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

net/publication/330353717

Becoming Agile in the Digital Transformation: The Process of a Large-Scale Agile

Transformation

Conference Paper · December 2018

CITATIONS READS

22 6,246

2 authors:

Christoph Fuchs Thomas Hess

Ludwig-Maximilians-University of Munich Ludwig-Maximilians-University of Munich
9 PUBLICATIONS   42 CITATIONS    511 PUBLICATIONS   6,002 CITATIONS   

SEE PROFILE SEE PROFILE

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

E-Commerce View project

PhD Thesis View project

All content following this page was uploaded by Thomas Hess on 13 January 2019.

The user has requested enhancement of the downloaded file.

 Becoming Agile in the Digital Transformation

Becoming Agile in the Digital

Transformation: The Process of a Large-
Scale Agile Transformation
Completed Research Paper

Christoph Fuchs Thomas Hess

LMU Munich LMU Munich
Ludwigstr. 28, 80539 Munich, Ludwigstr. 28, 80539 Munich,
Germany Germany
fuchs@bwl.lmu.de thess@bwl.lmu.de

Abstract
Confronted with the imperatives of a digital world, firms are striving to become agile,
resulting in a large-scale agile transformation as part of their organizational digital
transformation. Although initial research exists, empirical literature on the process,
challenges, and actions of a large-scale agile transformation is scarce. Consequently,
this paper conceptualizes the agile transformation process through the lens of socio-
technical systems theory and employs a qualitative research approach comprising two
in-depth case studies. As a result, a large-scale agile transformation can be interpreted
as an episodic change process that comprises a sequence of multiple agile
transformation phases. These phases include radical and incremental change and are
delimited by barriers which are formed by emerging challenges. Such barriers are
encountered by specific actions that serve scaling and coping purposes. Besides
pertinent theoretical insights, the paper offers guidance for managers that direct an
agile transformation in the wake of their firms’ digital transformation.

Keywords: Digital transformation, large-scale agile transformation, agile transformation

process, barriers of an agile transformation

Introduction
In an increasingly digital world, firms are confronted with a multitude of challenges such as volatile
customer demands, increasing market dynamics, and the continuous emergence of novel advancements
in information technology (IT) (Porter and Heppelmann 2015). As a result, firms are urged to undergo an
IT-enabled organizational business transformation (Besson and Rowe 2012) that sparks changes for the
entire firm and especially for aspects such as digital offerings, processes, and business models (Bharadwaj
et al. 2013). To master their respective organizational digital transformation, firms are striving to become
agile (Porter and Heppelmann 2015; Rogers 2016). One of the most popular options to foster the
associated organizational agility and to address relevant issues of an organizational digital transformation
(e.g., rapid and adaptive development of novel digital offerings) is to introduce agile methods (Bharadwaj
et al. 2013; Lee and Xia 2010). In contrast to plan-based approaches, which entail detailed upfront
planning and extensive documentation, agile methods, such as Scrum or eXtreme Programming,
represent iterative development approaches that embrace quick deployment, responsiveness to change,
and an emphasis on customer needs (Abrahamsson et al. 2009; Beck et al. 2001a; Beck et al. 2001b). In
the digital era, agile methods are not only increasingly applied by firms outside the IT and software
industry, but are also employed beyond the borders of their traditional application areas (i.e., IT-related
projects and software development [SD]) in fields such as new product development (Bharadwaj et al.
2013; Rigby et al. 2016). These wider application areas often facilitate an overarching implementation of

Thirty Ninth International Conference on Information Systems, San Francisco 2018 1

 Becoming Agile in the Digital Transformation

agile methods within firms, resulting in the large-scale application of agile methods (Dikert et al. 2016).
With agile methods entailing implications for work procedures, organizational structures, and cultures
(Zheng et al. 2011), such extensive agile approaches often necessitate a large-scale agile transformation
as part of the organizational digital transformation (Dikert et al. 2016; Paasivaara et al. 2018). However,
the process of a large-scale agile transformation is not trivial and entails key managerial challenges and
consequences for the entire firm. As a result, the transformation process characterized by the interplay of
occurring challenges, related coping as well as scaling actions is of relevance for research and practice.
In existing empirical literature on the digital transformation of firms, aspects of agile methods are only
referred to in terms of the changing role of the IT function (e.g., Jöhnk et al. 2017; Leonhardt et al. 2017).
Current empirical literature on the large-scale application of agile methods discusses specific topics such
as inter-team coordination in multi-team agile settings (e.g., Scheerer et al. 2014) or the challenges of
realizing a large-scale agile transformation (e.g., Dikert et al. 2016). However, only a few pertinent articles
on the process of large-scale agile transformations are available (e.g., Paasivaara et al. 2018) and a call for
empirical research on this topic has been issued (Dikert et al. 2016; Paasivaara et al. 2018). A processual
approach to study agile transformations appears particularly fruitful since such a viewpoint enables a
profound understanding of “underlying organizational change processes” (Poole et al. 2000, p. 4) and
thus fosters an in-depth grasp of the transformation itself. Therefore, we intend to answer the outlined
call for research and discuss large-scale agile transformations in the current research context of the digital
transformation. To achieve this, we investigate a large-scale agile transformation through a processual
lens by capturing the interplay of challenges, coping, and scaling actions. We derive as research questions:
RQ1: How does the process of a large-scale agile transformation in the context of the digital
transformation progress?
RQ2: How do challenges of a large-scale agile transformation as well as related coping and scaling actions
shape the transformation process?
To answer these questions, we build on literature on large-scale agile transformations as well as related
challenges and actions. We apply an explorative, qualitative-empirical approach by conducting two in-
depth case studies of firms that currently undergo a large-scale agile transformation in the course of their
digital transformation. To capture the change process, we employ socio-technical systems (STS) theory as
a research lens and identify challenges and two types of actions (i.e., scaling and coping actions) as well as
their interplay. We then compare the results in a cross-case analysis and develop an abstract explanation
for a large-scale agile transformation from a processual viewpoint. We offer clarification on the nature of
challenges and actions in such a transformation as well as their role in shaping the agile transformation
process. In this way, we contribute to literature on large-scale agile transformations in the context of the
digital transformation and provide guidance for firms to manage an agile transformation process.
Our paper is structured as follows: We begin with a discussion of relevant literature on large-scale agile
transformations, related challenges and actions. We then propose our conceptual research lens building
on STS theory. Next, we portray our qualitative case study approach, the selection as well as compilation
of our sample and our data analysis approach. Afterwards, we present the results of the case studies and
discuss these results as part of a cross-case analysis. We close our study with theoretical and practical
implications as well as a conclusion including limitations and suggestions for areas of future research.

Literature Background
Large-Scale Agile Transformation
Research on the large-scale application of agile methods often entails a discussion about what “agile in the
large” (Rolland et al. 2016, p. 2) means and how large-scale agile development can be conceptualized
(Dingsøyr and Moe 2014). The large-scale application of agile methods has multiple interpretations: a)
the use of agile methods in large firms, b) the application of agile methods in large projects or large teams,
c) the usage of agile methods in large multi-team settings, and d) the employment of agile practices and
principles in firms as a whole (Dingsøyr and Moe 2014). For this study, we focus on the last two options.
In terms of definitions, we understand the large-scale application of agile methods on an organizational
level with multi-team settings that consist of “50 or more people or at least six teams” (Dikert et al. 2016,
p. 88). The second part in the term large-scale agile transformation, namely transformation, refers to the

Thirty Ninth International Conference on Information Systems, San Francisco 2018 2

 Becoming Agile in the Digital Transformation

switch from a different development approach or work organization concept to agile methods. A large-
scale agile transformation can cover a one-time big bang transfer to agile methods in a large setting (e.g.,
switch in a firm’s entire SD unit) or a stepwise approach where an agile pilot is subsequently scaled up
into a large setting. Consequently, we understand the process of scaling up as the extension of an initial
adoption of agile methods. This scaling up can take the form of more organizational members employing
agile methods (e.g., hiring employees to extend the SD), extending the application of agile methods within
the firm (e.g., transforming further business units) or deepening the application of agile methods (e.g.,
integrating further agile practices from various agile methods) (Dikert et al. 2016; Paasivaara et al. 2018).
Empirical literature on large-scale agile transformations is mainly characterized by in-depth single case
studies which, for instance, examine inter-team coordination approaches (e.g., Scheerer et al. 2014) or
portray specific actions (e.g., communities of practice) supporting the transformation (e.g., Paasivaara
and Lassenius 2014). Given that the work of Paasivaara et al. (2018) represents one of the first studies to
depict a large-scale agile transformation process of a firm from an empirically founded perspective,
further research in this area is called for (Dikert et al. 2016; Paasivaara et al. 2018).
In addition to such academic literature, practitioners and consultants offer frameworks for large-scale
agile approaches in SD, such as the Scaled Agile Framework (SAFe) (Scaled Agile 2017), Large Scale
Scrum (LeSS) (Vodde and Larman 2014) and Disciplined Agile Delivery (DAD) (Ambler and Lines 2012).
Although these frameworks provide pertinent insights on how large-scale agile approaches manifest
themselves, they cannot explain the process of a large-scale agile transformation (Paasivaara et al. 2018).
Overall, an initial empirical basis for research on large-scale agile transformations exists. However, more
research is required to extend the understanding of the process and the inherent interplay of challenges
and actions of such a transformation, specifically in the novel context of the digital transformation.

Challenges and Actions of Large-Scale Agile Transformations

Challenges are a common topic in agile methods’ research also besides the specific field of large-scale agile
transformations. Cao et al. (2009), for instance, investigate the adaptation of agile methods and thereby
build on development process-related, customer-related, developer-related and organization-related
challenges to better comprehend firms’ motivations to customize agile methods. Additionally, early works
such as Nerur et al. (2005) report on the challenges of implementing agile methods in organizations’ SD
departments, considering issues such as the selection of an appropriate agile method or the problem of
lacking developers’ competences. Nonetheless, also current empirical studies provide pertinent insights
on this topic. For instance, Hekkala et al. (2017) examine an information systems (IS) development team
that transitioned to agile methods, revealing, amongst others, the challenges of misunderstanding agile
practices and unsuitable organizational structures for agile principles and values being in place. However,
large-scale agile transformations entail unique challenges that go beyond the issues of an initial agile
transition or the introduction of agile methods in small settings such as individual teams. In large-scale
agile settings, additional challenges such as coordination issues between agile teams as well as between
agile and non-agile teams emerge (Scheerer et al. 2014; Zheng et al. 2011).
Overall, an extensive body of literature comprising empirical investigations on (e.g., Hekkala et al. 2017)
and comprehensive summaries of (e.g., Dikert et al. 2016; Gregory et al. 2015) challenges of the (large-
scale) application of agile methods exists. However, we do not aim to validate, revise or extend this body
of knowledge. Instead, we build on the comprehensive findings of the existing works and aim to capture
relevant challenges of large-scale agile transformations. We intend to carve out more in-depth insights on
their nature as well as their role in shaping the agile transformation process. We specifically focus on the
interplay of such challenges with actions coping with these issues and actions scaling the overall agile
transformation. Considering the nature of these challenges, the question remains whether issues exist that
completely block the transformation process or lead to its termination. Challenges are mainly seen as
hindering factors that require mitigation, whereas a perspective including a potentially manifold nature of
agile transformation challenges is not yet applied (Dikert et al. 2016; Paasivaara et al. 2018). We propose
an abstract categorization of large-scale agile transformation challenges to which commonly examined
issues can be assigned. While we provide a general description of these categories, the empirical
validation and exact depiction of individual challenges is provided elsewhere in the literature (e.g., Dikert
et al. 2016; Gregory et al. 2015; Hekkala et al. 2017). To arrive at these categories, we reviewed diverse
studies: a) those with a focus on the large-scale application of agile methods and on large-scale agile

Thirty Ninth International Conference on Information Systems, San Francisco 2018 3

 Becoming Agile in the Digital Transformation

transformations (e.g., Dikert et al. 2016; Paasivaara et al. 2018; Rolland et al. 2016), b) those that address
general challenges of agile methods’ adoption without an emphasis on large settings (e.g., Boehm and
Turner 2005; Gregory et al. 2015; Hekkala et al. 2017; Nerur et al. 2005), and c) those discussing
challenges with specific focuses such as people-related issues (e.g., Conboy et al. 2011). By building on
existing categorizations (e.g., Dikert et al. 2016; Nerur et al. 2005) and revising the groupings on the basis
of the allocation of the challenges, we developed the categorization proposed in Table 1 which includes a
general explanation of the categories and pertinent examples of individual challenges. This categorization
provides us with an overview of challenges that can occur in a large-scale agile transformation.
Category Explanation: Challenges regarding… Examples of Challenges
…the appropriate application of agile  Misunderstanding of agile methods
Method-
methods and the respective employment  Poor customization of agile methods
related
areas within organizations.  Inappropriate application area of agile methods
…the infrastructural features of firms and
Technology-  Inappropriate technological equipment
the supporting structures of technological
related
tools within firms.  Inappropriate IT infrastructure

…the organizational structures, occurring  Problematic coordination with other business units
Organization
coordination issues and organizations’  Inappropriate organizational structures
-related
overall management.  Lack of top management engagement
Culture- …the social and overall cultural aspects of  Inappropriate leadership dynamics
related organizations.  Incompatible social structures
Ability- …the abilities of organizational members  Lack of hard skills
related involved in the agile transformation.  Lack of knowledge transfer
…the attitudes about and opinions on the
Motivation-  Missing agile mindset
transformation of organizational members
related
involved in the agile transformation.  Fear of consequences

Table 1. Proposed Categorization of Challenges of a Large-Scale Agile Transformation

Referring to the processual viewpoint of our study, it is crucial to capture the actions of an organizational
change process to gain a holistically understanding of it (Poole et al. 2000). By building on pertinent
empirical literature on large-scale agile transformations (e.g., Dikert et al. 2016; Paasivaara et al. 2012;
Paasivaara et al. 2018), we derive two types of actions that are undertaken in an agile transformation. We
differentiate them according to the purpose that is to be achieved by performing them in the course of a
large-scale agile transformation process. As a result, we identify coping actions which depict actions with
the purpose to mitigate emerging challenges of the change process (Paasivaara et al. 2018). Communities
of practice, which represent expert panels that meet to deepen their knowledge on one common topic, are
one example of coping actions. We classify communities of practice as a coping action since their
implementation is often undertaken to tackle issues such as a lacking knowledge transfer between agile
team members in the course of an agile transformation (Paasivaara and Lassenius 2014; Paasivaara et al.
2018). As a second type of relevant actions for a large-scale agile transformation, we derive scaling
actions. According to our aforementioned understanding of “scaling up”, the purpose of these scaling
actions is to advance a large-scale agile transformation, for instance, by including more people in the
change process, extending the agile approach by introducing additional agile practices, or implementing
specific organizational structures to enable further organizational change. One pertinent example of
scaling actions is the involvement of additional organizational members in the large-scale agile
transformation process, for instance, as new agile team members (e.g., by hiring new employees or
transferring existing organizational member into agile teams) (Dikert et al. 2016; Paasivaara et al. 2012).

Conceptual Research Lens: Socio-Technical Systems Theory

We aim to better comprehend a large-scale agile transformation as part of the organizational digital
transformation from a processual view by observing pertinent challenges and actions of the organizational
change as well as their interplay. Therefore, we select STS theory as our underlying research lens, because
it allows for an holistic perspective on transforming firms and acknowledges that organizational change

Thirty Ninth International Conference on Information Systems, San Francisco 2018 4

 Becoming Agile in the Digital Transformation

(i.e., a large-scale agile transformation) can be viewed as an interplay of organizational structures, work
procedures, technologies and organizational members (Alter 2013; Sarker et al. 2013). The socio-technical
lens also enables us to apply a processual approach towards agile transformations (Poole et al. 2000;
Sarker et al. 2013). However, we do not aim to make explicit contributions to STS theory but employ it as
a lens that guides our overall empirical research approach as well as the interpretation of our findings.
In the socio-technical approach firms consist of two subsystems: a social subsystem that comprises “the
individuals and the knowledge, skills, attitudes, values, and needs they bring to the work environment, as
well as the reward system and authority structures that exist in the organization” (Bostrom et al. 2009, p.
18) and a technical subsystem that includes “the tools, mechanisms, and techniques used within the social
subsystem to carry out organizational work” (Ryan et al. 2002, p. 89). Since these two subsystems are
inseparably connected, harmony between them is optimal for organizational outcomes (e.g., productivity)
and social aspects (e.g., employees’ work satisfaction) (Mumford 2003; Sarker et al. 2013). The socio-
technical approach has sparked manifold applications in IS literature, for instance in the field of IT-
induced change (e.g., Lyytinen and Newman 2008). Through the socio-technical research lens we
understand a firm as an organizational work system, defined as “a system in which human participants
and/or machines perform work (processes and activities) using information, technology, and other
resources to produce specific products/services for specific internal and/or external customers” (Alter
2013, p. 75). We assume reciprocal interactions between the social subsystem – consisting of structure
(i.e., organizational structures, hierarchies, and authority system) and people (i.e., employees of the firm,
their skills, knowledge, attitudes, and values) – and the technical subsystem – encompassing technologies
(i.e., IS, software, hardware, or machines) and business processes (i.e., techniques employed in the task
area) (Bostrom et al. 2009; Lyytinen and Newman 2008; Ryan et al. 2002; Sarker et al. 2013).
To examine the process of a large-scale agile transformation, we capture the actions and challenges of the
transformation by means of these four dimensions. To this end, we allocate observed challenges and
actions to the appropriate socio-technical dimensions and thus retrace the dimensions that are addressed
in the change process. Based on this mapping of actions and challenges to the abstract dimensions of STS
theory, we are able to capture their interplay as well as the more abstract co-evolution of the social and
technical subsystem. Therefore, we can portray the transformation process within the organizational work
system. This enables us to make more generalized statements about an organizational large-scale agile
transformation and its processual flow (Orlikowski 1992; Sarker et al. 2013). The starting point of a large-
scale agile transformation lies within the process dimension since such a change process is typically
triggered by the introduction of agile methods (i.e., techniques of the organizational work system).

Research Method
Case Study Research
We employ a positivist, multiple-case study research design. We postulate, given our positivist stance
towards epistemology, that a physical world with fixed social relationships exists independently from the
researcher that can be analyzed objectively (Myers 2009; Paré 2004). Also, we follow an exploratory
approach with the intent to define “questions, constructs, propositions, or hypotheses to be the object of a
subsequent empirical study” (Paré 2004, p. 235). We select qualitative case studies, since they are best
suited to answer how and why questions and to explore a current phenomenon within its real life context
(Yin 2013). In accordance with the requirements of Benbasat et al. (1987) our research objective (to study
the process of a large-scale agile transformation in the context of the digital transformation) a) cannot be
studied outside its natural setting, b) represents a contemporary development, c) requires no control or
manipulation of research subjects, and d) does not have the advantage of an established theoretical basis.
Overall, we closely follow common guidelines and recommendations from literature (e.g., Dubé and Paré
2003; Paré 2004) to foster the rigor of our empirical approach and control for quality criteria of positivist
case studies such as construct validity, internal validity, external validity, and reliability (Yin 2013).

Case Selection and Sample Description

We apply a multiple-case study research approach, because it allows for cross-case analysis and is less
vulnerable to critique concerning the generalizability of results (Yin 2013). Following Patton (2002), we
purposefully chose firms in a two-step process. First, we followed a criterion sampling logic and selected

Thirty Ninth International Conference on Information Systems, San Francisco 2018 5

 Becoming Agile in the Digital Transformation

firms outside the IT and software industry that undergo a large-scale agile transformation as part of their
digital transformation. Second, we employed a theoretical replication logic to generate contrasting results
by choosing heterogeneous cases, thus enhancing the study’s external validity (Yin 2013). We selected two
firms that differ in size, industry, and approach to a large-scale agile transformation. With sample sizes of
case study research often being a focal point of critique, we captured two case studies on suitable firms in-
depth to generate insights on large-scale agile transformations and to be able to offer meaningful answers
to our research questions (Paré 2004). We added information in form of interviews and secondary data to
the cases until we were no longer receiving relevant, new information on the change process (Yin 2013).
Since firm 1 is significantly larger, thus entailing a more extensive change process, we conducted twice the
number of interviews for case 1 than for case 2. All our interviewees had detailed knowledge of the firms’
present and past agile transformation activities enabling us to reconstruct both transformation processes.
We employed a case study protocol, including all relevant data (e.g., project overview, interview guide,
rules for field procedures) to ensure the reliability of our research (Paré 2004; Yin 2013). An overview of
the sample is shown in Table 2. To ensure confidentiality, we call the firms InsureTech and EventCom.

Case 1: InsureTech Case 2: EventCom

Industry Insurance and finance Retail and events
Revenue (2017) approx. 32 billion € approx. 75 million €
Employees (2017) approx. 29,000 >550
Number/Dates of Interviews 11 / April 2017 – January 2018 5 / May 2017 – December 2017
Interviewees by Position  Chief Information Officer (CIO)  Manager of agile transformation*
(Numbers represent the amount  Manager of agile transformation (2x) †  Product owner (2x)
of interviewees in this position)
 Agile coach (2x) †  Scrum Master (1x)
*Interviewees work outside the IT  Senior project manager (2x) †  Agile team member (1x)
or SD department.  Product owner (1x)
† One of two interviewees works  Scrum Master (1x)
outside the IT or SD department.  Agile team member (2x)

Table 2. Overview of Sample and Interviewees

Data Collection and Analysis

Data collection took place April 2017 – January 2018 and data on both firms was collected continuously
for at least half a year mainly by means of semi-structured interviews. This time span was needed for us to
be able to capture the large-scale agile transformations of the firms and include current developments in
this regard. The authors used their professional networks to arrange the interviews. All but one interview
were conducted face-to-face, on-site at the firms (one via telephone) and lasted 45 – 90 minutes. A semi-
structured interview guide with open-ended questions was used. The guide included four sections on: a)
the organizational digital transformation, b) the implementation and scaling history of agile methods used
in the firms, c) the challenges and coping actions of the firms in the course of the agile transformation and
d) the adaptations made to the organizational work system of the firms. Selected interviews were carried
out by two researches of which one was a senior researcher. All interviews were tape-recorded and
transcribed verbatim. We also prepared field notes following our on-site visits as well as memos based on
informal talks with organizational members. We also had access to selected internal documents such as
presentations relevant for the agile transformations. In case of InsureTech, one author also participated in
a two-day, on-site agile workshop. Besides this internal data, we also performed data triangulation and
included external secondary data in our analysis (e.g., press releases, firm websites, and management
reports) to enhance the rigor and construct validity of our study. Finally, we employed the software
ATLAS.ti to store, code, and analyze all gathered data (Miles et al. 2013; Paré 2004; Yin 2013).
We performed two cycles of coding. The first cycle consisted of descriptive coding to build an initial
understanding of the cases. This stage included codes deduced from our literature background (e.g.,
challenges) and emerging codes which were inductively added to the coding scheme (see Table 3). In the
second round of coding, we revised and refined our initial codes and searched for patterns in the codes to
cluster our data. This clustering was guided by the four dimensions of the STS theory. Considering the

Thirty Ninth International Conference on Information Systems, San Francisco 2018 6

 Becoming Agile in the Digital Transformation

challenges of a large-scale agile transformation, we initially coded the occurring issues descriptively in
reference to the literature. In addition, we allocated these challenges to the six abstract categories that we
developed in our literature background and noted the category in the descriptive codes. In the second
cycle of coding, we revised and refined these codes and abstracted the challenges to the four dimensions
of the STS theory. Consequently, method-related challenges refer to the process dimension, technology-
related issues to the technology dimensions, organization- and culture-related problems to the structure
dimension and ability- and motivation-related challenges to the people dimension. In terms of the actions
performed, we initially coded them descriptively and noted whether we classify them as scaling or coping
actions. In the second coding cycle, we allocated them to the dimensions of the STS theory. By abstracting
the codes and mapping them to our socio-technical research lens, we were better able to capture the
interplay of the two types of actions and the challenges, thus revealing the mutual adaptations to as well
as the co-evolution of the organizational work system undergoing a large-scale agile transformation. The
coding was conducted by two researchers to ensure data analysis quality. Each coder performed the first,
descriptive cycle of coding independently, and then a consensual approach was used for the second cycle
of coding. Significant differences in refining the codes, clustering patterns and mapping to the socio-
technical dimensions were discussed bilaterally and consensually resolved (Miles et al. 2013).

Revised, Descriptive Code Code Cluster

Exemplary Interview Statement
(First Cycle) (Second Cycle)
“There was the [top management commitment] to develop an
Action: Hiring new agile Scaling actions:
app and in 2015 an entire team of [another firm] switched to
employees (scaling) People
EventCom and this is how we acquired a further team.”
“[…] to address our coordination issues [between the agile
teams and the business units] we now thankfully work with Action: New alignment approach Coping actions:
[the strategic goal alignment tool] which really helps us to (coping) Structure
reach general alignment.”
“Especially the middle management is having a hard time Challenge: Fear of consequences Challenges:
since they cannot identify their role in the new [agile] setting.” (motivation-related) People

Table 3. Illustration of the Coding Scheme

We present our case study results in two stages. First, we describe the two case studies individually (i.e.,
within-case results) using narratives, because they provide an accessible and easily comprehendible
vehicle for our results (Miles et al. 2013; Myers 2009). A narrative is a story written by researchers that
represents “an edited reconstruction of events or aspects that are seen as relevant to the subject or theory
being discussed” (Myers 2009, p. 212). Narratives are useful for exploratory case studies with a positivist
stance towards epistemology (Myers 2009) and also represent an appropriate medium to convey our
processual perspective to the reader (Pentland 1999). The structure of our narratives emerged in the
course of the two cycle coding process. We discovered that the large-scale agile transformation process of
both firms could be depicted in four phases (i.e., before agile, 1. agile phase, 2. agile phase and 3. agile
phase) which both firms passed through. Consequently, we present a chronological narrative with a
consistent structure for both cases. Subsequent to the single case narratives, we compare and discuss our
findings by means of a cross-case analysis and derive common insights into the process of a large-scale
agile transformation and the role of challenges and both types of actions in shaping this process.

Results
Within-Case Results
Case 1: InsureTech

Firm Context and Before Agile Phase: InsureTech is the national division of a large, multi-national
insurance and finance group. It has a longstanding firm history and is currently undergoing a large-scale
agile transformation as part of its organizational digital transformation. The firm is built on a matrix
structure and a complementary project organization. Prior to its agile transformation, IT-related projects
were completed using a waterfall approach and involved the demanding business unit, the IT unit and a
specific organizing unit that coordinated the projects and defined the projects’ technical specifications.

Thirty Ninth International Conference on Information Systems, San Francisco 2018 7

 Becoming Agile in the Digital Transformation

1. Agile Phase: In 2007, a first, bottom-up agile pilot in the IT department of InsureTech was initiated.
Initially, the pilot was well accepted in the IT department and was successfully scaled up to 60 agile teams
by 2011. The agile approach was developed to a holistic concept including practices from Scrum (e.g.,
team roles and meetings) and eXtreme Programming (e.g., pair programming), combined with insights
from change management (e.g., implementing an agile transition team). A training program with agile
coaches was available for IT teams and thus the initiative “was able to train teams, build teams and
accompany new teams during their first sprints” (agile coach_1). In the beginning, the bottom-up
initiative had management backing since IT projects were under pressure owing to the trend of IT
outsourcing. However, this backing crumbled with the introduction of new middle managers and the
emerging perception of the agile transformation approach as being “too dogmatic and too sophisticated
resulting in too big of a change” (manager of agile transformation_1). Also, the overall backend legacy
system has always been monolithic and hard to integrate into the short-cycled, iterative workflows of agile
methods. In addition, the initiative was never able to scale outside the IT department, resulting in firm-
wide skepticism of the initiative. This aspect is especially crucial since only overarching, structural change
could have prevented the agile teams from “[getting] grinded between several projects, whereas the agile
[project] was only one of them” (agile coach_1). Consequently, the combination of these challenges
evolved into an insuperable problem for the agile transformation culminating in its termination in 2013.
2. Agile Phase: With changes in the top management (TM) of InsureTech and the increasing importance
of the organizational digital transformation, the topic of agile methods was again placed on the agenda
around 2015. However, owing to the TM’s perception that the previous bottom-up agile initiative “left
scorched earth” (manager of agile transformation_1), an entirely fresh implementation approach for agile
methods was chosen. Therefore, a top-down, large-scale digital change process was initiated, including
the large-scale agile transformation of InsureTech as one focal point. As a result, in early 2016 an agile
boot camp was launched. The boot camp was, and still is, in a separate location where IT teams are
trained for about one year in “agile working, but also […] to learn to work cross-functional and enable
self-initiative [to become] autonomous teams” (manager of agile transformation_2). A holistic approach
to the agile transformation was developed which not only enabled co-location (e.g., separate rooms) and
provided the newest technological equipment (e.g., hardware and software), but also offered an adapted
agile framework outlining team roles and practices as well as integrating aspects of “lean start-up,
minimum viable products, and 100 day cycles [of funding]” (CIO). The camp was launched exclusively
for front-end SD teams, in response to the yet unsolved issues of the incompatible IT infrastructure on the
back-end side. The initial approach started with one agile team (ten members), whereas this number was
subsequently upped and more teams were invited, based on a top-down decision considering the
suitability of the projects for an agile way of working. In 2017, 200 agile team members were part of the
agile boot camp working in various teams of up to twelve people on diverse products. Multiple teams were
responsible for one product coordinated within or even across the structures of the boot camp. Novel
technological achievements (e.g., cloud services) were integrated in the development stack. However,
several issues emerged across time, for instance, the approach was initially promoted in terms of “only the
best can go there” (agile coach_2). This led to the perception of a two-tier society within the firm sparking
envy and a shared fear of consequences on the part of all IT members left out of the agile transformation
process. Also, the social structures of the organization became complicated with team members and entire
teams being relocated to a shielded site hindering communication with colleagues and further team
members which were not included in the agile boot camp. The agile boot camp also increased firm-wide
interest in agile methods. However, outside of the agile boot camp “only two people [had] the role [of an
agile coach] in the entire firm” (senior project manager_1). Projects outside of the boot camp aiming to
work agile had to implement agile methods on their own, reach out to one of the agile coaches or bring in
external consultants. Consequently, besides the top-down agile boot camp, bottom-up agile initiatives
existed and still exist that require synchronization and management. Finally, there was no exit strategy for
the agile teams in the boot camp. Originally, the teams were to stay in the training facility for a year and
then return to their original organizational context, however, as agile coach_2 stated: “we recognized that
we cannot transform the organization outside [of the boot camp] fast enough to bring the teams back.”
3. Agile Phase: To face these accumulated challenges, in 2017, InsureTech founded an agile hub within its
core structures to enable the return of the trained agile teams to their traditional working environment
and thus allow further teams to enter the boot camp. The returning teams kept the shielded status of the
boot camp but at the same time introduced the agile mindset into the traditional structures by “showing

Thirty Ninth International Conference on Information Systems, San Francisco 2018 8

 Becoming Agile in the Digital Transformation

what [they] do” (agile team member_1). With this new structure for the large-scale agile transformation
some of the challenges were addressed, such as the flow jam at the agile boot camp and the issues related
to social structures within the firm. Nonetheless, the two-tier society issue remained. Therefore, the agile
transformation approach was further scaled by implementing an agile transformation team that offers
coaching support outside the agile boot camp as well as beyond the IT and SD contexts. However, further
challenges erupted such as the issue of leadership dynamics. Middle managers were typically not included
in the agile teams trained in the boot camp. Consequently, these managers lost parts of their staff and
thus control and power. With the institutionalization of the agile hub, this loss became permanent and
some of the team leaders “found it very difficult to let their people go and let them actually work 100% in
agile teams” (product owner). Additionally, owing to the progress of the large-scale agile transformation,
projects and products grew and thus could no longer be kept isolated in the front-end SD, but increasingly
developed interfaces to non-agile, mostly back-end development tasks. This meant a “clash of two
[development] speeds” (agile team member_1). One of the most current issues is that “skill represents the
limiting factor” (CIO), referring to hard skills, soft skills, as well as an agile mindset, what makes it
difficult to select the right projects and the respective fitting teams to enter the agile boot camp.

Case 2: EventCom

Firm Context and Before Agile Phase: EventCom operates in a multi-national market environment and
retails event and entertainment offers. In the course of its ongoing digital transformation, EventCom
adopted facets of electronic commerce and online retail and engages in a large-scale agile transformation.
Prior to the agile transformation, the firm employed only two IT members that cooperated with external
IT service providers to ensure IT operations. With firm growth and the rising importance of digital
offerings, more IT members were hired and less tasks were outsourced. However, the newly staffed IT/SD
unit was not able to handle the enquiries and “could not manage to actually develop and deliver a
specific piece of software” (product owner_2) by means of traditional, plan-based development methods.
1. Agile Phase: To address this issue, EventCom built a central project management office (PMO) with ten
team members in 2011 that followed initial agile practices and employed Kanban boards (i.e., typically
physical whiteboards that show tasks arranged in three categories: to do, doing, done) to organize the
workflow of the IT/SD unit. The approach’s focal point was a monthly meeting where one representative
of each business unit came together with the IT/SD team and presented the demands of the unit “in the
form of a user story. Everybody had three to five minutes [to pitch] and there was a template for a
poster that needed to be presented. Afterwards, one could ask questions and then, by means of play
money, all those present – that was the [TM], the department heads and the representatives – could
invest in the ideas” (product owner_1). Although this idea worked initially, problems occurred over time.
The main challenge was that the PMO was not able to handle the growing enquiries, and the stakeholders
became increasingly frustrated with the fact that only a minimal percentage of ideas could be realized.
Therefore, if a project was “bought” the internal clients designed their projects to “get as much as possible
if [they] had access [to the SD unit] for once” (product owner_2). As a result, a vicious circle of excessive
demands and increasing frustration emerged. In the end, departments refused to send representatives to
the PMO meeting because they thought there was little chance of success.
2. Agile Phase: To solve the paralysis of the PMO, a bottom-up agile pilot project was initiated in 2013.
This pilot included one cross-functional, agile team (six members) working according to Scrum and
focusing on one specific customer topic, whereas the rest of the PMO continued working according to
Kanban. Also, external coaching support was sought to assist the agile transformation. The approach was
subsequently scaled by hiring more SD members, resulting in six cross-functional, agile teams with six to
nine team members working according to Scrum and focusing on specific customer groups. New tools
(e.g., Jira and Confluence) were implemented to improve the coordination of the six agile teams. With the
presence of these six teams, a structure for the coordination between the six teams as well as between the
teams and other business units emerged. The resulting steering board included the TM, which was
directly responsible for the agile teams, and department heads. However, challenges developed along the
way. Initially, the pilot project and its scaled successor were lacking visibility and acceptance outside the
SD unit owing to their strong bottom-up approach. One team member that started off in another business
unit and then became Scrum Master explained: “To be honest, I did not realize everything going on, it felt
like [the agile approach] exploded […]. It was really like a parallel world, a blackbox.” This invisibility
was also perceived as a signal of absent TM support since “the test of one single cross-functional team

Thirty Ninth International Conference on Information Systems, San Francisco 2018 9

 Becoming Agile in the Digital Transformation

shows that not a 100% commitment of [the TM] is present” (manager of the agile transformation). In fact,
the steering board partially lost trust that the agile teams “were actually developing the right things”
(product owner_1) since the coordination between business units and agile teams was problematic.
3. Agile Phase: At the beginning of 2017, the TM of EventCom restructured the entire SD department to
address problems of the agile transformation and to holistically state its support of the transformation. A
new organizational role, the product lead, was implemented. EventCom currently has two product leads
responsible for three cross-functional agile teams each, broken down into electronic commerce topics and
further topics (e.g., community building). The product leads build the structure between the agile teams,
the TM, and the business units and are responsible for a value-based SD. To ensure that the “right”
products are developed, the firm introduced a strategic alignment tool that serves as a goal achievement
measure and enables a goal-oriented coordination between the agile teams and the business units. On the
basis of this alignment tool and the new role of the product leads, Scrum-of-Scrums like meetings are
employed to bring together department heads, product owners and product leads to discuss future
development projects. The agile transformation at EventCom was also advanced by implementing the
strategic alignment tool beyond the SD unit to establish a common coordination structure. Agile methods
spread beyond the boundaries of the SD unit and there are now “teams [e.g., marketing] that think about
agile and already adopt individual agile practices” (Scrum Master). A continuous knowledge transfer is
fostered by diverse communities of practice that “enable general knowledge and experience exchange”
(agile team member) and which are available for all agile team roles (e.g., product owner, Scrum Master,
front-end developer). However, EventCom still faces challenges that hinder its large-scale agile
transformation. Although agile practices spread beyond the SD unit, not all business units are equally
suitable for implementing agile methods. In some areas agile methods might be “too much of a good thing
since in the haptic product development things are not going as fast as in the context of digital [products
and services]” (manager of agile transformation). Consequently, there are two worlds within EventCom,
one agile and one non-agile, which might be getting closer but still are far enough apart to require careful
coordination. Additionally, although the strategic alignment tool was implemented beyond the SD unit, it
was not introduced in the entire firm resulting in interfaces of the two worlds that require coordination.

Cross-Case Analysis and Discussion

Large-Scale Agile Transformation Process

We discovered that the large-scale agile transformation process at both firms can be portrayed according
to four phases. Thereby, the starting point of each of the three agile change phases is characterized by a
significant organizational effort which noticeably advances the transformation. The resulting leap
between the phases enables us to distinguish them. Although the contents of the phases (i.e., precise
challenges and actions) differ between the cases, an in-depth look reveals a similar sequence of processual
events in terms of challenges and actions in the course of both large-scale agile transformations (see Table
4 for an overview). Both firms used a stepwise approach to an agile transformation. First, the initial
transition from plan-based approaches to agile methods was undertaken in both firms. InsureTech began
with a bottom-up initiative in the IT unit that was initially promising but later was terminated owing to
diverse issues. At EventCom the TM implemented a central PMO in “Kanban mode” (product owner_2).
Although the concepts differ between the cases (bottom-up vs. top-down), both approaches represent the
initial engagement with agile methods in the respective firm. In both cases, a second agile phase was
started with several coping actions addressing challenges that developed during the first agile phase. At
InsureTech, the TM addressed earlier mistakes and initiated an extensive, top-down agile approach. In
contrast, at EventCom the accumulated challenges were addressed by means of a bottom-up initiative
characterized by cross-functionality, focus realignment and external coaching support. Scaling actions
were undertaken in both firms during this second phase ranging from extending the technological stack
(InsureTech) to hiring new employees for the agile teams (EventCom). Again, multiple challenges piled up
and were then addressed by coping actions that led to a third phase of the respective transformations.
Thereby, the TM of EventCom recognized the need to signal its support for the agile transformation and
addressed this issue in combination with further challenges by restructuring the SD unit, implementing
the new organizational role of product leads, and introducing a strategic alignment tools to foster
coordination within the entire firm. At InsureTech, the “traffic jam” (CIO) within the agile boot camp and
the lack of an exit strategy were approached by the introduction of a new organizational structure (i.e., the

Thirty Ninth International Conference on Information Systems, San Francisco 2018 10

 Becoming Agile in the Digital Transformation

agile hub). Again, the agile approaches were scaled on the basis of various aspects (e.g., establishing agile
approaches outside the SD unit) in both cases. At the end of our data collection, we identified current
challenges hindering the progress of the firms’ agile transformation. To sum up, although our two cases
are heterogeneous in terms of firm features, encountered challenges, undertaken actions and overall
approaches to the agile transformation, they can be compared on the basis of the processual flow of their
agile transformations. In both cases, this process can be clustered according to agile phases leading to the
assumption that the process of a large-scale agile transformation appears as a sequence of transformation
waves, whereas the transition between the phases seems more as a leap than a continuous flow.

Challenges of a Large-Scale Agile Transformation

Considering the nature of challenges of large-scale agile transformations, we observed that challenges
may arise collectively and form barriers that substantially hinder the progress of the agile transformation
process. Such barriers require explicit and extensive coping actions that go beyond the mitigation of
individual issues. Based on our data analysis and interpretations, we derive three archetypes of barriers
that comprise multiple challenges that occurred together and that match the socio-technical dimensions
of the challenges. First, the coordination of different organizational worlds pertains to issues of structure
such as coordination problems between multiple agile teams or between agile teams and other business
units. This archetype includes issues, such as inappropriate organizational structures, difficult leadership
dynamics, and cultural issues. Second, the barrier difficult selection of the right people pertains to people-
related challenges such as organizational members involved in the agile transformation lacking abilities,
motivation and/or an agile mindset. Finally, the barrier suitability of agile methods pertains to a
combination of challenges attached to the process as well as the technology dimension. Such challenges
encompass the overall fit of agile methods to their focal application field but also firms’ IT prerequisites
that may not be feasible for an agile approach. In addition to the archetypes, we found that challenges
may occur repeatedly throughout the process of a large-scale agile transformation. This can be seen in the
case of EventCom where the coordination of agile and non-agile teams remains an ongoing issue. Finally,
if challenges are not addressed, this can lead to the complete termination of an agile transformation as in
the case of InsureTech and its first, bottom-up agile initiative. In terms of the role of challenges in shaping
the large-scale agile transformation process, the presented barriers may be one reason why no continuous
flow between the processual phases is possible and a leap is necessary to transition between the phases.

Actions of a Large-Scale Agile Transformation

Since it is not the aim of this paper to provide best practices on how to react to certain challenges, we do
not discuss the wide range of scaling and coping actions that we captured in detail. Instead, we view these
actions as a means to depict the process of a large-scale agile transformation. Similar to the challenges
and barriers, both types of actions can be clustered according to the dimensions of the STS theory.
However, some actions of both types can be viewed in the light of multiple socio-technical dimensions
since they are general and can, for instance, be performed to address diverse challenges. As an example,
providing external coaching support as in the case of EventCom, can pertain to the process (e.g.,
supporting agile methods’ tailoring in the firm) and the people dimension (e.g., training soft skills and
fostering an agile mindset) in terms of coping actions. In addition to the finding that individual coping
actions can address multiple challenges, we also came to the conclusion that our initial differentiation of
actions in scaling and coping actions might be too narrow and may thus not represent the reality of large-
scale agile transformations appropriately. By examining the nature of the observed actions in-depth, we
found that actions initially classified as coping actions do not solely follow the purpose of mitigating
emerging transformation challenges but can also support the scaling of agile approaches. We came to the
same interpretation in terms of actions initially classified as scaling actions. For instance, we initially
categorized the introduction of new organizational structures as scaling action, since we determined the
underlying purpose as an approach to advance an agile transformation. However, as in the case of
InsureTech, the implementation of a new organizational structure (i.e., agile hub) for the return of trained
agile teams can be considered as scaling action – since more agile teams can be trained in the boot camp
and thus more agile teams emerge – but also as a coping action – since it represents an exit strategy that
mitigates the “traffic jam” (CIO) issue. Additionally, by establishing communities of practice at
EventCom, the firm did not only address the people-related challenge of missing knowledge transfer
between agile team members (i.e., coping action), but also advanced the overall agile approach (i.e.,

Thirty Ninth International Conference on Information Systems, San Francisco 2018 11

 Becoming Agile in the Digital Transformation

scaling action). These examples indicate that the two types of actions are not exclusive to each other but
overlap and that the purpose of actions in an agile transformation process can be diverse. By referring
back to literature, we found support that communities of practice, for instance, are not solely applied to
mitigate occurring challenges but can also “[support] continuous organizational improvements [in the
context of agile transformations]” (Paasivaara and Lassenius 2014, p. 1556). Consequently, by taking into
account our assumptions about the process of a large-scale agile transformation and its progression as a
sequence of transformation waves, we propose a differentiation of actions in between-phase and within-
phase actions, whereas both types of actions can follow the purpose of scaling and coping. This illustrates
the role of actions as enabler of the transition between and progression within the transformation phases.
Case 1: InsureTech Case 2: EventCom
 Bottom-up agile pilot (2007 – 2013)  Initiating central PMO (2011)
Within-Phase Actions:  Application of first agile practices (e.g., Kanban)
1. Agile Phase

 Developing an integrated agile concept Challenges:

 Offering sophisticated trainings  Poor customization of agile methods
 Transition of further IT teams (60 teams)  Problematic coordination with other business units
Challenges/Combination Resulting in Termination:  Lack of stakeholder commitment
 Lack of top management engagement Barrier:
 Inappropriate IT infrastructure  Coordination of different organizational worlds
 Inappropriate agile implementation approach
 Institutionalization of an agile boot camp (2016)  Initialization of a bottom-up, agile pilot project
Between-Phase Actions: (2013)
 Implementing a top-down, holistic approach for Between-Phase Actions:
large-scale agile transformation  Establishing cross-functionality and Scrum
 Restricting boot camp to suitable front-end SD  Realigning focus on customer groups
 Establishing independent organizational structure  Including external coaching support
2. Agile Phase

Within-Phase Actions: Within-Phase Actions:

 Integrating new technological development tools  Integrating new collaboration tools
 Inviting more teams into the boot camp  Building a structure for the agile teams
Challenges:  Hiring more agile team members
 Problematic synchronization of agile initiatives Challenges:
 Inappropriate organizational structures  Inappropriate agile implementation approach
 Inappropriate social structures  Problematic coordination with other business units
 Fear of consequences  Lack of top management engagement
Barrier: Barrier:
 Coordination of different organizational worlds  Coordination of different organizational worlds
 Implementation of an agile hub in the traditional  Top-down restructuring of bottom-up agile initiative
structures of the organization (2017) (2017)
Between-Phase Actions: Between-Phase Actions:
 Providing organizational structures for the return of  Establishing new coordination role and structure
the trained agile teams  Providing top management commitment
3. Agile Phase

Within-Phase Actions:  Implementing strategic alignment tool

 Offering training beyond the agile boot camp Within-Phase Actions:
Challenges:  Establishing communities of practice
 Inappropriate IT infrastructure  Initiating agile pilots beyond SD department
 Problematic coordination with other business units Challenges:
 Inappropriate leadership dynamics  Inappropriate application area for agile methods
 Lack of hard skills/soft skills/agile mindset  Problematic coordination with other business units
Barriers: Barriers:
 Difficult selection of the right people  Suitability of agile methods
 Coordination of different organizational worlds  Coordination of different organizational worlds

Table 4. Results of the Cross-Case Analysis

Thirty Ninth International Conference on Information Systems, San Francisco 2018 12

 Becoming Agile in the Digital Transformation

Combined Insights

Thus far, we presented our qualitative-empirical findings through the lens of STS theory in an individual
manner. By combining the insights about the process, challenges, and actions of a large-scale agile
transformation and blending them with literature on socio-technical change in the context of IS, we
conclude that a large-scale agile transformation appears as an episodic organizational change process that
can be interpreted according to the punctuated equilibrium paradigm. In contrast to evolutionism, a
punctuated equilibrium approach to organizational change posits that such a change is episodic, whereas
periods of radical and incremental change alternate (Besson and Rowe 2012; Gersick 1991; Lyytinen et al.
2009). The punctuated equilibrium paradigm is constituted by three main features: deep structures,
equilibrium periods and revolutionary periods (Gersick 1991). Deep structures represent, from an
organizational work system perspective, the fundamental, routinized characteristics of the four socio-
technical dimensions (Lyytinen and Newman 2008). Equilibrium periods are the manifestation of these
deep structures, where only incremental adaptations to the organizational work system are possible. In
contrast to this stand the revolutionary periods where the deep structures are broken up and a radical
change in the form of a holistic upheaval is possible (Gersick 1991; Lyytinen and Newman 2008). This
concept of alternating incremental and radical change fits our postulated sequence of agile waves in a
large-scale agile transformation process well. An initial upheaval is necessary to break up existing deep
structures and begin the transformation process. Subsequently, an equilibrium period follows where the
implications of agile methods, related values, and structures become routinized and only incremental
changes are possible. These new deep structures hinder the further agile transformation as we can observe
by the emerging challenges and barriers between the agile phases. These deep structures, for instance,
manifest in challenges of organizational and social structures, leadership dynamics and individuals’
mindsets and again require an upheaval, represented by a revolutionary period that begins a new agile
transformation phase. With the overall process of a large-scale agile transformation consisting of multiple
agile waves, each phase encompasses a revolutionary period with radical changes and an equilibrium
period with incremental changes. The end of each phase is characterized by one or multiple barriers that
hinder the further process. Consequently, barriers of a large-scale agile transformation are overcome by
between-phase actions that break up deep structures and facilitate radical change, whereas within-phase
actions foster incremental change and the progression of a large-scale agile transformation process. In
Figure 1, we depict this interpretation of our findings graphically. In addition, we derive two propositions:

Figure 1. Large-Scale Agile Transformation Process as Episodic Change

Proposition 1: The process of a large-scale agile transformation can be characterized as an episodic, socio-
technical change that can be portrayed by a sequence of transformation phases. Each phase encompasses
alternating aspects of radical and incremental change, whereas each phase begins with an initial upheaval
that represents a revolutionary period of the agile transformation process. Subsequently, a period of
incremental adaptations follows until the point where one or more barriers of a large-scale agile
transformation are encountered, again sparking the necessity of an upheaval.

Thirty Ninth International Conference on Information Systems, San Francisco 2018 13

 Becoming Agile in the Digital Transformation

Proposition 2: A large-scale agile transformation is mainly driven by two types of actions: between-phase
and within-phase actions. Between-phase actions refer to actions that are required for radical change and
thus needed to overcome barriers of a transformation phase and to reach the next agile transformation
phase. Within-phase actions refer to the incremental adaptations of a transformation phase undertaken in
the course of a large-scale agile transformation to foster the incremental progression of the process. Both
actions can thereby serve coping (e.g., mitigating occurring challenges) as well as scaling purposes (e.g.,
advancing the overall agile approach).

Theoretical and Practical Implications

Considering our first research question – How does the process of a large-scale agile transformation in
the context of the digital transformation progress? – we conclude that the change process of a large-scale
agile transformation can be interpreted as an episodic, socio-technical change representing a sequence of
multiple agile phases. These phases encompass two periods consisting of radical change (i.e.,
revolutionary period) and incremental change (i.e., equilibrium period) according to the punctuated
equilibrium paradigm (Besson and Rowe 2012; Gersick 1991; Lyytinen and Newman 2008).
By referring to our second research question – How do challenges of a large-scale agile transformation
as well as related coping and scaling actions shape the transformation process? – we can provide deeper
insights into the large-scale agile transformation process. We can state that the interplay of challenges,
barriers, as well as between- and within-phase actions majorly shape the process of a large-scale agile
transformation, resulting in episodic change. Throughout the course of an agile phase, multiple challenges
accumulate that collectively constitute barriers to the agile transformation. To overcome these barriers
and enable a new phase beginning with a revolutionary period, an upheaval, often represented by
extensive organizational actions (e.g., creating new structures) is necessary. Overall, this depiction of a
large-scale agile transformation process reflects the current state of literature on the topic well. From our
qualitative data, we can confirm that such a process is rarely straightforward or easy to handle (Dikert et
al. 2016) and is typically approached in a stepwise manner (Paasivaara et al. 2018). Additionally, there is
no perfect roadmap on how to master an agile transformation since firms encounter diverse constellations
of the manifold challenges that can emerge in the course of a large-scale agile transformation process
(Paasivaara et al. 2018). Consequently, the alternating periods of radical and incremental change are
required to evaluate and refine the organizational approach applied in the agile transformation process.
Besides offering an explorative interpretation of large-scale agile transformations as episodic change
processes and explaining the role of challenges and actions in shaping this process, we also provide
insights on the nature of challenges and actions. In terms of the challenges of an agile transformation
process, we provide contribution by initially deriving three archetypes of barriers that are formed by
collectively emerging challenges: a) coordination of different organizational worlds, b) difficult selection
of the right people, and c) suitability of agile methods. Especially the latter barrier may have extensive
implications for organizations. Although, agile methods are increasingly applied in various contexts
outside the SD units (e.g., human resources), the question remains whether agile methods are feasible for
all organizational activities and thus the entire organizational work system (Cappelli and Tavis 2018;
Rigby et al. 2016). Consequently, the suitability of agile methods for firms based on their individual
features, such as size, industry, products and service offerings, might cap the ceiling of a large-scale agile
transformation. Various types of such natural barriers might exist, such as legal issues (e.g., contractual
regulations), processual standards (e.g., ISO norms), IS security requirements or simply a lack of
performance improvement by implementing agile methods. Literature addresses some of these aspects
(e.g., Gregory et al. 2015), however, since we can only initially assume the possible characteristics of
natural barriers, further research is required in this regard. Additionally, it should be considered that a
large-scale agile transformation is often embedded in an overarching approach to firms’ digital
transformation entailing a scope for the overall organizational agile transformation. In addition to the
presentation of potential barriers that occur in large-scale agile transformations, we also revised our
initial categorization of coping and scaling actions based on our overall empirical findings. Consequently,
our proposition of between- and within-phase actions takes into account that undertaken actions of firms
may not follow a sole purpose of, for instance, tackling a specific challenge, but may also serve several
goals and thus scaling as well as coping purposes. This matches our interpretation that overcoming
barriers can require a combination of scaling and coping actions and thus can trigger a new agile phase

Thirty Ninth International Conference on Information Systems, San Francisco 2018 14

 Becoming Agile in the Digital Transformation

starting with radical change. Since current literature mainly discusses success factors of large-scale agile
transformations (e.g., Dikert et al. 2016) and does not distinguish types of coping actions (e.g., Paasivaara
et al. 2018), we offer an initial classification of actions to advance large-scale agile transformations.
Our research also offers relevant practical implications. Our proposition of an agile transformation as
episodic change supports managers in comprehending and structuring the change process in their firms.
Additionally, practitioners should be aware that a large-scale agile transformation can lead to a division
within the firms into agile and non-agile worlds. Although agile methods may not be applied throughout
the entire firm, transparency about the agile world including practices, principles, and mindsets should be
fostered and spread throughout the organization. Additionally, potential organizational interfaces require
feasible coordination. This can be achieved by dedicated organizational roles. In addition, coordination
can be supported by strategic alignment tools, such as in the case of EventCom, where the tool is applied
to summarize overall organizational goals and break them down into measurable key figures for the units
involved (agile and non-agile). These figures can also foster the transparency and visibility of the agile
approaches in a firm. Agile initiatives may start as bottom-up approaches instigated by individuals.
However, if firms aim at becoming holistically agile, such bottom-up approaches need to be supported
and at some point transferred into more extensive top-down approaches that are backed by the TM.
Otherwise, these bottom-up initiatives will lack long-term tenacity and eventually rather hinder a large-
scale agile transformation as in the case of InsureTech. Overall, as there is no perfect guideline on large-
scale agile transformations, managers are best advised to follow an “experimental transformation
approach” (Paasivaara et al. 2018, p. 36) that not only fits the agile mindset but also enables the
continuous evaluation of the change process and a rapid reaction to challenges and barriers.

Conclusion, Limitations, and Areas for Future Research

Based on the empirical findings of our study, we propose that a large-scale agile transformation can be
interpreted as an episodic, socio-technical change process according to the punctuated equilibrium
paradigm that includes a sequence of agile phases. We also identify barriers to an agile transformation
that are formed by the challenges of the change process and that hinder the continuous flow and transfer
from one phase to the next. In an effort to overcome these barriers, firms engage in specific actions that
trigger radical change. These between-phase actions stand in contrast to within-phase actions that
advance the agile transformation by incremental adaptations. Overall, by means of our study, we provide
initial insights into a large-scale agile transformation process as part of the organizational digital
transformation and thus take up a current call for research (Dikert et al. 2016; Paasivaara et al. 2018).
Nonetheless, we acknowledge the limitations of our study. First of all, our results stem from qualitative-
empirical research and thus might be of limited generalizability. Nonetheless, processual insights require
an in-depth understanding of the research topic, which is best facilitated by qualitative research
approaches such as case studies (Myers 2009). Since we only observed stepwise approaches to large-scale
agile transformations in our case studies, we cannot generalize our findings on the second option of such a
transformation, the big bang approach, where the entire change process is performed at once. However,
firms are typically advised to undertake radical transformations, such as a large-scale agile
transformation, stepwise (Paasivaara et al. 2018). In addition, we acknowledge that we do not offer best
practices or a guide on how to respond effectively to specific challenges and barriers since we are of the
opinion that a generalization of “best reactions” is not appropriate given that a large-scale agile
transformation is a highly complex change process impacted by various aspects as well as the interplay of
diverse firm features, emerging challenges, and performed actions. The derivation of such best reactions
would be a promising direction for future research. We also call for further research on large-scale agile
transformations in the context of the digital transformation to deepen the understanding of the change
processes as well as validate or revise our findings and test our propositions in comparable organizational
settings. Finally, our study raised the topic of possible natural barriers to an agile transformation that cap
the ceiling of the process and thus also limit the potential increase in the organizational agility of firms.
Therefore, we call for research on this topic since it may not only spark intriguing theoretical insights but
also pertinent practical implications, which are especially relevant for a fast-changing, digital world.

Thirty Ninth International Conference on Information Systems, San Francisco 2018 15

 Becoming Agile in the Digital Transformation

References
Abrahamsson, P., Conboy, K., and Wang, X. 2009. "'Lots Done, More to Do': The Current State of Agile
Systems Development Research," European Journal of Information Systems (18:4), pp. 281-284.
Alter, S. 2013. "Work System Theory: Overview of Core Concepts, Extensions, and Challenges for the
Future," Journal of the Association for Information Systems (14:2), pp. 72-121.
Ambler, S. W., and Lines, M. 2012. Disciplined Agile Delivery: A Practitioner's Guide to Agile Software
Delivery in the Enterprise. Boston, USA: IBM Press.
Beck, K., Beedle, M., Van Bennekum, A., Cockburn, A., Cunningham, W., Fowler, M., Grenning, J.,
Highsmith, J., Hunt, A., Jeffries, R., Kern, J., Marick, B., Martin, R. C., Mellor, S., Schwaber, K.,
Sutherland, J., and Thomas, D. 2001a. "Manifesto for Agile Software Development."
Beck, K., Beedle, M., Van Bennekum, A., Cockburn, A., Cunningham, W., Fowler, M., Grenning, J.,
Highsmith, J., Hunt, A., Jeffries, R., Kern, J., Marick, B., Martin, R. C., Mellor, S., Schwaber, K.,
Sutherland, J., and Thomas, D. 2001b. "Principles behind the Agile Manifesto."
Benbasat, I., Goldstein, D. K., and Mead, M. 1987. "The Case Research Strategy in Studies of Information
Systems," MIS Quarterly (11:3), pp. 369-386.
Besson, P., and Rowe, F. 2012. "Strategizing Information Systems-Enabled Organizational
Transformation: A Transdisciplinary Review and New Directions," The Journal of Strategic
Information Systems (21:2), pp. 103-124.
Bharadwaj, A., El Sawy, O. A., Pavlou, P. A., and Venkatraman, N. 2013. "Digital Business Strategy:
Toward a Next Generation of Insights," MIS Quarterly (37:2), pp. 471-482.
Boehm, B., and Turner, R. 2005. "Management Challenges to Implementing Agile Processes in
Traditional Development Organizations," Project Management (22:5), pp. 30-39.
Bostrom, R. P., Gupta, S., and Thomas, D. 2009. "A Meta-Theory for Understanding Information Systems
within Sociotechnical Systems," Journal of Management Information Systems (26:1), pp. 17-47.
Cao, L., Mohan, K., Xu, P., and Ramesh, B. 2009. "A Framework for Adapting Agile Development
Methodologies," European Journal of Information Systems (18:4), pp. 332-343.
Cappelli, P., and Tavis, A. 2018. "HR Goes Agile," Harvard Business Review (96:2), pp. 46-52.
Conboy, K., Coyle, S., Wang, X., and Pikkarainen, M. 2011. "People over Process: Key Challenges in Agile
Development," IEEE Software (28:4), pp. 48-57.
Dikert, K., Paasivaara, M., and Lassenius, C. 2016. "Challenges and Success Factors for Large-Scale Agile
Transformations: A Systematic Literature Review," The Journal of Systems and Software (119), pp.
87-108.
Dingsøyr, T., and Moe, N. B. 2014. "Towards Principles of Large-Scale Agile Development: A Summary of
the Workshop at XP2014 and a Revised Research Agenda," Proceedings of the International
Conference on Agile Software Development, Rome, Italy, pp. 1-8.
Dubé, L., and Paré, G. 2003. "Rigor in Information Systems Positivist Case Research: Current Practices,
Trends, and Recommendations," MIS Quarterly (27:4), pp. 597-636.
Gersick, C. J. G. 1991. "Revolutionary Change Theories: A Multilevel Exploration of the Punctuated
Equilibrium Paradigm," Academy of Management Review (16:1), pp. 10-36.
Gregory, P., Barroca, L., Taylor, K., Salah, D., and Sharp, H. 2015. "Agile Challenges in Practice: A
Thematic Analysis," Proceedings of the International Conference on Agile Software Development,
Helsinki, Finland, pp. 64-80.
Hekkala, R., Stein, M.-K., Rossi, M., and Smolander, K. 2017. "Challenges in Transitioning to an Agile
Way of Working," Proceedings of the 50th Hawaii International Conference on System Sciences,
Hawaii, USA, pp. 5869-5878.
Jöhnk, J., Röglinger, M., Thimmel, M., and Urbach, N. 2017. "How to Implement Agile IT Setups: A
Taxonomy of Design Options," Proceedings of the 25th European Conference on Information
Systems, Guimarães, Portugal, pp. 1521-1535.
Lee, G., and Xia, W. 2010. "Toward Agile: An Integrated Analysis of Quantitative and Qualitative Field
Data on Software Development Agility," MIS Quarterly (34:1), pp. 84-114.
Leonhardt, D., Haffke, I., Kranz, J., and Benlian, A. 2017. "Reinventing the IT Function: The Role of IT
Agility and IT Ambidexterity in Supporting Digital Business Transformation," Proceedings of the 25th
European Conference on Information Systems, Guimarães, Portugal, pp. 968-984.
Lyytinen, K., and Newman, M. 2008. "Explaining Information Sytems Change: A Punctuated Socio-
Technical Change Model," European Journal of Information Systems (17:6), pp. 589-613.

Thirty Ninth International Conference on Information Systems, San Francisco 2018 16

 Becoming Agile in the Digital Transformation

Lyytinen, K., Newman, M., and Al-Muharfi, A.-R. A. 2009. "Institutionalizing Enterprise Resource
Planning in the Saudi Steel Industry: A Punctuated Socio-Technical Analysis," European Journal of
Information Systems (24:4), pp. 286-304.
Miles, M. B., Huberman, A. M., and Saldana, J. 2013. Qualitative Data Analysis - A Methods Sourcebook,
(3rd ed. ed.). Thousand Oaks, USA: SAGE Publications.
Mumford, E. 2003. Redesigning Human Systems. Hershey, USA: Information Science Publishing.
Myers, M. D. 2009. Qualitative Research in Business & Management. London, UK: SAGE Publications.
Nerur, S., Mahapatra, R., and Mangalaraj, G. 2005. "Challenges of Migrating to Agile Methodologies,"
Communications of the ACM (48:5), pp. 72-78.
Orlikowski, W. J. 1992. "The Duality of Technology: Rethinking the Concept of Technology in
Organizations," Organization Science (3:3), pp. 398-427.
Paasivaara, M., Behm, B., Lassenius, C., and Hallikainen, M. 2018. "Large-Scale Agile Transformation at
Ericsson: A Case Study," Empirical Software Engineering), pp. 1-47.
Paasivaara, M., and Lassenius, C. 2014. "Communities of Practice in a Large Distributed Agile Software
Development Organization – Case Ericsson," Information and Software Technology (56:12), pp.
1556-1577.
Paasivaara, M., Lassenius, C., and Heikkilä, V. T. 2012. "Inter-Team Coordination in Large-Scale Globally
Distributed Scrum: Do Scrum-of-Scrums Really Work?," Proceedings of the International
Symposium on Empirical Software Engineering and Measurement, Lund, Sweden, pp. 235-238.
Paré, G. 2004. "Investigating Information Systems with Positivist Case Research," Communications of the
Association for Information Systems (13:1), pp. 233-264.
Patton, M. Q. 2002. Qualitative Research & Evaluation (3rd ed. ed.). Thousand Oaks, USA: SAGE
Publications.
Pentland, B. T. 1999. "Building Process Theory with Narrative: From Description to Explanation,"
Academy of Management Review (24:4), pp. 711-724.
Poole, M. S., Van de Ven, A. H., Dooley, K., and Holmes, M. E. 2000. Organizational Change and
Innovation Processes - Theory and Methods for Research. Oxford, UK: Oxford University Press.
Porter, M. E., and Heppelmann, J. E. 2015. "How Smart, Connected Products Are Transforming
Companies," Harvard Business Review (93:10), pp. 96-114.
Rigby, D. K., Sutherland, J., and Takeuchi, H. 2016. "Embracing Agile - How to Master the Process that's
Transforming Management," Harvard Business Review (94:5), pp. 41-50.
Rogers, D. L. 2016. The Digital Transformation Playbook - Rethink Your Business for the Digital Age.
New York, USA: Columbia University Press.
Rolland, K. H., Fitzgerald, B., Dingsøyr, T., and Stol, K.-J. 2016. "Problematizing Agile in the Large:
Alternative Assumptions for Large-Scale Agile Development," Proceedings of the 37th International
Conference on Information Systems, Dublin, Ireland, pp. 1-21.
Ryan, S. D., Harrison, D. A., and Schkade, L. L. 2002. "Information-Technology Investment Decisions:
When Do Costs and Benefits in the Social Subsystem Matter?," Journal of Management Information
Systems (19:2), pp. 85-127.
Sarker, S., Chatterjee, S., and Xiao, X. 2013. "How "Sociotechnical" Is Our Research? An Assesment and
Possible Ways Forward," Proceedings of the 34th International Conference on Information Systems,
Milan, Italy, pp. 1-24.
Scaled Agile. 2017. "SAFe® 4.5 Introduction - Overview of the Scaled Agile Framework® for Lean
Enterprises." Boulder, USA: Scaled Agile, Inc.
Scheerer, A., Hildenbrand, T., and Kude, T. 2014. "Coordination in Large-Scale Agile Software
Development: A Multiteam Systems Perspective," Proceedings of the 47th Hawaii International
Conference on System Science, Hawaii, USA, pp. 4780-4788.
Vodde, B., and Larman, C. 2014. "LeSS Framework." The LeSS Company B.V.
Yin, R. K. 2013. Case Study Research - Design and Methods, (5th ed. ed.). Thousand Oaks, USA: SAGE
Publications.
Zheng, Y., Venters, W., and Cornford, T. 2011. "Collective Agility, Paradox and Organizational
Improvisation: The Development of a Particle Physics Grid," Information Systems Journal (21:4), pp.
303-333.

Thirty Ninth International Conference on Information Systems, San Francisco 2018 17

View publication stats