia Analytics 2014). Indeed, on the one hand, a more accurate
picture of the past enables better conscious decisions in the future. On the other
hand, Essentia’s coaching partners identify behavioral patterns and correlations, and
reveal actionable insights, enabling awareness of the unconscious influences,
helping fund managers convert that awareness into behavioral change, thus,
increasing the probability of success.
8.5.1 Developer
Formerly a fund manager, Clare Flynn Levy is founder and CEO at Essentia
Analytics. She is an entrepreneur specialized in hedge-fund and financial software,
behavioral finance and financial analysis. Essentia Analytics is at the nexus of
Behavioral Finance and Data Science, developed by leading neuroscientists,
software engineers and ex-fund managers (Essentia Analytics 2014). The result is a
simple, accurate, continuous feedback loop, the same sort that professional athletes
use to achieve excellence. Essentia Analytics, which is headquartered in London,
was founded in 2013. The software was launched in March 2014.
In Table 8.7 competitiveness appears to be high on time-to-market drivers. The
company is growing and the technology is developing and integrating with more
partner technologies. The hedge fund sector is interested, especially in design-
intensive companies and for advanced decision making activity.
8.5.2 Applications
Appropriate for traditional active managers, hedge funds and multi-manager

platforms, high and low turnover portfolios, the software has already been adopted
by several of the world’s leading investment managers, including GLG Partners
Table 8.7 Company

Solution Essentia Analytics
No. of products 1
Clients Fund managers
Partners Growing
Competitors No
Enabling infrastructure Not relevant
8.5 Essentia Analytics 165

User interface Medium
User experience Medium
User feedback Very positive
“Wow” effect High
(Man Group), Talisman Global Asset Management, and Union Investment, despite
being less than 2 years old.
Table 8.8 suggests a high level of perception for the User Value, but with the
caveat of expectations on the application results.
8.6 Ayasdi Core
Ayasdi is a solution using a visual approach to take the guesswork out of big data
(Ayasdi Core 2014). Indeed, users of Ayasdi, such as, e.g., companies like Merck
and Citigroup, deposit related data into its platform to generate a 3D map that
uncovers new trends (Fast Company 2014b). The mathematic technique it adopts is
called topological data analysis, which analyzes the shape of complex data,
identifying clusters and their statistical significance that, finally, deliver a more
accurate “big picture” of the data, representing, for example, clusters of patients or
customers. Thus, the Ayasdi’s tools allow for automated discovery instead of the
previous model of having people type in search-style queries, which works only if a
person knows the right question to ask, as pointed out by Gurjeet Singh, co-founder
and chief executive of Ayasdi (Lohr 2013).
8.6.1 Developer
The data analysis start-up has been founded with the expertise of Gunnar Carlsson,
a professor of mathematics at Stanford University. Dr. Carlsson was the principal
investigator on research projects that were financed, from 2000 to 2008, with
$10 million from the National Science Foundation and the Defense Advanced
Research Projects Agency (DARPA) (Lohr 2013). In 2008, when Dr. Carlsson and
the other co-founders decided to attempt to commercialize the research they
received $1.25 million as a Small Business Innovation Research grant, for “high-
risk, high-payoff research” (Lohr 2013). The prototype technology has been
developed 2 years later and in 2013 it has its commercial launch. Ayasdi is not the
only big data start-up in Silicon Valley, but while most of the others are focused on
producing software to more efficiently manage data and handle queries, Ayasdi
focused and produced a new important technique of mathematical analysis.
Table 8.9 Company

Solution Ayasdi Core
No. of products 1
Clients Enterprise
Partners Growing
Competitors No
In Table 8.9 time-to-market drivers show a ready and on the market company
facing a growing demand. The company development is continuing. Competitors
are few and exhibit a more traditional approach. The infrastructure is ready as there
is no need for anything other than a PC and an Internet connection.
8.6.2 Applications
Ayasdi’s work has revealed genetic traits of cancer survivors and tracked the source
of an E. coli outbreak to name some of its applications (Fast Company 2014b).
Then, it created a visualization to help the Institute for the Study of War, a
Washington, D.C.-based think tank, to map terrorist behavior in and around
Baghdad during a campaign to free some imprisoned Al Qaeda members (Fast
Company 2014b). Ayasdi’s technology has also been used to investigate cross-
species databases to accelerate biomedical discovery and treatment and it has been
reported to be extremely good at detecting patterns and presenting them (Lohr
2013). As said above, Ayasdi has about 20 customers including Merck, the Food
and Drug Administration, the United States Department of Agriculture, and several
unnamed large corporations. They are using the company’s technology mostly for
drug discovery, oil and gas exploration, and financial-market research.
Table 8.10 shows a good level of perceived User Value, based on a great user
interface and user experience, a very high level of novelty and “Wow” effect.
Feedback from users has been very positive.

8.7 Cogito Dialog 167
8.7 Cogito Dialog
Cogito’s systems aim to constantly collect and interpret the social signals underlying
telephone conversations, video chats, and smartphone behavior (Cogito Dialog
2014). The underlying algorithms have been developed based on clinical studies and
post-traumatic stress disorder (PTSD) patients. The system automatically tracks and
analyzes facial expressions, body posture, acoustic features, linguistic patterns and
higher-level behavior descriptors (for example, attention and fidgeting), in real-time.
Cogito Dialog helps people interact on the phone by giving the accounts a real-time
visualization of how they and the customer sound as they are speaking. This
feedback promotes active listening, leading to more caring and empathetic
communications. Dialog can monitor large populations for specific behavior patterns
such as pain or engagement and inform managers as those patterns are recognized
(Cogito Dialog 2014).
8.7.1 Developer
Founded in 2006, Cogito is a Boston-based technology company developing new

ways of analyzing and understanding human communication and perception (see
also our discussion in Morabito (2014)). They combine a decade of MIT Human
Dynamics Lab research, represented by the Co-founder Professor Alex Pentland in
the Management Team, with advanced signal processing and computational
engineering capabilities, in order to track and model social signals exchanged in
conversation and monitored through Smartphone-enabled mobile sensing systems.
Cogito has been supported by Venture Mentoring System of MIT, which matches
entrepreneurs with skilled volunteer mentors selected for their experience in areas
relevant to the needs of new entrepreneurs and for their enthusiasm for the program.
In Table 8.11 the time-to-market competitiveness appears to be high, with a solid
company, facing some competition but with a large market to tap, showing no need
for additional enabling infrastructure.
Table 8.11 Company

Solution Cogito Dialog
No. of products 3
Clients Enterprises
Partners Research institutes
Market dimension Greenfield
Competitors Few

Process impact Medium to high
8.7.2 Applications
Besides its commercial application in supporting communication rich activities,

Cogito has been partnering with Raytheon-BBN and the United States Government’s
Defense Advanced Research Projects Agency (DARPA) to develop and validate a
privacy-assured, secure and scalable mobile sensing platform that gathers trustworthy
data from mobile phones, analyzes that data for patterns of psychological distress,
stores the data securely, and allows each individual to see and share feedback on their
overall mental health status. The platform has been deployed as part of a major
Boston-area clinical trial that looks at behavior from individuals experiencing PTSD
and depression and generates automatic symptom assessment models.
Companies use Cogito to gain insight into buying behavior, satisfaction, follow-
through, and intervention needs of their customers. This improves decision-making,
workflows, consistency of service, and successful interactions. Feedback from
clients reports benefits such as, e.g., increase in productivity, improved sales
performance, better data collection, and higher customer retention.
In Table 8.12 the User Value is quite high, with positive feedback from users.
The impact on existing processes is quite high, due to the fact that the solution
requires to set specific actions and procedures to make use of it, unless it is
integrated in an existing activity as an improvement of sentiment recognition.
8.8 Tracx
Tracx is a software indexing social media and delivering the most relevant, high
impact audiences and conversations about a brand (Tracx 2014). The platform allows
marketers to sort through streams of social media data and drill down to provide
geographic, demographic, and psychographic insights and to monitor performance
against competitors while planning, monitoring, engaging, and measuring influencers
from a centralized dashboard. The technology analyzes masses of data across all
social media and channels, providing insights into customer, competitor, and
influencer behaviors. It delivers the most relevant, high impact audiences and
conversations by capturing a 360° view of activity around a brand, product, or
ecosystem (Tracx 2014).
8.8 Tracx 169
Table 8.13 Company

Solution Tracx
No. of products 3
Clients Enterprises
Partners Integrators
Competitors Few
Enabling infrastructure None
8.8.1 Developer
Established in 2008, the Tracx team is made of social enthusiasts and technology
experts united in the desire to build the most compelling social media management
system. Since going live, Tracx has conquered the trust of some of the biggest
brands in the world including Microsoft, Johnnie Walker, BMW and Coca Cola.
In Table 8.13 the time-to-market competitiveness appears to be high, with a solid
company, facing some competition but with a large market to tap, showing no need
for additional enabling infrastructure as it is provided as a cloud solution.
8.8.2 Applications
The key to the startup’s value proposition is its data engine. To give its customers a
more accurate, real time view of audience segmentation across marketing, sales,
and customer relations, Tracx goes beyond “listening and monitoring” platforms
(Empson 2012). The platform satisfies different use cases, such as competitive
benchmarking, audience analysis and segmentation, influencers’ identification and
management. Part of its value and growth are also due to its development, in part
through local partnerships, of solutions also in languages other than English.
As shown in Table 8.14, the User Value is quite high, with positive feedback
from users. The impact on existing processes is not too high as it blends easily with
existing processes and systems. Bringing big data to social CRM can have a big
effect on how deep marketers are able to go in planning the best ways to reach and
engage their customers.

User interface Positive
User experience High
User feedback High
8.9 Kahuna
Kahuna’s engine provides real-time analytics, creating people profiles, which reflect
individual’s behavior and preferences across mobile and tablet (Kahuna 2014a).
People are automatically divided into subgroups based on defined criteria and
presented in clear visual images. Kahuna allows sending out personalized push
notifications without requiring spreadsheets, data analysis or SQL. Marketers can
use specific, highly targeted rules for engagement. Analytics solutions, typically,
offer dashboards that show what is going on, but the user is still needed to figure out
what to do about the analysis, which actually involves manually building marketing
campaigns through other email/push/notification systems. When the number of
campaigns increases this becomes a huge burden. Kahuna prompts marketing
campaigns automatically, checking every day to see who has fallen into “dormant”,
or signed up as s “newbie”, or has become active. Then, Kahuna automatically
activates a campaign as needed as soon as a user changes state, whether that is a
push notification, email, or any other means (Dunlap 2013).
8.9.1 Developer
Kahuna was founded in 2012, funded by Sequoia Capital, by Adam Marchick, its
CEO (Kahuna 2014b), who started coding when he was 16, building GPS tracking
systems in Matlab. The first start-up he joined had a $3B IPO (iBeam Broadcasting)
(Kahuna 2014b). For the first 7 years of his career, Adam Marchick held
Engineering and Product roles at Oracle (Siebel), Facebook and Jarna (Mobile
Apps). He received his BS in Computer Science from Stanford University, then
went back to get his MBA at the Stanford some years later. Jacob Taylor, the CTO,
started his first funded programming project at age 14, prototyping his dream of a
massively multiplayer online role playing game. He spent several years helping
startups with their technology needs and building innovative technologies. He
received his BS in Computer Science and Engineering and his MS in Artificial
Intelligence from University of California, Los Angeles.
In Table 8.15 the time-to-market drivers show a very promising company, also
given the relevance of its funding partner (Sequoia Capital). Enabling infrastructure
takes a little to get ready, but once it gets running, it just gathers results and
improves. Since most clients don’t have a lot of extra bandwidth to maintain
campaigns, Kahuna is found to be a great ROI of time invested to returns, and
anyone from an entry level marketing manager to engineer can do it.
8.9 Kahuna 171
Table 8.15 Company

Solution Kahuna
No. of products 1
Clients Digital enterprises
Partners Enterprises
Competitors Some
Enabling infrastructure In progress
8.9.2 Applications
Successful mobile-focused companies engage with their users at every stage of the
mobile lifecycle, where an important step is tailoring the communication to each
user’s current engagement state with the brand (Kahuna 2014c). There are different
types of notification-push campaigns that can be carried out, e.g., to inform new
users about the benefits of creating an account or signing in through an app or else
encouraging them to make their first purchase by targeting new users who have
shown purchase intent (such as, e.g., adding an item to their shopping cart) but have
not yet completed their purchase. In this case it is worth to send personalized
notifications based on previous user behavior and areas of interest to dormant users,
that were active in the past, making sure segmentation so that dormant users do not
receive a notification about a feature they have already tried (Marchick 2014).
These are just few examples of the possible applications, with the fundamental
purpose of making sure the app delights its users. Customized push notifications
that add real value to users’ lives are one of the way in which to drive higher app
adoption. Table 8.16 shows a good User Value, with very positive feedback from
the first adopters. The impact on existing processes is positive in terms of time-
saving and, even if there is no “Wow” effect, the value resides in simplification and
reduction of unwanted expenses.

User interface High
Process impact Low
User feedback High
8.10 RetailNext
RetailNext is leader in Applied big data for brick-and-mortar retail, delivering real-
time analytics that enable retailers and manufacturers to collect, analyze, and
visualize in-store data (RetailNext 2014a). The solution uses video analytics, Wi-Fi
detection, on-shelf sensors, data from point-of-sale systems and other sources to
automatically inform retailers about how people engage with their stores (Retailnext
2014d). RetailNext platform is highly scalable and easily integrates with promo-
tional calendars, staffing systems, and even weather services to analyze how internal
and external factors impact customer shopping patterns, thus, allowing retailers to
identify opportunities for growth, implement changes, and measure achievements
(Retailnext 2014c).
8.10.1 Developer
RetailNext was founded by engineers during at Cisco Systems, recognizing that to

stay competitive with e-commerce providers, retailers need business intelligence
tools dedicated to in-store analytics (RetailNext 2014b). Having developed
advanced solutions for a number of other industries at Cisco, the team applied their
experience to build an analytics platform that would address the needs of retailers at
global level.
In 2007, RetailNext established its headquarters in San Jose, California. It
received the support of talented professionals from companies such as, e.g., Google,
Oracle, Salesforce, Cisco, Motorola, IBM, Symantec, Intel, VeriSign, Palm
Computing, and Accenture (RetailNext 2014b). In addition, RetailNext resulted in a
combination of technology experts and professionals who have had years of
experience at companies such as Saks Fifth Avenue, Tiffany & Co, Bloomingdales,
Macys, Lancôme, Tommy Hilfiger, L’Oréal and Unilever (RetailNext 2014b).
In Table 8.17 the drivers for time-to-market describe a solid company facing a
large and growing market, with a strong link to partners in the same technological
ecosystem. Enabling infrastructure is quite ready and the demand is very large.
Table 8.17 Company

Solution RetailNext
No. of products 1
Clients Retail enterprises
Competitors Few
Enabling infrastructure Sensors and store data
8.10 RetailNext 173

User interface Good
User feedback High
“Wow” effect Low to medium
8.10.2 Applications
RetailNext is tracking over 500 million shoppers per year, collecting data from
nearly 100,000 in-store sensors across locations in 33 countries. Companies that use
RetailNext include Bloomingdales, American Apparel, Brookstone, Montblanc,
Ulta and Family Dollar. In 2013 RetailNext acquired Nearbuy, the in-store, opt-in
data tracking service, which offers shoppers free Wi-Fi in return for letting retailers
track where the customer is browsing physically and online as they traverse the
store (Shieber 2014). As reported by Perez et al. (2013), actually the company
focuses on crunching retailers’ so-called “big data” from different sources such as,
e.g., video surveillance, passive Wi-Fi tracking, point-of-sale systems, workforce
management tools, credit card transactions, weather data.
In Table 8.18 the User Value is high, with very good feedback on user interface
and experience, and a minimal process impact (it’s actually an improvement and
streamlining of existing procedures). The “Wow” effect is not very high, as in most
innovations focused on cost reduction and simplification.
8.11 Evrythng
Evrythng (2014a) is among the British startups attempting to change the face of
retail, considering the ambitious company’s plan to give every single object in the
world a unique web-addressable URL (Hern 2014). Furthermore, as pointed out by
Hern (2014). Evrythng offers also a way to solve specific “smart” issues, such as the
fact that smart fridges don’t work without smart food, and smart food won’t be sold
without smart fridges. Indeed, if manufacturers start placing unique RFID chips in
individual products to enable promotions as well as analytics, smart fridges could
take advantage of that, for example, to advise owners when food is about to go off.
8.11.1 Developer
Evrythng was founded by Andy Hobsbawm, Dominique Guinard, Niall Murphy,

and Vlad Trifa. It is based in London, United Kingdom. Niall Murphy, the CEO,
has authored and presented numerous papers around the world, including at TED.
He was a co-author of the WiFi International Roaming Access Protocols framework
Table 8.19 Company

Solution Evrythng
No. of products 1
Clients Retail enterprises
Competitors Few
Enabling infrastructure In progress
in 2005 (Evrythng 2014b). Andy Hobsbawm, Chief Marketing Officer, has been
listed among the 100 top digital influencers by Wired UK and has been a weekly
columnist about the new economy for the Financial Times, a member of
GartnerG2’s first advisory board on online advertising and spoken at numerous
conferences, including TED (Evrythng 2014b).
Dom Guinard got his Ph.D. from ETH Zurich, where he worked on laying down
the foundations of the Web of Things. He also worked 4 years for SAP on the
software aspects of the next generation platform for integrating real-world services
with business systems. Early in 2012, his Ph.D. research was granted the ETH
Medal (Evrythng 2014b). Vlad Trifa is a recognized expert in the interconnection of
networked embedded devices (sensor networks, robots, mobile phones, RFID, etc.)
with higher-level applications using Web technologies. He also worked as a
researcher in urban and mobile computing at the Senseable City Lab at MIT in USA
and Singapore and in human-robot interaction and neurosciences the ATR Inter-
national Research Center in Kyoto (Japan) (Evrythng 2014b).
In Table 8.19 the representation shows a longer time-to-market, if we consider
the absence of a true ecosystem and the lack of enabling infrastructure. However
growing, there is some doubt about the market dimensions, as the implementation
of sensors is only useful from the perspective of a network.
8.11.2 Applications
The prototypes in Evrythng’s offices in Clerkenwell, London, use QR codes and

RFID chips to achieve the goal of tracking and communicating with items, and
show applications that include, for example, bottles of whisky that can have pro-
motions personalized to the location or time they were bought, or biscuits that can
automatically redeem a free coffee on purchase from a machine (Evrythng 2014b).
In Table 8.20 the User Value is high in terms of learning and user interface,
although the process impact is quite significant. Furthermore, the “Wow” effect is
high, considering the potential impact on customer experience improvement and
supply chain management.
8.11 Evrythng 175

User interface Good
User experience Not available
Process impact High
User feedback Not available
8.12 Summary
This chapter has discussed examples of big data and analytics solutions in practice,
providing fact sheets of 10 of the most interesting ones available worldwide in
2014. The evolution trends are going to concern a further focus on convergence of
mobile services and social sensing, that is an increased exploitation of advanced
analytics for behavioral analysis from intensive data streams as well as from big
data.
Businesses and organizations from all sectors began to gain critical insights from
the structured data collected through various enterprise systems and analyzed by
commercial relational database management systems. However, over the past several
years, web intelligence, web analytics, web 2.0, and the ability to mine unstructured
user generated contents have ushered in a new data-driven era, leading to unprece-
dented intelligence on consumer opinion, customer needs, and recognizing new
business opportunities. By highlighting several applications such as e-commerce,
market intelligence, retail and sentiment analysis and by mapping important initiatives
of the current big data and analytics landscape, we hope to contribute to future sources
of value.
References
Ayasdi Core: http://www.ayasdi.com/product/ (2014). Accessed 11 Nov 2014
Cogito Dialog: http://www.cogitocorp.com/ (2014). Accessed 12 Nov 2014
Crunchbase: http://www.crunchbase.com/person/max-simkoff (2014). Accessed 18 Nov 2014
Dunlap, S.: What differentiates Kahuna mobile analytics from other mobile analytics companies? http://
www.quora.com/What-differentiates-Kahuna-mobile-analytics-from-other-mobile-analytics-
companies (2013). Accessed 18 Nov 2014
Empson, R.: Tracx secures $4.4 million to bring big data to social media management.
http://echcrunch.com/2012/02/16/tracx-secures-4-4-million-to-bring-big-data-to-social-media-
management/ (2012). Accessed 18 Nov 2014
Essentia Analytics: http://www.essentia-analytics.com (2014). Accessed 11 Nov 2014
Evolv: www.evolv.net (2014). Accessed 11 Nov 2014
Evrythng: http://www.evrythng.com (2014a). Accessed 11 Nov 2014
Evrythng: https://evrythng.com/about-us/our-team/ (2014b). Accessed 18 Nov 2014
Fast Company: http://www.fastcompany.com/3034748/worth-the-risk/how-evolv-is-arming-companies-
with-predictive-data-on-employees (2014a). Accessed 18 Nov 2014
Fast Company: http://www.fastcompany.com/most-innovative-companies/2014/industry/big-data

(2014b). Accessed 18 Nov 2014
cHern, A.: Cisco invests in UK internet of things startup Evrythng. http://www.theguardian.com/
technology/2014/apr/30/cisco-internet-of-things-startup-evrythng (2014). Accessed 18 Nov 2014
Invenio: Cognitive technology. http://www.cognitive.com.mt/ (2014). Accessed 11 Nov 2014
Kahuna: http://www.usekahuna.com/ (2014a). Accessed 12 Nov 2014
Kahuna: https://www.kahuna.com/about/ (2014b). Accessed 18 Nov 2014
Kahuna: https://www.kahuna.com/blog/2014/07/ (2014c). Accessed 18 Nov 2014
Lawrence, J.: Big Data Renders College Diplomas Worthless; Billionaires Nonplussed, San Diego
Free Press. http://sandiegofreepress.org/2014/04/big-data-renders-college-diplomas-worthless-
billionaires-nonplussed/#.VGsbJ4dJOjl (2014). Accessed 18 Nov 2014
Lohr, S.: Ayasdi: a big data start-up with a long history. Bits. http://bits.blogs.nytimes.com/2013/
01/16/ayasdi-a-big-data-start-up-with-a-long-history/?_r=0 (2013). Accessed 18 Nov 2014
Marchick, A.: 8 push notification campaigns your App should be running, right now. https://www.
linkedin.com/today/post/article/20140728170145-2185326-8-push-notification-campaigns-your-
app-should-be-running-right-now (2014). Accessed 18 Nov 2014
Morabito, V.: Trends and Challenges in Digital Business Innovation. Springer, Cham (2014)
Pentland, A.: Honest Signals—How They Shape Our World. The MIT Press, Cambridge (2008).
ISBN: 9780262162562
Perez, S.: RetailNext acquires Eric Schmidt-Backed Wi-Fi analytics company, Nearbuy Systems.
http://techcrunch.com/2013/12/03/retailnext-acquires-eric-schmidt-backed-wi-fi-analytics-company-
nearbuy-systems/ (2013). Accessed 18 Nov 2014
Retailnext: http://retailnext.net/ (2014a). Accessed 11 Nov 2014
Retailnext: http://retailnext.net/about-us/ (2014b). Accessed 18 Nov 2014
Retailnext: http://retailnext.net/press-release/retailnext-picks-up-five-accolades-recognizing-growth-
big-data-innovation-and-customer-results/ (2014c). Accessed 18 Nov 2014
Retailnext:https://www.linkedin.com/company/retailnext (2014d). Accessed 18 Nov 2014
Shieber, J.: RetailNext raises another $30 million to track in-store data. http://techcrunch.com/2014/
07/08/retailnext-raises-another-30-million-to-track-in-store-data/ (2014). Accessed 18 Nov 2014
Sociometric Solutions: http://www.sociometricsolutions.com/ (2014). Accessed 16 Nov 2014
Tracx: http://www.tracx.com/ (2014). Accessed 12 Nov 2014
Conclusion
9
Abstract
The book has discussed and presented challenges, benefits, and experiences in
big data and analytics to a composite audience of practitioners and scholars. In
this chapter conclusive remarks are provided as well as key advices for strategic
actions as a result of the issues discussed and analyzed in this volume.
9.1 Building the Big Data Intelligence Agenda
In this book we have discussed the key issues and impacts of big data and analytics
to a composite audience of practitioners and scholars. In particular, we have
focused the attention on their main strategic and organizational challenges and
benefits. Thus, we have first framed big data and analytics to question how they can
be utilized for achieving competitive advantage (Chap. 1). However, business is
only one of the domains impacted by big data and analytics, other areas concern the
public sector (Chap. 2) and education (Chap. 3), whose challenges have been
consequently investigated. No matter the context and the sector, big data and
analytics ask for a new understanding of the potential use of the actual information
growth to design appropriate business models (discussed in Chap. 4). Furthermore,
we have pointed out what needed at organizational level for improved big data
governance (Chap. 5), business oriented evaluation (Chap. 6), and managing
change for big data driven innovation (Chap. 7).
The different facets considered can be summarized in the key areas we have
identified in 2014 (Morabito 2014a) as representative of a digital business inno-
vative organization (see Fig. 9.1). In the case of big data we consider “business” in
a more general sense as “an activity that someone is engaged in” or “work that has
to be done or matters that have to be attended to” (Oxford Dictionaries 2014), thus,
encompassing both private and public organizations. Consequently, to be innova-
tive by exploiting big data and analytics in their digital business they need to take
into account:

DOI 10.1007/978-3-319-10665-6_9
178 9 Conclusion
Fig. 9.1 Key areas for digital

business innovative CONTROL COLLABORATION
organization. Adapted from
Morabito (2014a)
Digital Business
Innovative
Organization
INNOVATION
• innovation through appropriate business models (Chap. 4),

• collaboration through an effective governance (Chap. 5), and
• control through evaluation frameworks (Chap. 6), and accuracy in managing
change (Chap. 7).
As a consequence, IT leaders have to be able to combine an appropriate knowledge

of benefits and drawbacks related to big data and analytics utilization, in order to design
effective digital strategies and implement them within they organization.
As for these issues, this book has tried to provide insights as well as inspiring
“templates” for putting in practice digital business innovation through big data and
analytics. A practice we inaugurated in a former volume (Morabito 2014a), and we
actually find useful for managers know-how.
Accordingly, Chap. 8 tells what could be called “10 short stories” about those
which have been selected as interesting “global” experiences of the 2014. As for
this selection, it is worth mentioning that also in this case as in Morabito (2014a),
the choice concerns innovations that are actually applied and “in use”; thus, a
pragmatic approach have been adopted, balancing between the so called “Wow”
effect (i.e. the perceived novelty and interest in the idea), feasibility, and actual user
adoption. Consequently, not only digital innovations potentially inedited if not
disruptive, but also “ready-to-use” ones, have been selected and analyzed.
The above arguments and cases lead us to the big data lifecycle management
(shown in Fig. 9.2 and discussed in our 2014 contribution on big data, Morabito
2014b) and the main challenges and IT actions identified there for big data for
business value:
• Convergence information sources: IT in the organization must enable the con-

struction of a “data asset” from internal and external sources, unique, integrated
and of quality.
• Data architecture: IT must support the storage and enable the extraction of
valuable information from structured, semi-structured as well as unstructured
data (images, recordings, etc.).
• Information infrastructure: IT must define models and adopt techniques for
allowing modular and flexible access to information and analysis of data across
9.1 Building the Big Data Intelligence Agenda 179
Technologicalperspective Business perspective
Analytics
BIG DATA Business

Information systems Processes
Management
Low Value of information High

50% of managers say they
Executives often have to make do not have access to the
decisions based on information information they need to
they do not trust , or they do perform their jobs.
not have. High Volume of data Low
Fig. 9.2 Big Data management challenges. Adapted from Morabito (2014a), Pospiech and Felden
(2012)
the enterprise. Furthermore, organizations must commit human resources in

recruiting and empowering data scientist skills and capabilities across business
lines and management.
• Investments: The IT and the business executives must share decisions on the
budget for the management and innovation of information assets.
Taking these issues into account, as also discussed in Morabito (2014b), big data
and analytics are key components of the digital asset of today’s organizations (as
shown in Fig. 9.3). Indeed, business decisions and actions rely on the digital asset
of an organization, although requiring different types of orientation in managing the
Decisions Actions
IT PORTFOLIO
Δ (Services)
Process Change
Orientation Orientation
Application
ENVIRONMENT
COMPETITIVE
(Outer Context)
integration
Integration DIGITAL
Orientation IS Organizational Absorptive Capacity
ASSET
Data Integration
Analytics Information
Orientation Orientation
Δ (Data)
DATA ASSET
Fig. 9.3 A framework for managing digital asset. Adapted from Morabito (2014b)
180 9 Conclusion
information systems (IS). As for decisions, integration orientation seems to be

required for satisfying the needs for optimization and effective data management of
big data. Indeed, the greater the integration of a company’s information system, the
faster the overall planning and control cycles (Morabito 2013). Therefore, applying
to big data and analytics issues our SIGMA model (discussed in Morabito 2013),
we argue that integration orientation constitutes a fundamental lever for facilitating
the absorption and transformation of information and knowledge coming from big
data and analytics into evidence-driven actions, helping manager’s decision-making
and employees perform their work (Morabito 2014b). Furthermore, integration
orientation is one of the determinants of organizational absorptive capacity, which,
in turn, is theorized to affect business performance (Morabito 2013; Francalanci and
Morabito 2008), thus, measuring the ability of an organization to cope with IT
complexity or in our case with big data management and use by businesses. As a
consequence, moving from decisions to action calls for an organization to improve
IS absorptive capacity (Morabito 2013; Francalanci and Morabito 2008) in terms of
the set of key orientations considered in the above mentioned SIGMA approach:
analytics, information, process, and change orientation.
Considering these issues and what discussed in previous chapters, we point out
that the framework in Fig. 9.2 is suitable to provide a systemic and integrated
“working” representation of factors and drivers involved in managing digital assets,
which aim to exploit the opportunities of big data and analytics for business per-
formance and value. Finally, taking all the above issues into account, we hope the
book has provided a toolbox for managerial actions in building what we call a big
data intelligence agenda (Morabito 2014b).
References
Francalanci, C., Morabito, V.: IS integration and business performance: the mediation effect of
organizational absorptive capacity in SMEs. J. Inf. Technol. 23, 297–312 (2008)
Morabito, V.: Business Technology Organization—Managing Digital Information Technology for
Value Creation—The SIGMA Approach. Springer, Berlin (2013)
Morabito, V.: Trends and Challenges in Digital Business Innovation. Springer, Berlin (2014a)
Morabito, V.: Big data. In: Trends and Challenges in Digital Business Innovation, pp. 3–21.
Springer, New York (2014b)
Oxford Dictionaries: business. http://www.oxforddictionaries.com/definition/english/business. Accessed
16 Nov 2014
Pospiech, M., Felden, C.: Big data—a state-of-the-art. In: AMCIS 2012
Index
A Cognitive code, 69
Academic analytics, 54 Competitive advantage, 4
Adobe, 125, 147–150 Coolest cooler, 74
Airbnb, 72 Cornerstone, 162
Amazon, 12, 15, 108, 109, 111, 115, 118, 120, Coursera, 50–52, 58
122 Course Resource Appraisal Model (CRAM),
Amazon web service, 120 56
Apple, 8 Crowdflower, 38
Applied big data, 172 Crowdfunded, 9
Artificial Intelligence (AI), 72 Crowdfunding, 10, 74, 75
Asos.com, 8 Crowdreporting, 26, 27
Crowdsourced, 9
Crowdsourcing, 23, 24, 26, 29, 30, 38, 39, 42,
B 67, 70, 73, 75, 76
22@ Barcelona, 36 Customer Relationship Management (CRM),
Banorte, 71 71
Barcelona, 23, 36–38, 42
Behavior, 110, 111, 113
Behavioral finance, 164 D
Big data driven business models, 66, 79 3D printer, 10
Big data governance, 83 3Sage, 87
Big data management (BDM), 99 Data analysts, 35
Bitcoin, 68, 72, 73 DataFlux, 94, 95
Business model, 65, 66, 68, 69, 73, 77–79 Data governance, 84, 87, 94, 95–97
Business process integration, 89 Data mining, 107, 122
Data ownership, 23, 31, 33, 42
Data quality, 34, 39, 86, 89, 98
C Data science, 164
California Report Card, 41 Data scientists, 12
Capability Maturity Model (CMM), 95 The Data Warehousing Institute (TDWI), 91
Center of Excellence (CoE), 134 Deep thunder, 68
Chief Executive Officer (CEO), 130 Demographics, 110, 111
Chief Information Officers (CIOs), 130 Distance learning, 47, 48, 62
Chief Marketing Officer (CMO), 130
Citidirect BeMobile, 136
Citysourced.com, 24 E
Cloud computing, 31, 32, 120, 121 eBay, 108, 109, 111, 114, 115, 123
Cloudera, 138 e-commerce, 4
CMOOCs, 52–54, 58, 62 e-government, 23, 32
Cognitive city, 27 Edification, 62

DOI 10.1007/978-3-319-10665-6
182 Index
edX, 50–52, 60 Key Performance Indicators (KPIs), 75

Engine Health Management (EHM), 135 Kickstarter, 74
Enterprise information management (EIM), 95 Kickstarter.com, 10
Enterprise resource planning (ERP), 4
L
F Learning analytics, 55–57
Facebook, 26, 30, 34, 76, 85, 106, 108, 116– Learning Management System (LMS), 52
119, 121, 122, 126, 140, 170 Lending, 127
Financial markets, 127 Lifestyle, 110
FutureLearn, 50, 51 LinkedIn, 6, 85
Livemocha, 61, 63
G
Gamification, 6, 57, 61, 63 M
Gartner maturity model, 95, 96 MapMyRun, 141, 142
Google, 11, 12, 15, 73, 76, 105, 108, 111, 123, Marine Institute of Ireland, 136
126, 134, 140 Master data integration, 89
Google ventures, 73 McKinsey Global Institute, 27, 32
GPS, 83, 86 Megacities, 25
Greece, 25, 31 Metadata management, 89
Grid computing, 32 Meteorological Assimilation
Groupon, 5, 10, 14, 15, 19 Data Ingest System (MADIS), 27
Microsoft, 115
Money management, 127
H MySpace, 116
Hadoop, 34, 71, 114, 118, 122, 123, 132, 137–
139
Haiti, 23, 38, 42 N
HarvardX, 60 Neighborhood planning, 30
Health informatics, 33 Netflix, 108, 110, 116, 118–123
Her Majesty’s Revenue and Customs Network intelligence, 158
(HMRC), 131 NoSQL, 132, 139, 151
Hewlett Packard, 125, 150
Human metric identification, 158
O
One Laptop Per Child (OLPC), 48
I Open Courseware, 49
IBM, 11, 68, 95–97, 115, 136, 138, 149 Open government, 23, 24, 31, 36
Information governance, 85, 87 Open innovation, 9
Information lifecycle management, 86, 89 Open Universities, 47
Innocentive, 9 Oracle, 115, 170, 172
Instagram, 116 Ownership of, 79
InStedd, 38
Intellectual property (IP), 9
Internet of Everything, 27, 28 P
Internet of things (IoTs), 8, 27, 66, 75, 76 PartyX, 25
iPhone, 8 Peer 2 Peer University, 50
IT service delivery, 32 Pharmaceuticals, 5, 15–18
Predictive analytics, 30, 33
Privacy, 34
K Privacy and security, 114, 122
Kaggle, 70 Profiling, 34, 35, 41
Kearney, A.T., 6 Prosumers, 42
Index 183
Public–Private Partnerships, 31 T
Public service management, 31 Topological data analysis, 165
Twitter, 26, 85, 116, 123
R
Real-estate, 127 U
RedLaser, 109 Udacity, 50, 51, 58
RFID, 83, 86, 173, 174 Udemy, 50
Riff, 38 Urban Habitat, 36
Ripple, 73 Ushahidi, 38
Rolls-Royce, 135 Utility from, 78
S V
SeeClickFix.com, 26 Virtonomics, 57
Security and privacy, 89 Virtualization, 32
Sequoia Capital, 170 Visualization, 108, 122
Servitization, 75
Skype, 5, 14
Smart city, 26, 28, 29, 36–40 W
Smart City Personal Management Waze, 136
Office, 36 Weather underground, 27
Smart grid technology, 33
Social entrepreneurs, 9
Social sensing technologies, 158 X
Social signals, 160, 167 Xerox, 163
Software Engineering Institute (SEI), 95 xMOOCs, 51–55, 57, 62
SpigitEngage, 73
SQL, 170

Big Data and Analytics: Vincenzo Morabito

Uploaded by

Copyright:

Available Formats

Big Data and Analytics: Vincenzo Morabito

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

Big Data and Analytics: Vincenzo Morabito

Uploaded by

Copyright:

Available Formats

How does the book aim to develop the strategic and organizational impacts of Big Data and analytics?

How does the book aim to develop the strategic and organizational impacts of Big Data and analytics?

What are some examples mentioned of how organizations can extract insights and value from big data?

What are some examples mentioned of how organizations can extract insights and value from big data?

Vincenzo Morabito

Big Data and

Big Data and Analytics

ISBN 978-3-319-10664-9 ISBN 978-3-319-10665-6 (eBook)

Library of Congress Control Number: 2014958989

Springer Cham Heidelberg New York Dordrecht London

Printed on acid-free paper

Springer International Publishing AG Switzerland is part of Springer Science+Business Media

dynamic ties, pricing of digital media, anonymity in online relationships, to

Big Data Drivers and Characteristics

• Support customer-facing processes: e.g., to identify fraud or medical patients’

Marketing/ BIG DATA Utilities /

Fig. 1 Big Data Applications. Adapted from (Morabito 2014)

Outline of the Book

Scattaretico, Claudio Basso, Mauro Pianezzola, Marco Zanussi, Davide Carteri,

1 Big Data and Analytics for Competitive Advantage . . . . . . . . . . . . 3

2 Big Data and Analytics for Government Innovation. . . . . ....... 23

3 Big Data and Education: Massive Digital Education Systems . . . . . 47

4 Big Data Driven Business Models . . . . . . . . . . . . . . . . . . . . . . . . . 65

5 Big Data Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

6 Big Data and Digital Business Evaluation . . . . . . . . . . . . . . . . . . . 105

6.3 Organizational Advantages and Opportunities . . . . . . . . . . . . . 108

7 Managing Change for Big Data Driven Innovation . . . . . . . . . . . . 125

7.8 Big Data Change Key Issues . . . . . . . . . . . . . . . . . . . . . . . . . 143

Part III Innovation Practices

8 Big Data and Analytics Innovation Practices . . . . . . . . . . . . . . . . . 157

8.10 RetailNext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

ACID Atomicity, Consistency, Isolation, and Durability

MOOCs Massive open online courses

The concept of competitive advantage has created kaleidoscope of perspectives

© Springer International Publishing Switzerland 2015 3

Before we discuss, however, the potential implication of big data on business

1.2 Competitive Advantage Definition: Old and New

As popularized by Michael Porter’s back in the 80s, competitive advantage denotes

1.2.1 From Sustainable to Dynamic

Globalization and e-commerce are radically changing consumer preferences and a

1.2.2 From Company Effects to Network Success

Such unpredictably changing market conditions were deﬁned by Williamson as

1.3 The Role of Big Data on Gaining Dynamic

1.3.1 Big Data Driven Target Marketing

Cross-selling, for example, leverages a company’s afﬁnity and knowledge of its

1.3.2 Design-Driven Innovation

1.3.3 Crowd Innovation

1.4 Big Data Driven Business Models

The emergence of mobile phones is a potentially lucrative media platform for

1.5 Organizational Challenges

1.5.1 Skill Set Shortages

1.5.2 Cultural Barriers

While market-driven innovations require insight, design-driven innovations call for

1.5.3 Processes and Structures

1.5.4 Technology Maturity Levels

1.5.5 Organizational Advantages and Opportunities

a. Improve decision making by lowering the cost of better quality information

c. Improve collaboration by developing a common, enterprise-wide business

1.6 Case Studies