See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319543740

The Future of Software Development Methods

Chapter · January 2018

DOI: 10.4324/9781315619361-11

CITATION READS

1 1,792

2 authors:

Brian Fitzgerald Klaas-Jan Stol

University of Limerick University College Cork
254 PUBLICATIONS   9,794 CITATIONS    81 PUBLICATIONS   2,049 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Episodic Volunteering in Free/Libre/Open Source Software View project

SCALARE: Scaling Software View project

All content following this page was uploaded by Klaas-Jan Stol on 10 September 2018.

The user has requested enhancement of the downloaded file.

B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

9
THE FUTURE OF SOFTWARE
DEVELOPMENT METHODS

Brian Fitzgerald and Klaas-Jan Stol

Introduction
There is a distinct half-life obsolescence with respect to software development methods, in that the
principles and tenets that become enshrined in methods can become obsolete due to the emergence of
new challenges and issues as the technological environment evolves. This has been labeled the ‘problem
of tenses’ (cf. Fitzgerald 2000; Friedman 1989). The software development methods and processes that
are popular today have been derived from principles that were first identified many decades ago – the
systems development life cycle, object orientation, agile and lean methods, open source software,
software product lines, and software patterns, for example. However, there are a number of
fundamentally disruptive technological trends that suggest that we need an ‘update of tenses’ in relation
to the software development methods that can cater to this brave new world of software development.
Fitzgerald (2012) has coined the term Software Crisis 2.0 to refer to this new disruptive age. The original
software crisis (Software Crisis 1.0, as we might refer to it), identified in the 1960s, referred to the fact
that that software took longer to develop and cost more than estimated, and did not work very well
when eventually delivered. The diligent efforts of many researchers and practitioners has had a positive
outcome in that this initial software crisis has been resolved to the extent that software is one of the
success stories of modern life.

Software Crisis 2.0

Notwithstanding the achievements made that have helped to address Software Crisis 1.0, this success has
also helped to fuel Software Crisis 2.0. Fitzgerald (2012) identifies a number of ‘push factors’ and ‘pull
factors’ that combine to cause Software Crisis 2.0. Push factors include advances in hardware such as that
perennially afforded by Moore’s law, multiprocessor and parallel computing, big memory servers, IBM’s
Watson platform for cognitive computing, and quantum computing. Also, developments such as the
Internet of Things (IoT) and Systems of Systems (SoS) have led to unimaginable amounts of raw data
which fuel the field of data analytics – a field that is commonly referred to as “big data.” Pull factors
include the insatiable appetite of ‘digital native’ consumers – those who have never known life without
computer technology – for new applications to deliver initiatives such as the quantified self, life-logging,
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

and wearable computing. Also, the increasing role of software is evident in the concept of software
defined * (where * can refer to networking, infrastructure, data center, or enterprise). The Software
Crisis 2.0 bottleneck arises from the inability to produce the volume of software necessary to leverage
the absolutely staggering increase in the volume of data being generated in turn allied to the enormous
amount of computational power offered by the many hardware devices also available, and both
complemented by the demands of the newly emerged digital native consumer in a world where
increasingly software is the key enabler. Figure 9.1 summarizes these ‘push’ and ‘pull’ factors.

“Software is eating the world!”

This quote from Netscape founder Marc Andreessen colorfully captures the extent to which software is
becoming predominant, to the extent that we claim that all companies are becoming software companies.

Our organization has become a software company. The problem is that our engineers haven’t
realized that yet!

This is how the vice president for research of a major semiconductor manufacturing company,
traditionally seen as the classic hardware company, characterized the context in which software solutions
were replacing hardware in delivering his company’s products. This organization knew precisely the
threshold of reuse level for its hardware components before designing for reuse became cost-effective.
However, this level of sophistication was not yet present in its software development processes. There
was a tendency to approach each software project in a once-off fashion, and a repeatable and formalized
software development process had not been fully enacted.

Hardware Advances
• Moore’s Law, Butter’s
Law, Kryder’s Law
• Parallel computing
• IBM Watson platform
• Quantum computer/
Memcomputer

Digital Natives
• Crowdsourcing/
Big Data
crowd-science
•
•
IoT
Ubiquitous devices
Software • Quantified self,
• Systems of Systems Crisis 2.0 Life-logging,
Wearable computing
• Columnar databases
• Mass customization
• Continuous Deployment/
Innovation

Software-Defined *
• SD Networking
• SD Infrastructure
• SD Data Center
• SD Enterprise

Figure 9.1 Software Crisis 2.0

B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

We have seen this situation replicated across several business domains as the transformation to software
has been taking place for quite some time. The telecommunications industry began the move to
softwareization in the 1970s with the introduction of computerized switches, and currently, the mobile
telephony market is heavily software focused. The automotive industry has very noticeably been moving
toward softwareization since the 1960s – today, 80%–90% of innovations in the automotive industry are
enabled by software (Mossinger 2010; Swedsoft 2010). This is evidenced in the dramatic increase in the
numbers of software engineers being employed in proportion to the numbers employed in traditional
engineering roles. A striking example of the growing importance of software in the automotive industry
is conveyed in the following: In 1978, a printout of the lines of code would have made a stack about
twelve centimeters high; by 1995, this was already a stack three meters high; and by 2015, it is a staggering
830 meters, higher than the Burj Khalifa – the tallest man-made structure in the world (see Figure 9.2).
Another example of a domain in which software has increased dramatically in importance is the
medical device sector. Traditionally, medical devices were primarily hardware with perhaps some
embedded software. However, a 2010 EU Medical Device Directive classifies stand-alone software
applications as active medical devices (McHugh et al. 2011). This has major implications for the software
development process in many organizations, as they now find themselves subject to regulatory bodies
such as the US Food and Drug Administration (FDA).
Thus, we have a context where software, traditionally seen as secondary and a means to an end in
many sectors, moves center stage. The implications of this global shift are frequently underestimated. It
requires the software development function to transform itself in order to provide the necessary
foundation to fulfill this central role, which in turn requires an expansion of the set of concerns that need
to be integrated into any software solution. The increasing importance of software is not only a matter
of ‘scale’ in the traditional sense, measured in lines of code, transactions per second, or capacity, but the
scaling of software causes the need for changes in other dimensions, too. This has implications for the
methods that we use to guide software development practice.

Figure 9.2 Height of software printout stack in Mercedes S-Class

Source: Schneider (2015).

B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

The future of software development methods: an expanded view

This expanded role for the software development function requires a multi-dimensional perspective.
Figure 9.3 illustrates three inter-related dimensions that have been drivers for software scaling over the
past 20 years. The inner, dark-shaded circle represents the traditional view that was concerned with
individual products developed within traditional organizational boundaries using conventional software
development processes – or as is the case in many organizations, no formal development process at all.
However, the move to services, the emergence of topics such as open innovation, organizational
ecosystems, various forms of sourcing (outsourcing, open-sourcing, innersourcing, crowdsourcing; we
use the term ‘*-sourcing’) (Ågerfalk et al. 2015a; 2015b), and IoT have caused an expanded focus in
each of these dimensions. We characterize each of these dimensions in more detail later in this chapter.

Scaling products, systems and services

Companies are also seeking to extend beyond the current state-of-the-art in service-oriented
architectures (SOA) to build products. Many companies are actively using open source components and
are also providing middleware functionality to provide the glue layer which allows them to incorporate
various popular social media applications into their product and service offerings. Software reuse,
component-based software development and software product line (SPL) approaches are established
means that facilitate the scaling of software systems in the traditional sense.

Products, Systems Organizations and

and Services Partnerships and Business domains
ecosystems
Exemplar trends:
Closed and Outsourcing
Softwareization in-house Opensourcing
and Servicetization development
Exemplar trends: Crowdsourcing
SOA Standalone Innersourcing
SPL systems Ecosystems
IoTS GSD
CBSD Traditional
Big Data Software
Cloud computing functions

The Software
Enterprise Exemplar trends:
DevOps / Continuous Delivery
Enterprise Agile
Processes and Methods Lean software development
Continuous Software Engineering
Agile in regulated domains

Figure 9.3 Software scaling as a multi-dimensional phenomenon

B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

It is estimated that 35 billion devices are currently connected in the IoT scenario, a figure that is
expected to rise to 100 billion by 2020 (Feki et al. 2013). The exponential increase in the volume of
available data – Big Data – allied to the ubiquitous devices that are available to process the data also leads
to a demand for additional software systems to process into market relevant knowledge. So, not only do
software systems become bigger, the number of systems (however small these may be) is growing
exponentially.
Another trend is that of servitization. Rather than selling products to customers, organizations sell
services that could not exist without software. Airbnb is the world’s largest supplier of accommodation,
but the company owns no real estate. Likewise, Uber is the world’s largest taxi company, but owns no
taxis of its own. This shift from products to services is continuing and has significant implications for
how companies conduct business.

Scaling processes and methods

Software processes and methods are subject to scaling, as building larger systems requires adjusting those
processes. New software development paradigms, such as Open Source (Feller et al. 2005), Inner Source
(Stol and Fitzgerald 2015), lean software development (Fitzgerald et al. 2014; Poppendieck and
Cusumano 2012), and DevOps (Fitzgerald and Stol, 2017), are emerging, and are at varying levels of
maturity. Global software development with distributed teams, for example, requires that methods that
were not intended for such settings (e.g., agile methods) be tailored. Furthermore, the increasing reliance
and prevalence of software solutions in regulated domains (e.g., automotive, medical, and financial
sectors, as mentioned earlier) also requires tailoring of both new and traditional development approaches.
While the move from traditional to agile methods is well underway, it is worth noting that agile
methods were originally proposed as only suited to small projects with co-located teams developing
software in non-critical domains (Boehm 2002). While agile has moved well beyond this space, with
frequent use of agile methods on projects with globally distributed teams, many challenges remain in
relation to social issues such as cultural differences, communication breakdowns, and optimum practices
for distributing development work across sites. Furthermore, the need for Enterprise Agile, or Agile 2.0,
whereby agile principles need to permeate beyond the software development function to other
organizational functions such as finance, marketing, and human resources, is very apparent. Individual
successes such as mentioned earlier rely heavily on their particular organizational context, which is
increasingly recognized as an important factor (Dybå et al 2012).
Agile methods and regulated environments are often seen as incommensurable (Turk et al. 2005).
The reason for this is probably evident in the Agile Manifesto, which identifies a number of fundamental
value propositions for agile – for example, that working software and developer interaction should be
viewed as more important than documentation or processes and plans. In regulated environments where
traceability is paramount, documentation, processes, and plans are critical. Likewise, in regulated
domains, the agile goal of improving time-to-market is secondary to the need to ensure safety is never
compromised to satisfy market demands, as the consequences of system failure can lead to loss of life as
well as multi-million-dollar compensation claims. Nevertheless, there are encouraging signs to suggest
that agile methods can be successfully scaled to regulated environments (Fitzgerald et al. 2013), although
this will require organizational changes as well.
Software systems are becoming larger. The obvious expansion is first in terms of lines of code, as
demonstrated in Figure 9.2. For example, an average Fortune 100 company maintains 35 million lines
of code, adding about 10% each year in enhancements alone – as a result, the amount of code maintained
was previously estimated to double every seven years (Muller et al. 1994), but with the ever-increasing
ubiquity of computing devices, this is likely to be a conservative estimate. Furthermore, many
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

organizations are faced with mission-critical legacy software systems that are written in older technologies
such as COBOL or even assembly language. This represents major risks for many organizations that wish
to modernize their systems. Again, there is a tension in relation to the use of agile methods for such
modernization projects. Certainly, the use of agile methods in large development projects remains a
significant challenge (Booch 2015), with little empirical evidence of the successful deployment of agile
in large projects (e.g., Cao et al. 2004; Kähkönen 2004). The extant literature tends to assume that scaling
agile methods can be done in a linear fashion; Rolland et al. (2016) have argued that this approach has
limitations and that such assumptions must be re-evaluated.

Scaling organizations and business domains

As software assumes greater importance in companies, traditional concepts such as software architecture,
software product lines, and software reuse, formalized software development processes become more
important. These in turn require significant organizational change in order to be successfully
implemented. A major task here is the education of management and the broader engineering workforce
to the complexities and challenges of software development and the benefits of a systematic approach in
the areas just mentioned.
Software organizations are now adopting global software development (GSD) strategies such as
opensourcing, outsourcing and offshoring, and even crowdsourcing to accommodate the increasing need
to deliver larger software systems more quickly in order to stay competitive (Stol and Fitzgerald 2014).
Acquisition of companies that possess critical technology or software resources is another common trend.
Ensuring a proper governance model and selecting the right partnerships are key challenges. Another
trend is that organizations are entering new domains as they increasingly rely on software. Domains that
were traditionally dominated by hardware-based solutions are now moving to software solutions.
Consequently, companies find themselves in new business markets with new competitors.
Another phenomenon which is relevant in this context is the erosion of strict organizational
boundaries as organizations seek to leverage open innovation. One concrete form of open innovation is
the building of organizational partnerships or ecosystems with suppliers (e.g., commercial off-the-shelf
(COTS) vendors and open source software (OSS) communities) (Bosch and Bosch-Sijtsema 2010). A
topic that has started to draw more attention is the creation of open source–style communities within
one or a consortium of commercial organizations: (proprietary) software is developed in-house using
OSS development practices (Stol and Fitzgerald 2015). This is called Inner Source, and has been adopted
in several large organizations such as Allstate, Bloomberg, Bosch, Ericsson, Paypal, Philips, and Sony
Mobile. There is growing evidence that an organization’s people and culture play a pivotal role in
adopting new methods (Conboy et al. 2011). However, there is little insight into how “Communities
of Practice” (CoP) can be built within commercial organizations, which must take business
considerations and sustainability into account (Tamburri et al. 2013).

Industry vignettes
We believe the three dimensions in Figure 9.3 are inter-related and suggest that to scale successfully in
any one dimension requires that related challenges in the other dimensions be resolved simultaneously.
This can be difficult as the software development function in many organizations today is not in a position
to mandate organizational change in other organizational functions, yet the latter is critical for success.
A case in point is the current focus on DevOps and continuous deployment, which requires buy-in from
an operations team to be able to deliver new versions at a fast pace (Stol and Fitzgerald 2014).
We illustrate these challenges with a number of industry vignettes describing real-world scenarios
where organizations are faced with these challenges. These vignettes represent real organizational
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

contexts in which we have been involved through funded research projects (Fitzgerald et al. 2013). Each
vignette is primarily driven from one dimension but challenges in terms of the other dimensions are
clearly evident. The first vignette presents the case of QUMAS and relates to the Scaling Processes and
Methods dimension.

Scaling processes and methods: the case of QUMAS

QUMAS is an example of a company developing software for a regulated market. The software
development process in QUMAS must adhere to a number of standards. QUMAS previously
employed a waterfall-based approach. However, this approach resulted in a long time-to-market
and a large release overhead, which were seen as drawbacks in the quickly changing market that
QUMAS is operating in. As a consequence, they have adopted and augmented the Scrum
methodology.
Agile methods were initially assumed to be limited to small development projects with co-
located teams working on non-safety critical development projects. In order to introduce a new
agile process, QUMAS have to interact with external organizations in the regulatory bodies that
monitor compliance with the relevant standards. Also, the agile process that QUMAS have adopted
requires the Quality Assurance (QA) function to assess compliance of the software produced at the
end of each three-week sprint. This was a fundamental change in work practice for QA as they
currently assess regulatory compliance more or less annually to coincide with new releases of the
product. Because QA are required to be independent of the software development function in a
regulated environment, senior management need to support such a large organizational change
initiative.
QUMAS see this approach as also altering the product and services they offer. New
functionality in interim software products at the end of sprints can be demonstrated to customers
and their feedback sought more frequently than in the typically annual big-bang release of a new
product as per the traditional waterfall-based process.

QUMAS’s transformation started on the processes and methods dimension, but resulted in significant
changes in the product and organizational dimensions. The need to involve other parts of the
organization to make a process transformation is clearly relevant to the organizational dimension. In
regulated domains, the QA function is required to be an independent department from the software
development function, and thus the agile transformation also required other parts of the organization to
transform. The change to an incremental process facilitated the concept of continuous compliance.
Whereas the waterfall approach validated the product under development only at the very end, in the
new process the compliance auditing took place incrementally as well. The incremental development
and frequent delivery also facilitated presales of the product before the software was finished – something
unheard of in a waterfall approach. Finally, given the new incremental development approach also
affected the implementation of the product, as its design now had to facilitate incremental addition of
features.
The second vignette presents the case of Husqvarna, a leading manufacturer of gardening machinery,
and represents a company transforming primarily in relation to the Products, Systems and Services
dimension, but with major implications for the Organization and Business Domains and the Processes
and Methods dimensions.
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

Scaling products, systems and services: the case of Husqvarna

Husqvarna is an example of a company on the journey of transformation to becoming a software

company. Ten years ago, Husqvarna was primarily a retailer of gas-powered machinery. Now, an
increasing number of machines are battery powered. This changes dramatically the competitive
business domain in which Husqvarna operates, as they are now faced with new competitors.
Software is now becoming a key factor. In the past, software development was often done by
the robotics staff at Husqvarna, who would have taken programming courses at college. Software
development was perceived in Husqvarna as largely equivalent to coding. However, it is estimated
that coding only represents about 7% of the effort in software development, and the majority of the
work has to do with requirements, architecture and system design, documentation, testing, and
other higher-level activities. Husqvarna has moved to the adoption of more formalized software
development processes, and a software product line approach is being introduced to rationalize
software development across the Husqvarna product range.
A range of new services will be enabled through software in the future. For example,
Husqvarna currently incurs a high cost through products being returned as faulty under warranty,
which in many cases is due to customer misuse. Adopting an IoT approach, Husqvarna will
interrogate built-in sensors in their products deployed in the field worldwide to recover diagnostic
and statistical information on the functioning of their equipment. This will enable Husqvarna to
proactively advise customers as to whether the equipment needs servicing, and whether it is being
used properly and efficiently. If not, a message will be sent to the customer advising them that the
product is likely to malfunction, which will require it to be returned to Husqvarna and thus not
available to the customer for some period of time.
Also, Husqvarna robotic lawn mowers will be able to monitor weather forecasts to ensure grass
is cut before imminent rainfall, and also will remember previous grass-cutting routes to ensure that
new routes will be chosen each time. All these innovations are enabled by software. 


There are several key lessons in the vignette. First, as Husqvarna was moving from gas-driven to
electrically powered machinery, they found themselves in a new business domain that was already
inhabited by other companies producing similar devices, and hence were faced with new competitors.
In order to introduce new products and services, new approaches such as software product lines and a
formalized development process were necessary. Furthermore, the need to deliver more software
required an organizational expansion as a result of the hiring of trained software developers.
The final vignette relates to the Scaling Organizations dimension, and we draw on the experiences of
Sony Mobile.

Scaling organizations: the case of Sony Mobile

Sony Mobile is an example of a large company in the mobile phone domain, a highly dynamic and
competitive market with a number of significant competitors looming. Staying innovative is key
and delivering new features is of the utmost importance. This in turn results in a higher demand
for a large pool of developers that can deliver those features quickly. To overcome this
organizational barrier to scale their software development capacity, Sony Mobile is expanding the
development organization in two ways. First Sony Mobile is actively participating and contributing
in open source communities. Sony Mobile’s phones are based on the Android platform and more
than 85% of their software is based on open source. Second, Sony Mobile is adopting Inner Source
(adopting open source practices within organizational boundaries). Inner Source facilitates
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

developers across the organization to collaborate on common projects without formal team
membership. Engagement with OSS projects and adopting Inner Source has implications for the
product and process dimensions as well. Teams interacting with communities of ‘unknown’
developers must comply with community norms and common practices so as to ensure that their
contributions can be integrated.

Similar to the companies in the other vignettes, Sony Mobile has also been facing the challenge to deliver
more software and more quickly. In their domain, they chose to adopt the Android platform for their
mobile phones, and consequently were facing the need to collaborate with other stakeholders outside
their organization – in this case, open source communities. At the same time, the company is also
adopting inner source: managing projects that involve such external and internal communities requires
considering new and unknown forces (Höst et al. 2014). Many other large organizations are actively
involved in open source communities, including Hewlett-Packard and Wipro. These organizations must
learn how to interact with these external communities of developers that may have a different agenda
and motivations. To that end, many of these companies are now employing dedicated Open Source
Community Experts.
In each of the three vignettes described earlier, the primary focus was on scaling in one dimension,
but the companies involved ultimately had to deal with changes in all three dimensions. For example,
QUMAS started their scaling transformation in the process dimension, but consequentially had to make
changes to their organization and product dimensions. Likewise, Husqvarna is scaling primarily in the
product dimension, leading to subsequent changes to their processes and organization. Sony Mobile
focused primarily on expanding their organization by leveraging open source and inner source
communities, which in turn led to changes in the product architecture (e.g., dependency on the open
source Android platform) and process changes.

Implications for research

The three industry vignettes demonstrate how different organizations are scaling in different ways. What
they have in common, however, is that no matter which direction they scale in initially, they will
eventually have to scale along all three dimensions. This interdependency across these three dimensions
has thus far received limited attention; studies tend to focus on one dimension exclusively. For example,
cloud computing, software product lines (SPL) and service-oriented architectures (SOA) are all means
to scale software products. While the SPL community has also recognized the impact on process and
organization, for example through the BAPO framework (van der Linden et al. 2007), there has not
been a systematic approach to considering all three interlinked dimensions for all the contemporary
developments that we can identify in the modern software research or industry landscapes.
We believe future research on software development should include considerations across all three
key dimensions discussed earlier. In Table 9.1, we summarize exemplar scaling scenarios for each
dimension, together with the implications for other dimensions, and also some relevant implications for
practice, research and education.
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK Stein (Eds.) The Routledge Companion to Management
Information Systems. Routledge. Pp. 125-137
Table 9.1 Exemplar scaling scenarios in each dimension and their implications

Scaling Dimension Exemplar Scaling Scenario Implications for Other Dimensions Implications for Practice, Research and
Education
Products, Systems Organizations are changing their Companies need to scale on organizational aspects, e.g., hire trained software The range of software development
and Services product offerings and moving to engineers instead of relying on hardware engineers, and adopt an appropriate contexts will become much more
‘softwareization’ and software development process that suits the product development context. varied, which increases the importance
‘servitization.’ of context in research and education of
software engineers.
Organizations and Organizations are becoming Trends such as open sourcing, outsourcing and crowdsourcing require Organizations must carefully consider
Business Domains increasingly dependent on changes to an organization’s internal development processes. Outsourcing the implications of their sourcing and
external suppliers and workforces, organizations should also consider the product architecture in coordinating partnership strategies. Software may
and offer participation to third- external workforces and integrating externally developed software. Keystone take an organization to new business
party suppliers in ecosystems. players in ecosystems offering platforms to third parties to develop plug-ins domains that are already inhabited by
or apps must consider constraints of the product architecture. others, thus facing new competition.
Processes and Organizations are adopting Organizations need to consider other stakeholders in the organization Software transcends the software
Methods contemporary and emerging affected by a changing process; e.g., to adopt DevOps and continuous development function, which must be
software development practices delivery, teams responsible for operations must be involved. Top-level linked to the whole organization – the
and techniques, including agile management support is required to get different departments involved. If software enterprise. Software
methods, DevOps, continuous applicable, constraints due to regulatory compliance must be considered. engineering education must address
deployment, and continuous Product architectures must be amenable to a continuous delivery approach. the interactions with other functions,
software engineering. including marketing and sales.
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

Conclusion
The domain of software development is expanding and reaching far beyond mere technical solutions that
involve algorithms, software architectures, and ultra-large systems. The ability to deliver this increasing
amount of software is an immediate and pressing concern for many organizations – this has been termed
Software Crisis 2.0 (Fitzgerald 2012). Software development must take a holistic view and consider the
various interdependencies between trends such as softwareization and servitization, the organizational
aspects that are affected by an increasing ‘pull’ for software-based solutions, and the processes and methods
that are used to deliver this software.
These trends will have far-reaching consequences. First, the tension that arises due to the push and
pull factors discussed earlier may increase the pressure to deliver software more quickly, which inevitably
will lead to compromises in the quality of the software produced. Software that is constructed quickly,
without proper design, review, and quality assurance is far more likely to exhibit defects down the road,
thus increasing the cost of maintenance, which in turn exacerbates Software Crisis 2.0.
Second, the pull factors also imply an increasing need for talented software developers. As software
systems are becoming increasingly complex, the reliance on self-educated and software hobbyists is no
longer sufficient; instead, companies need well-trained staff. The lack of well-trained staff is an oft-cited
reason for companies to outsource their software development.
Another challenge that is perhaps not receiving enough attention is our society’s ever-increasing
reliance on software systems. Software systems are truly ubiquitous, but many of these systems are very
dated legacy systems. Many critical systems (e.g., operating at banks, insurance companies, and
government institutions) were developed in the sixties and seventies on technology platforms that are
now considered dated, such as COBOL. While developers who are able to maintain these systems are
currently still available, this may no longer be the case 20 or 30 years from now. An incredible amount
of software is being written today on a variety of technologies that rapidly become obsolete and will
inevitable have to be replaced at some point. As systems are becoming ever more intertwined and
interdependent, this will become increasingly problematic.
With the realization that software development transcends algorithms and techniques, we believe
research should include considerations across all three key dimensions: products, services, and systems;
processes and methods; and organizations and business domains. An increased understanding of the
different types of scaling challenges that organizations face will help the software industry to overcome
them, and will help in deriving new software development methods better suited to the needs of the
prevailing software development environment.

Acknowledgements
This work was supported, in part, by Enterprise Ireland, grant IR/2013/0021 to ITEA-SCALARE,
Science Foundation Ireland grant 15/SIRG/3293 and 13/RC/2094 and co-funded under the European
Regional Development Fund through the Southern & Eastern Regional Operational Programme to Lero
– the Irish Software Research Centre (www.lero.ie).

References
Ågerfalk, P. J., B. Fitzgerald and K. Stol (2015a) Software Sourcing in the Age of Open: Leveraging the Unknown
Workforce, Springer.
Ågerfalk, P. J., B. Fitzgerald and K. Stol (2015b) Not so shore anymore: The new imperatives when sourcing in the
age of open, Proceedings of the European Conference on Information Systems (ECIS) Münster, Germany.
Boehm, B. (2002) “Get ready for agile methods, with care,” IEEE Computer, vol. 35, pp. 64–69;
doi:10.1109/2.976920.
B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

Booch, G. (2015) Keynote at the International Conference on Software Engineering, SEIP Track, Florence, Italy.
Bosch, J. and P. Bosch-Sijtsema, “From integration to composition: On the impact of software product lines, global
development and ecosystems,” Journal of Systems and Software, vol. 83, no. 1, 2010, pp. 67–76;
doi:10.1016/j.jss.2009.06.051.
Cao, K., P. Mohan, P. Xu and B. Ramesh, (2004) “How extreme does extreme programming have to be? Adapting
XP practices to large-scale projects,” in 37th Hawaii International Conference on System Science.
Conboy, K., S. Coyle, X. Wang and M. Pikkarainen, “People over process: Key challenges in agile development,”
IEEE Software, vol. 28, no. 4, 2011, pp. 48–57.
Dybå, T., D.I.K. Sjøberg and D. S. Cruzes, “What works for whom, where, when, and why? On the role of Context
in empirical software engineering,” Proceedings of Empirical Software Engineering and Measurement , 2012,
pp. 19–28.
Feki, M. A., F. Kawsar, M. Boussard and L. Trappeniers, “The internet of things: The next technological revolution,”
IEEE Computer, vol. 46, no. 2, 2013, pp. 24–25.
Feller, J., B. Fitzgerald, S. Hissam, and K. Lakhani. (2005) Perspectives on Free and Open Source Software. MIT
Press. Cambridge, MA.
Fitzgerald, B. “Systems development methodologies: the problem of tenses,” Information Technology & People, vol.
13, no. 3, 2000, pp. 174–185.
Fitzgerald, B., “Software Crisis 2.0,” IEEE Computer, vol. 45, no. 4, 2012, pp. 89–91.
Fitzgerald, B., M. Musial and K. Stol, “Evidence-based decision making in lean software project management,” in
Proceedings of International Conference on Software Engineering , vol. 2, Hyderabad, India, 2014.
Fitzgerald, B. and K. Stol, “Continuous software engineering and beyond: A roadmap and agenda,” Journal of Systems
and Software, vol. 123, 2017, pp. 176-189.
Fitzgerald, B., K. Stol, R. O’Sullivan and D. O’Brien, “Scaling agile methods to regulated environments: An industry
case study,” Proceedings of International Conference on Software Engineering , vol. 2, 2013, pp. 863–872.
Friedman, A. (1989) Computer Systems Development: History, Organisation and Implementation , Wiley & Sons,
Chichester.
Höst, M., K. Stol and A. Orucevic-Alagic (2014) “Inner source project management,” in: G. Ruhe and C. Wohlin
(Eds.) Software Project Management in a Changing World , Springer, pp. 343-367.
Kahkonen, T. (2004) “Agile methods for large organisations – building communities of practice,” in Agile
Development Conference.
McHugh, M., F. McCaffery and V. Casey, “Standalone software as an active medical device,” Proceedings of 11th
International SPICE Conference, 2011, pp. 97–107.
Mössinger, J., “Software in automotive systems,” IEEE Software, vol. 27, no. 2, 2010, pp. 92–94.
Müller, H., K. Wong, and S. Tilley, “Understanding software systems using reverse engineering technology,”
Proceedings of the 62nd Congress of L’Association Canadienne Française pour l’Avancement des Sciences
(ACFAS), vol. 26, no. 4, 1994, pp. 41–48.
Poppendieck, M. and M. Cusumano, “Lean software development: A tutorial,” IEEE Software, vol. 29, no. 5, 2012,
pp. 26–32.
Rolland, K., B. Fitzgerald, T. Dingsøyr and K. Stol, “Problematizing agile in the large: Alternative assumptions for
large-scale agile development,” Proceedings 37th International Conference on Information Systems , Dublin,
Ireland, 2016.
Schneider, J. “Software-innovations as key driver for a green, connected and autonomous mobility,” ARTEMIS-
IA/ITEA-Co-Summit, 2015.Stol, K. and B. Fitzgerald, “Inner source – adopting open source development
practices in organizations: A tutorial,” IEEE Software, vol. 32, no. 4, 2015.
Stol, K. and B. Fitzgerald (2014) “Two’s company, three’s a crowd: A case study of crowdsourcing software
development,” Proceedings of the 36th International Conference on Software Engineering (Technical Track),
Hyderabad, India
Swedsoft, A Strategic Research Agenda for the Swedish Software Intensive Industry , 2010.
Tamburri, D., P. Lago and H. van Vliet, “Uncovering latent social communities in software development,” IEEE
Software, vol. 30, no. 1, 2013, pp. 29–36.
 B Fitzgerald and K Stol (2018) The Future of Software Development Methods. RD Galliers and MK
Stein (Eds.) The Routledge Companion to Management Information Systems. Routledge. Pp. 125-137

Turk, D., R. France and B. Rumpe, “Assumptions underlying agile software-development processes,” Journal of
Database Management, vol. 16, no. 4, 2005, pp. 62–87.
van der Linden, F., K. Schmid and E. Rommes (2007) Software Product Lines in Action, Springer, New York, Inc.
Secaucus, NJ, USA.

View publication stats