Gupea 2077 39860 1
Gupea 2077 39860 1
Gupea 2077 39860 1
Purpose – Agility has been recognised as essential for firms operating in turbulent, uncertain
and complex environments, in order to preserve a competitive advantage and ensure survival.
However, not many organisations have considered the role of agility with respect to their
innovation activities. Agility has been recently recognised as a way of dealing with innovation
activities in turbulent times. Currently, there is little empirical evidence in the existing
literature that links agility to innovation. Because of its critical implications for the existence
and success of a business, the main purpose of this study is to gain new insights into the
impact of agility on the innovation productivity of large organisations in the European Union
(EU).
Approach – The concept of agility is fragmented into several agility attributes based on
previous empirical works. A generally positive relationship is expected between these agility
attributes and innovation productivity. This relationship is tested by considering large
organisations in the EU, by using a survey methodology, by implementing a structured
questionnaire as the research instrument to obtain information, and by performing three
independent Negative Binomial regressions in order to analyse the data.
Findings – Different agility factors depict a significant and positive impact on the
productivity of the three types of innovation.
Research limitations – It is difficult to generalise the results, due to the use of a non-
probability sampling method and the final sample size.
Originality – Most of the literature on agility is concerned with agile manufacturing. Scholars
have only recently started to recognise the importance of applying agility to innovation. This
gap presents an interesting area of research. The study tests the effects of agility on innovation
productivity in large organisations in the EU for the first time, and opens up to challenging
future researches. It also seeks to address the problem of innovation processes being slow and
unproductive.
II
Acknowledgements
First and foremost, the authors of this research extend their gratitude to their immediate
supervisor Daniel Ljungberg, who provided valuable feedback throughout the course of this
research and played a pivotal role in guiding this research work to its completion. The authors
also thank Johan Brink for his help in refining this study and providing critical feedback at the
very onset of this study, which helped the authors define a clear research question. In
addition, the authors appreciate the critical feedback provided by Rick Middel during the
terminal stage of this study, thereby helping the authors address some key aspects of the
study. Finally, the authors would like to express their gratitude to Thomas Hagbard, CEO and
Idea Pilot at Realize AB, for his highly valuable industry insight, which helped lay the
foundation for this research work.
III
Table of Contents
1 Introduction ................................................................................................... 1
1.1 The Innovation Problem: “It takes too much time” .................................................... 1
1.2 Need for Agility: The New Methodology ................................................................... 1
1.3 Motivation ................................................................................................................... 2
1.4 Research Question ....................................................................................................... 3
1.5 Scope of the Research.................................................................................................. 4
1.6 Thesis Content ............................................................................................................. 5
2 Theoretical Background ............................................................................... 6
2.1 Settings ........................................................................................................................ 6
2.1.1 Research Areas ................................................................................................................ 6
2.1.2 Scope of the Theoretical Background.............................................................................. 7
2.2 Agility .......................................................................................................................... 7
2.2.1 Firm Agility ..................................................................................................................... 7
2.2.2 Achieving Firm Agility .................................................................................................. 11
2.2.3 Summary on Agility ....................................................................................................... 16
2.3 Innovation .................................................................................................................. 19
2.3.1 Product Innovation ........................................................................................................ 19
2.3.2 Process Innovation ........................................................................................................ 19
2.3.3 Business Model Innovation............................................................................................ 20
2.3.4 Innovation Productivity ................................................................................................. 21
3 Methodology ................................................................................................ 22
3.1 Research Approach .................................................................................................... 22
3.1.1 Goal ............................................................................................................................... 22
3.1.2 Strategy.......................................................................................................................... 22
3.2 Research Design ........................................................................................................ 24
3.2.1 Field Study..................................................................................................................... 24
3.2.2 Cross-sectional Design.................................................................................................. 24
3.3 Research Methods...................................................................................................... 25
3.3.1 Survey Methodology ...................................................................................................... 25
3.3.2 Questionnaire ................................................................................................................ 25
3.3.3 Regression Analysis Methodology................................................................................. 26
3.4 Data Collection .......................................................................................................... 27
3.4.1 Sampling ........................................................................................................................ 27
3.4.2 Questionnaire Design .................................................................................................... 28
IV
3.5 Research Criteria ....................................................................................................... 31
3.5.1 Replicability................................................................................................................... 31
3.5.2 Reliability ...................................................................................................................... 32
3.5.3 Validity .......................................................................................................................... 32
3.6 Execution ................................................................................................................... 32
3.6.1 Survey Administration ................................................................................................... 32
3.6.2 Final Sample.................................................................................................................. 34
5 Conclusion.................................................................................................... 59
5.1 Interpretation ............................................................................................................. 60
5.2 Limitations ................................................................................................................. 61
5.3 Future Research ......................................................................................................... 63
6 References .................................................................................................... 64
7 Appendices ................................................................................................... 69
Appendix A: Agile Methods................................................................................................. 69
Appendix B: Contact Message Draft .................................................................................... 72
Appendix C: Questionnaire .................................................................................................. 73
Appendix D: Respondent Sample – Organisations .............................................................. 76
V
List of Figures
Figure 1.1 Agility as a catalyst for innovation.......................................................................... 4
Figure 2.1 Conceptual model of an agile enterprise ............................................................... 12
Figure 2.2 Conceptual model of an agile enterprise – Complete ............................................ 18
Figure 2.3 Business model canvas ........................................................................................... 21
Figure 2.4 Agility as a catalyst for innovation......................................................................... 21
Figure 3.1 Research approach ................................................................................................. 23
Figure 3.2 Survey administration across geographic area...................................................... 33
Figure 3.3 Responses across geographic area ........................................................................ 34
Figure 3.4 Distribution of respondents across positions ......................................................... 35
Figure 4.1 Visual representation of optimal number of factors ............................................... 40
List of Tables
Table 2.1 Key search words ....................................................................................................... 6
Table 2.2 Definitions of agility over time ................................................................................... 9
Table 3.1 The data rectangle in cross-sectional design ........................................................... 24
Table 3.2 List of dependent, independent and control variables ............................................. 31
Table 3.3 Survey data ............................................................................................................... 33
Table 4.1 Variable names and description............................................................................... 37
Table 4.2 Descriptive statistics ................................................................................................ 38
Table 4.3 Correlation matrix .................................................................................................. 38
Table 4.4 Factor analysis ......................................................................................................... 39
Table 4.5 Factor loadings ........................................................................................................ 40
Table 4.6 Rotated factor loadings ............................................................................................ 41
Table 4.7 Classification of independent variables into factor loadings .................................. 42
Table 4.8 Reliability of PCA .................................................................................................... 42
Table 4.9 Mean and variances of dependent variables ............................................................ 43
Table 4.10 Comparison mean / variances ................................................................................ 43
Table 4.11 Three independent regressions and their respective independent variables ......... 45
Table 4.12 Negative Binomial regression for Num_PDI ......................................................... 47
Table 4.13 Negative Binomial regression for Num_PRI.......................................................... 50
Table 4.14 Negative Binomial regression for Num_BMI ......................................................... 53
Table 4.15 Negative Binomial regression for manufacturing industry.................................... 55
Table 4.16 Negative Binomial regression for services industry .............................................. 56
Table 4.17 Summary of results ................................................................................................. 57
VI
Abbreviations
BM Business Model
BMI Business Model Innovation
CIS Community Innovation Survey
EU European Union
FA Factor Analysis
PDI Product Innovation
PCA Principal Component Analysis
PRI Process Innovation
VII
1 Introduction
The concept of Agile Innovation has been recently recognised as a way of dealing with
innovation operations in these turbulent times. Scholars understand the need for faster
innovation processes and acknowledge the importance of a more agile innovation
process. Agile innovation was described by Morris (2014) as “a radically new and eminently
practical approach to the challenge of survival”. However, a closer look at what agile practices
should be employed and what impact they may have on innovation is lacking in the current
literature on this subject.
Agility and agile methods have only recently been borrowed and applied to different divisions
of the organisation. There is often talk about “Agile Supply Chain”, “Agile Project
Management” and even an “Agile Organisation” as a whole, where being agile is defined as a
dynamic capability (Teece, 2009; Worley and Lawler, 2010). The primary goal of applying
agility on an organisational level is to respond to change, but more specifically to “manage” it
(Dove, 2001; Shafer, 1997). In present times, agility is considered as an essential ingredient for
not only preserving a competitive advantage, but also for survival (Ganguly, 2009; Morris,
2014).
A number of organisations have now started to apply such methods to their innovation
activities (Wilson and Doz, 2011). The current environment of accelerating change has made
innovation hazardous and tricky to time as argued in the previous section. Yet, innovation is
also perceived as crucial to survival (Schumpeter, 1939; Dodgson et al., 2008). Because of its
critical implication for the existence and success of a business, this study will shed light on the
implications of agility in the management of innovation.
1.3 Motivation
Although most of the scholars who have touched upon this area of research have acknowledged
the need for adopting agility within an organisation, especially manufacturing, there is almost
non-existent research on whether such agile practices have an impact on innovation in terms of
productivity. This presents an important area of research because most organisations nowadays
are engaged in either one or all types of innovations mentioned above. However, their
innovation processes often lack productivity and may take up too much time. This study will
help gain a better understanding into whether this problem can be addressed using the concept
of agility.
Furthermore, no quantitative study – to the extent of the authors’ knowledge – has been
conducted in this field. It is therefore a great opportunity to add to the literature and suggest
further studies based on this research.
2
Finally, this study will give a unique and collective insight into the different types of
innovations within an organisation i.e. Product Innovation, Process Innovation and Business
Model Innovation. It is of particular importance with regards to Process Innovation and
Business Model Innovation, due to the lack of existing literature on these concepts. In fact,
these two types of innovations are of great importance nowadays as they can bring a unique
competitive advantage to an organisation (Teece, 2009).
To what extent can agility impact the innovation productivity of large organisations?
‘Agility’ is defined as the aptitude of a firm to react to change by organising in a flexible way
at every level, thereby enabling the firm to prepare for change and to lead change in unforeseen
circumstances.
The research question also takes into account the term ‘large organisations’, which for the
purpose of this study depicts organisations with more than 250 employees 1. This criterion was
set in order to capture the effect on only large organisations and thereby eliminate the bias that
may be caused by the size of the organisation.
The particularity of this research is its unique focus on the three types of innovations present in
an organisation 2: Product 3 Innovation (PDI), Process Innovation (PRI) and Business Model
Innovation (BMI). For this reason, the term ‘Innovation Productivity’ refers to productivity
with regards to these three types of innovations. The definitions of these three types of
innovations were adapted from the Community Innovation Survey 2010. Therefore, Product
Innovation refers to the market introduction of a new or significantly improved good or service;
Process Innovation refers to the implementation of a new or significantly improved production
process, distribution method or supporting activity; and Business Model Innovation refers to
the creation of a new or significantly improved value offering, by taking into account
customers, infrastructure, and the revenue model.
The Innovation Productivity definition was built starting from the definition of productivity –
the rate at which goods/services (outputs) are produced from a standard set of inputs. The
authors identify the outputs in terms of the number of innovations (Product, Process or
1
A large-scale organisation employs more than 200 employees (Somers, Cain and Jeffery, 2011)
2
Based on the Global CEO study of IBM (2006)
3
Product takes into account both good and services
3
Business Model) produced by a firm during the three years 2012 - 2014. The inputs will be all
activities at the organisational level that are required to carry out innovation activities. The
following Figure 1.1 illustrates the role of agility as a catalyst.
Z
X+Y Agility (Output)
(Units of Input)
Number of Innovations
It is expected that a generally positive relationship exists between agility and innovation
productivity. The following recent industry example further elucidates this intuition.
The Volvo CarPlay Solution
When Apple released its CarPlay interface in 2013, the challenge for carmakers to be the first
to adopt the technology was real. Carmakers had to deal with a completely new set of design
features and figure out how to integrate the new Apple technology with their existing
dashboard display systems. Volvo Cars, by employing agile practices, was able to launch the
Volvo CarPlay solution in less than 50 percent of the industry standard time. In order to
accelerate the innovation process, Volvo collaborated with a small engineering firm from the
Silicon Valley, Symbio. Symbio brought external knowledge, promoted a fast and flexible
mind-set during the brainstorming sessions, and promoted the use of iterative strategies. Such
agile practices accelerated the innovation process at Volvo, and it can be implied that agility
played the role of a catalyst in the innovation process (Morris, 2015).
Furthermore, this study does not aim to measure the degree to which an organisation is agile or
has implemented agile practices. The research is designed to observe the independent impact of
agility practices that will be defined as agility attributes for the purpose of this study.
It is also important to note that the research will be conducted by only taking into account
organisations that are conducting activities within the European Union. The study is therefore
only a reflection of the organisations present in the EU and may or may not give an accurate
reflection of organisations outside the EU.
4
1.6 Thesis Content
This research study is structured as follows:
Section 1 – Introduction presents the research problem and introduces the research question
that will guide the study.
Section 2 – Theoretical Framework clarifies the concept of agility and then defines the three
types of innovations considered in this research, along with the notion of innovation
productivity.
Section 3 – Methodology lays the foundation of the research by providing a detailed and
argued presentation of the research approach, design and methods employed in this research.
This section also offers a complete description of the data collection process, its execution and
its outcomes.
Section 4 – Data Analysis summarises and models the data collected. A factor analysis will be
carried out to reduce the number of variables, followed by a regression analysis in order to
depict a relationship between the attributes of agility and innovation productivity (in terms of
PDI, PRI and BMI).
5
2 Theoretical Background
2.1 Settings
2.1.1 Research Areas
In order to create a theoretical framework for this research, a systematic literature review was
conducted. The literature review was broken down into two research areas: Agility and
Innovation.
The literature on these two research areas was gathered through journal articles, books and
dissertations. The databases used included Scopus, Science Direct, Emerald, SpringerLink and
Business Source Premier inter alia.
The following Table 2.1 summarises the key words used to collect the relevant literature on the
two research areas.
Agility
Agile
Agility
Agile Practices
Innovation
Product Innovation
Process Innovation
Business Model Innovation
Agile Innovation
Innovation Productivity
Innovation and Agility
6
2.1.2 Scope of the Theoretical Background
The first research area will be concerned with the concept of agility. It is a rather new
concept and the term agility was first employed in 1991 (Dove, 1999; Yusuf et al., 1999). The
literature on agility, however, is extremely broad. Scholars along with practitioners have
developed several definitions and frameworks of agility from various perspectives, which
renders the literature inconsistent and disjointed.
The purpose of this first part is not to question the existing theory of agility, nor to give a
complete overview of agility during the last two decades. Instead the literature review will
attempt to provide an overview of the most consistent views on agility.
Agility will be defined from a broad perspective i.e. embodying the concept of firm agility. In
addition, this concept will be deconstructed in order to reveal three levels of agility: (1) agility
attributes; (2) agility capabilities; (3) agility drivers.
It is important for the reader to note that the purpose of this research is not to measure agility.
Therefore, the degree to which a firm is agile will not be approached in the theoretical
background. Additionally, the reader should be aware that the literature review does not reflect
the entire complexity behind the term ‘agility’, but instead tries to give both a simplified and
exhaustive view of it.
The second research area will be concerned with the three types of innovations this
research focuses on. Product Innovation (PDI), Process Innovation (PRI) and Business Model
Innovation (BMI) will be defined, along with the concept of Innovation Productivity.
2.2 Agility
2.2.1 Firm Agility
Agility: “The ability to move readily and quickly, the ability to think and understand readily
and quickly; dexterity; alertness” 4
As straightforward as the dictionary definition may seem, the term ‘agility’ is multifaceted
from a managerial perspective, which makes it hard to define. Indeed, the literature on agility is
fragmented, heterogeneous and rather ambiguous (Sherehiy, Karwowski & Layer, 2007;
Audran, 2011) and only a few scholars have considered agility in its entirety (Charbonnier,
2011). As different aspects of agility have been emphasised by different scholars, various
interpretations of agility are reflected in the existing literature (Yusuf et al., 1999).
To gain a better understanding, the main definitions of agility were gathered in table 2.2. Three
converging characteristics were depicted: changing environment, time and responsiveness. The
following section will review these definitions in order to agree upon a definition of agility
which will be used throughout this research.
4
The Oxford English Dictionary
7
Changing
Authors Definition of agility environment Time Responsiveness
Iacocca/Leigh (1991) A system that shifts quickly among product
models/lines, ideally in real time in order to X X
respond to customer needs.
Goldman et al. (1995) Capability of an organisation to operate
profitably in a competitive environment
X X
comprised of continually changing
customer habits.
Vokurka and Fliedner Ability to successfully produce and market
(1998) a broad range of low cost, high quality
products with short lead times in varying
X
size, which provide enhanced value to
individual customers through
customisation.
Yusuf et al. (1999) A successful exploration of competitive
bases (speed, flexibility, innovation,
proactivity, etc.) through the integration of
reconfigurable resources and knowledge X X X
management, to provide customer driven
products/services in a fast changing market
environment.
Rigby et al. (2000) The ability of an organisation to thrive in a
constantly changing and unpredictable X
business environment.
Sharifi and Zhang (1999 Agility is the ability of enterprises to
and 2000) respond to change, to cope with unexpected
changes, to survive unprecedented threats X X
from the business environment, and to take
advantage of changes as opportunities.
Hooper et al. (2001) Ability of an enterprise to develop and
exploit its inter and intra-organisational
capabilities to successfully compete in an X
uncertain and unpredictable business
environment.
Sharifi et al. (2001) Two main factors: (i) responding to change
(anticipated or not) in due time, (ii)
X X X
exploiting changes and taking advantage of
changes as opportunities.
Dove (1999 and 2001) An effective integration of response ability
and knowledge management in order to
rapidly, efficiently and accurately adapt to X X X
unexpected changes in both proactive and
reactive needs and opportunities […].
8
Nirmal (2005) The capability to respond to new business
demands and opportunities effectively and
efficiently, rapidly shifting and aligning X X X
business assets to beat to competition to
market.
Ashrafi et al. (2005) An organisation’s ability to sense
environmental changes and respond X X X
effectively and efficiently to that change.
Sull (2009) The capacity to identify, capture, and
exploit opportunities more quickly than X X X
rivals do.
Tallon & Pinsonneault Ability to detect and respond to
(2011) opportunities and threats in the environment X X X
with ease, speed, and dexterity.
In management literature, agility originated in 1991 based on the awareness that the
environment of the firms started to move faster than the ability of the firms to adapt (Dove,
1999; Yusuf et al., 1999). At that time, a group of scholars from Iaccoca Institute at Lehigh
University were discussing the new paradigm of manufacturing (Yusuf et al., 1999). They
advocated “a holistic, rather than a sub-optimal, approach to manufacturing” and
acknowledged that “the main driving force behind agility is change”. Hence, agility was
defined as “a manufacturing system with extraordinary capabilities to meet rapid changing
needs of the marketplace […] a system that shifts quickly among product models or product
lines, ideally in real time response to customer demand” (Youssef, 1994). Not so long after, the
Agile Manufacturing Enterprise Forum (AMEF) was established in order to disseminate the
vision of “the agile enterprise” inside organisations in the United States (Dove, 2001). Since
then, scholars have refined the definition of agility, which has resulted in multiple
interpretations.
When reviewing the different definitions of agility, the vast majority of the scholars who
touched upon the topic defined agility as the aptitude of a firm to adapt quickly to external
changes (Dove, 1999; Sharifi and Zhang, 1999 and 2000; Rigby et al., 2000). From the
definitions gathered in table 2.2, agility is always defined as an answer to turbulent and
unstable markets and business environments. As claimed by Yusuf et al. (1999), the main
driving force behind agility is change. Therefore, one of the main characteristics of agility is
the Changing business environment. These changes can be predictable (e.g. an expected new
regulation affecting the industry) or unpredictable and unforeseen (e.g. market volatility caused
by a disruptive innovation).
9
Characteristic 2: Time
A second essential characteristic of agility embedded within the definitions gathered in table
2.2 is Time. Yusuf et al (1999), Sharifi et al (2001), Dove (2001), Nirmal (2005), Sull (2009)
and Tallon and Pinsonneault (2011) emphasised quickness and speed as a way to answer
changes within the business environment of firms. One of the main perceptions of agility, as
expressed by Dove (2001), is that agility is a combination of speed and flexibility (Agility =
Flexibility + Speed). Vokurka and Fliedner (1998) defined flexibility as the ability of an
organisation to transit between a variety of tasks as a routine and a predetermined process.
However, a firm’s agility does not only comprise of the ability to respond rapidly, but more
importantly to respond efficiently to unexpected changes (Goldman et al., 1995; Vokurka and
Fliedner, 1998). It can then be argued that in order to be agile, a firm needs to be flexible.
Following the logic of Wadhwa (2003) and considering the definition given by Vokurka and
Fliedner (1998) on flexibility, agility relies on flexibility by incorporating the ability to respond
to unforeseen changes in the market in a quick way. Therefore, it can be understood that the
concept of flexibility embeds the characteristic of time; by organising in a flexible way, a firm
can cope with changes occurring in the business environment with little time penalty (Wadhwa,
2003).
Nirmal (2005) discerned two sides of agility: the reactive side and the proactive side. The
reactive side is driven by external forces such as competition, market, customer needs etc. It
comprises of two distinct approaches: the firm can either react to change – when the latter
cannot be anticipated – or adopt a pre-emptive attitude i.e. by setting flexible structures. Here, a
distinction can be made between the concept of adaptability and agility: adaptability is
applicable when a firm deals with a predictable change, whereas agility is concerned with
unforeseen changes in turbulent times. The reactive side of agility can be linked to the previous
characteristic of agility i.e. being flexible.
The second side Nirmal (2005) referred to is the proactive side. This side is driven by internal
forces such as the firm’s vision, values and people. Nirmal put a clear emphasis on the ability
of the top management to have clear insights about the future and to anticipate future trends; in
fact, foresight activities such as scenario planning enhance agility (Vecchiato, 2014). The
author also underlined the importance of having creative actions at all levels within the firm.
The proactive side accentuates the characteristic of agility to lead change and take advantage of
it. This view was also shared by Zhang & Sharifi (2000) (see Table 2.2). Dove (2001) gave a
definition of agility which also stresses the two sides – proactive and reactive: “an effective
integration of response ability and knowledge management in order to rapidly, efficiently and
accurately adapt to any unexpected (or unpredictable) change in both proactive and reactive
10
business/customer needs and opportunities, without compromising the cost or the quality of the
product/process”. Therefore, a firm is characterised as being agile when it responds to change
in both a reactive and proactive manner. A McKinsey global survey dated 2006 clearly defined
an organisation’s agility as “its ability to change tactics or direction quickly – that is, to
anticipate, adapt to and react decisively to events in the business environment”.
Most of the definitions of agility as presented in the previous section have converging
characteristics such as turbulent and changing environments, time, and responsiveness. For the
purpose of this research, the following definition of agility was converged upon:
Agility is the aptitude of a firm to react to change by organising in a flexible way at every
level, thereby enabling the firm to prepare for change and to lead change in unforeseen
circumstances.
This definition covers the occurrence of changes within the business environment and recalls
the notion of time through the concept of flexibility (it was previously discussed that a flexible
firm can cope with changes within the business environment with little time penalty). Through
this definition, the authors put a strong emphasis on the characteristic of being responsive to
market turbulences. If a firm is agile, it will use its flexibility to react to change more
effectively instead of suffering from it. But most importantly, agility is about preparing and
anticipating changes and leading these changes – predictable or not – at the advantage of the
firm to drive growth.
In the previous section, different definitions of firm agility were presented and a general
definition was suggested for the purpose of this research study. But what are the specific
attributes i.e. indicators of an agile firm? The following section will discuss the different agility
attributes found in the existing literature. These agility attributes are a way to indicate whether
a company can be considered as agile or not. As the literature offers a broad range of studies
that aim to measure agility and identify agility attributes, only the most commonly cited
attributes will be selected. It should be noted that the majority of the studies have attempted to
measure agility within the manufacturing industry; however, agility is “not industry specific”
(Erande and Verma, 2008) and therefore the agility attributes reported in the following section
are attributes that can be adapted or extended to organisations belonging to various industries.
The following figure, adapted from Tseng and Lin (2011), provides a better understanding of
how agility may be articulated throughout an organisation.
11
Agility Drivers
Firm’s
Agility
Agility Capabilities
Agility Attributes
(Adapted from Tseng and Lin, 2011)
As evident from Figure 2.1 above, agility attributes form the underlying structure of an agile
organisation (Ren, Yusuf and Burns, 2003). Agility attributes must be adopted by firms
attempting to cope with a turbulent business environment. In fact, the main driving force
behind agility is change (Yusuf et al., 1999; Tseng and Lin, 2011). Therefore, agility attributes
are required to deal with the agility drivers, i.e. changes in a business environment that pressure
companies to embrace an agile strategy (Tseng and Lin, 2011). Agility attributes also determine
the agility capabilities and behaviour of an enterprise; therefore, agility attributes can be
understood as leading to agile capabilities. The following sections will review the agility
attributes in detail and also introduce the agility capabilities and agility drivers.
Achieving agility requires the presence of various agility attributes at the organisational level.
One of the first scholars who proposed the idea of agility attributes was Yusuf et al. (1999). In
his article, Yusuf identified thirty-two different attributes that determine the entire behaviour of
a firm. For the scope of this research paper, eleven attributes were selected based on their
relevance to the research topic and on their occurrence within research papers.
12
Outsourcing and partnering activities
Goldman et al. (1995) presented “cooperation with other companies” as one of the main
dimensions of agility. Similarly, Yusuf et al. (1999) defined “rapid partnering formation” as an
attribute of firm agility. In their paper, the authors argued that in the agile paradigm,
cooperating with other organisations – even with competitors – is very important. According to
the authors, cooperation among enterprises can help acquire “missing links” within a
company’s main capabilities, through activities such as insourcing or alliances. Similarly,
Combs (1999) argued that collaboration with other organisations is a tool for exploration; when
firms are facing a turbulent and unstable business environment, accessing the knowledge of
other firms improves strategic decisions. Also, networking and collaboration improve
innovation and performance (Gulati and Sytch, 2007 as cited by Bock et al., 2012).
Reliance on partners
One should note that relying on partners for novel opportunities or access to information may
increase coordination problems and result in survival-based learning that reduces flexibility.
(Harrigan and Newman, 1990; Denrell, 2003, as cited by Bock et al., 2012). If flexibility is
reduced, it can be assumed that agility will be negatively affected. Therefore, not relying on
partners is an attribute of firm agility.
Efficient IT systems
Most of the authors who have attempted to measure agility, agree on the importance of having
efficient IT systems within the organisation (Overby et al., 2006; Sherehiy et al., 2007; Tseng
and Lin, 2011). The need for IT systems has risen from the incredible volume of information in
today’s global and competitive environment (Overby et al, 2006). IT systems can help process
this information much faster than humans and also provide organisations with the required
tools to monitor and understand changes in the business environment. One such tool is known
as the Knowledge Management System, which has been acknowledged as an IT system helping
organisations progress through turbulent and changing environments (Dove, 2001, Overby et
al., 2006; Sherehiy et al., 2007).
Iterative strategies
Working in an iterative and incremental way within agile organisations is a concept that derives
from Agile Software Development 5 (Kettunen, 2009). Abrahamsson (2002) identified Agile
Software Development as being incremental (small releases and rapid cycles). Lindvall et al.
(2002) recognised agile methods as being incremental and iterative. Similarly Fietz (2013)
suggested that working in an iterative manner is a way to help support organisational agility
Creative climate
Promoting a creative climate has become essential for agile innovation in turbulent business
environments. In fact, as the environmental turbulence increases, a creative culture facilitates
innovative solutions to competitive threats (Goodstein et al., 1996; Amabile, 2008).
5
More information in Appendix A
13
Fast and flexible mind-set
Being fast and flexible is at the core of the concept of agility. The promotion of a fast and
flexible mind-set helps build operational dexterity (Highsmith, 2012). According to Yusuf et al.
(1999), having flexible people in an organisation is an attribute of agility. Flexible thinking
encourages novel approaches to value creation within an organisation (Bock, 2012).
Efficient communication
Sherehiy et al. (2007) argued that clearly communicated internal information is of great
importance when thriving for firm agility. In the same paper, the authors also mention that
organisations facing a turbulent environment should adopt open communication and smooth
information flow. An efficient flow of information among organisational structures, people, and
system components was classified as an attribute of agility by Yusuf et al. (1999) and Tseng
and Lin (2011).
Decentralisation
In the traditional pyramid model, the flow of information is tightly controlled and decision
making is highly centralised, which prevents big mistakes from happening (Audran, 2011).
However, the pyramid organisation is slow to react i.e. the business is not agile. Agile
organisations advocate a highly decentralised decision making process, in order to increase
responsiveness in turbulent and uncertain environments. Hage and Dewar (1973) showed that
the decentralised organisation leads to higher innovation rates. A decentralised decision making
process is an attribute of agility (Yusuf et al., 1999).
Managing complexity
Simplification of operations is a process that decreases the functions or business units overseen
by management via consolidation, elimination or delegation (Bock et al., 2012). By reducing
design complexity of operations within an organisation, managerial attention can be devoted to
solving problems or identifying opportunities stemming from turbulent, changing and complex
business environments (Bock et al., 2012; Rothaermel et al., 2006; Ocasio, 2007).
Modular approach
Modularity enables an organisation to meet the customer's specifications by quickly modifying
parts of a product (Yusuf et al., 1999). As argued by Worren et al. (2002), a modular structure
of product and service elements has been identified to increase flexibility, which implies a
higher rate of responsiveness to turbulent environments. Modular approaches help
organisations to be globally efficient by using standardised components in areas such as
product development and manufacturing. Modular structures are created in order to rapidly
innovate and to adjust to changing market needs through reduced coordination costs (Doz and
Kosonen, 2009).
As argued previously, firm agility is concerned with an unpredictable, turbulent and complex
business environment. It is critical for a firm to ensure that the attributes of firm agility can
14
satisfy the agility capabilities – and further cope with the agility drivers (Tseng and Lin, 2011).
The literature review led to various agility capabilities. The most commonly cited were the
following three capabilities.
Flexibility
Flexibility is the ability of a firm to implement different processes and achieve multiple goals
whilst keeping the facilities, resources and systems unchanged (Tseng and Lin, 2011). Vokurka
and Fliedner (1998) referred to flexibility as “the capability of an organisation to move from
one task to another quickly and as a routine procedure, with each situation defined ahead of
time so that the procedures needed to manage it are in place”. Christopher et al. (2001)
extended the notion of flexibility from the manufacturing context to a wider business context.
The author encompassed organisational structures, logistics processes and in particular mind-
sets as enablers of flexibility within a firm. Flexibility is often argued as being a pre-requisite to
firm agility: a firm needs to first organise in a flexible way to manage change under known
conditions before it can respond to unforeseen circumstances (Wadhwa, 2003).
Responsiveness
Responsiveness is the ability of a firm to respond to turbulent environments and most
importantly to identify the upcoming changes in uncertain times (Tseng and Lin, 2011). This
capability complements the capability of flexibility. In fact, it was earlier defined that
flexibility is used to cope with known changes; responsiveness on the other hand is more about
identifying future unforeseen changes and coping with them. A Knowledge Management
System is often used within organisations to improve organisational knowledge and monitor
changes within the business environment (Overby et al, 2005).
Intuitively, it can be argued that these three capabilities converge towards the notion of speed.
It was argued in part 2.2.1 that flexibility allows to cope with changes within the business
environment with little time penalty. Responsiveness was defined by Tseng and Lin (2011) as
the ability to respond quickly to change. Partnering and collaboration activities speed up the
learning process and the creation of knowledge. The notion of speed – or being quick – is
embedded in the primary definition of agility. Some authors defined speed as being a
capability, however as it was demonstrated that each capability encompasses the notion of
quickness, it will not be retained as a capability for the scope of this thesis.
15
2.2.2.3 Agility Drivers
As defined by Zhang and Sharifi (2001), agility drivers are “the pressures from the business
environment that necessitate a company to search for new ways of running its business to
maintain its competitive advantage”. Agility drivers would require a company to revise the
current company’s strategy, admit the need to become agile, and adopt an agility strategy
(Sherehiy et al., 2007). This part will review the different agility drivers found in the literature.
As the agility drivers are not the main focus of this research, they will be introduced briefly.
By summarising the literature on agility drivers, the most cited areas of change and turbulence
in the business environment can be categorised as follows: (1) Market forces - such as the
market structure, the demand, the market need, price consciousness etc.; (2) Industry
competition - defined in terms of competition environment, competitors’ responsiveness,
substitutes for products etc.; (3) Globalisation - which makes the business environment broader
and more complex; (4) Macroeconomic forces - which are unpredictable exogenous factors that
affect the economy directly or indirectly; (5) Technological advancements -represented by the
introduction of new technologies and their adoption; (6) Geopolitical issues - through
government policies pressures, legislation pressures etc.; and finally (7) Environmental issues -
caused by environmental protection pressures. (Yusuf et al., 1999; Sharifi and Zhang, 1999,
2001; Sharp et al., 1999; The IBM Innovation Study, 2006).
After reviewing numerous definitions of the concept of agility across time, three key
characteristics were identified: changing environment, time, and responsiveness. Based on
these characteristics, firm agility was defined as follow:
Agility is the aptitude of a firm to react to change by organising in a flexible way at every
level, thereby enabling the firm to prepare for change and to lead change in unforeseen
circumstances.
Then, the different attributes that characterise agility at the firm level were defined. These
attributes represent indicators of firm agility. They are:
Thereafter, the agility capabilities that stem from the agility attributes were presented. The
three agility capabilities identified are:
16
- Flexibility - Responsiveness
- Partnerships and Collaboration
Finally, the most cited agility drivers were presented. They are:
The information presented above will now be used to fill up the organisational agility
framework that was presented earlier. The completed framework is represented in figure 2.2.
Previously, it was stated that agility attributes determine the agility capabilities i.e. agility
attributes can be understood as leading to agile capabilities (Tseng and Lin, 2011). But which
attributes lead to what capability? Instinctively, it can be claimed that the attributes “Customer
collaboration”, “Partnering and outsourcing activities” and “Reliance on partners” will belong
to the capability “Partnerships and collaboration”. The capability “Responsiveness” was
defined as the ability to identify future changes and cope with it when unforeseen. A good
Knowledge Management System was identified as an IT tool used to increase responsiveness.
Adopting a modular approach towards products and/or services, as well as smooth
communication flow allow for greater responsiveness as well. Therefore, the attributes
“Efficient IT systems”, “Efficient communication” and “Modular approach” are expected to
belong to the capability “Responsiveness”. The remaining attributes “Creative climate”,
“Iterative strategies”, “Fast and flexible mind-set”, “Decentralisation” and “Managing
complexity” all present cultural values that are close to the natural values of start-ups (Audran,
2011). Most of the time, start-ups and young firms are classified as being highly flexible
(Knight, 2004). Therefore, it is expected that these attributes belong to the capability
“Flexibility”. These intuitions are summarised below in the form of propositions. These
propositions will in turn be tested in part 4 – Analysis.
- Efficient IT systems
Proposition 2 - Efficient communication Responsiveness
- Modular approach
- Creative climate
- Iterative strategies
Proposition 3 - Fast and flexible mind-sets Flexibility
- Decentralisation
- Managing complexity
17
Market forces Agility Drivers Technological
advancements
Industry
competition Geopolitical issues
Globalisation Environmental
issues
Macroeconomic
forces
Firm’s
Agility
Agility Capabilities
Agility Attributes
18
2.3 Innovation
Innovation is a complex activity (Trott, 2005) and therefore it is difficult to converge upon a
universal definition. As described by Schumpeter in his seminal work (1942), innovation
encompasses the creation and introduction of a new product/service, new or improved
production processes, materials and intermediate inputs, and management methods
(Schumpeter, 1942).
For the scope of this thesis, the definition by Hartley (2006) will be used. Hartley suggests that
innovation is the successful development, implementation and use of new or structurally
improved products, services, processes or business models (Hartley, 2006). This definition is
interesting as it covers the dimension of novelty (which can be associated with radical
innovation), the concept of improvement (which can be associated with incremental
innovation), emphasises that innovation is a process (development, implementation and use),
and most importantly it highlights the three types of innovation that will be addressed in this
study: Product Innovation, Process Innovation, and Business Model Innovation.
Product Innovation (PDI) can be defined as new or significantly improved goods and services,
which generate value for the final consumer. Product innovation is strategically important to
industrial firms. The design of a new product is not an isolated activity. Apart from the product
design, plans have to be drawn up for the manufacturing process, the layout of the factory, the
distribution, the sale, as well as the whole production and sales unit. While it can be complex
and costly, product innovation is also a major source of future income and competitive
advantage for companies (e.g. Brown and Eisenhardt, 1995; Clark and Fujimoto, 1991). A
firm’s ability to generate a continuous stream of product innovations may be more important
than ever in allowing a firm to improve business performance, because of increasing levels of
competition and decreasing product life cycles. Therefore, a major concern for product
innovation managers is managing the complexity of product innovation, whilst keeping in mind
the link between product innovation and business performance.
Process Innovation (PRI) is concerned with identifying new and more effective internal
operations (Cohen and Levin (1989) as cited by Martinez and Labeaga, 2009), and also with
introducing new elements into an organisation's operations such as input materials, task
specifications, work and information flow mechanisms, and equipment used to produce a
product or render a service (Afuah, 1998). These new materials, tools, task specifications, etc.
help reduce production costs, which in turn help strengthen the firm’s competitive advantage
(Freeman (1987) as cited by Martinez and Labeaga, 2009).
19
2.3.3 Business Model Innovation
The notion of a Business Model (BM) can be understood as being the mechanism and the
structural design by which a firm creates margins and/or growth. Osterwalder and Pigneur
(2010) define BM as “the rationale of how an organisation creates, delivers and captures
value”. However, due to the competitive environment in constant change, business models are
not static; rather they need to be adapted and strengthened (Ganguly, Euchner 2014).
Business Model Innovation (BMI) has been defined in various ways. It is defined by Markides
(2006) as “the discovery of a fundamental different business-model in an existing business”. It
is concerned with the search of new logics and new ways to create and capture value
(Casadeus-Masanell and Zhu, 2012). BMI also refers to the situation when a firm adopts a
novel approach to commercialising its underlying assets (Gambardella and McGahan, 2010).
BMI is about changing the current business model; however the extent to which the current
business model has to change in order to classify it as an innovation remains uncertain.
Markides (2006) argues that PDI and PRI solely do not lead to BMI. Markides claims that BMI
emphasises the creation of new value sources and new markets, by modifying existing systems.
Ganguly and Euchner (2014) add on the notion of creating new markets by stating that BMI is
any innovation that disrupts the competitive advantage of key competitors. Consequently, firms
can compete through their BMs (Casadesus-Masanell, 2007). Novel BMs may be a source of
disruption (Christensen, 1997), by changing the logic of entire industries and replacing the old
ways of doing things to become the standard for the next generation of entrepreneurs to beat
(Magretta, 2002). According to Chesbrough (2007), BMI may have more important strategic
implications than other forms of innovation. This view was sustained by a recent IBM global
CEO survey (2006), which revealed that organisations that have grown their operating margins
faster than their competitors were putting twice as much emphasis on BMI as the
underperformers.
Osterwalder and Pigneur (2010) defined BMs with the help of the business model canvas
(figure 2.3). It can be instinctively assessed that if a firm changes at least one of the key
elements of the business model canvas (i.e. “who”, “what”, “why”, “where”, “how” and “how
much”), and by doing so create a new market, generate a new value source, or disrupt the
competitive advantage of key competitors, then the firm has successfully engaged in BMI.
Zott and Amit (2012) suggested that managers can innovate a BM in three ways: by adding
new activities, linking activities in novel ways, and changing which parties perform an activity.
In other words, BMI consists of innovating the content (i.e. the nature of the activities), the
structure (i.e. linkages and sequencing of activities) or the governance (the
control/responsibility over an activity) of the activity system between a firm and its network.
20
Source: Alex Osterwalder and Yves Pigneur (2010)
Figure 2.3 Business model canvas
The Innovation Productivity definition was built starting from the definition of productivity –
the rate at which goods/services (outputs) are produced from a standard set of inputs. Outputs
were identified in terms of the number of innovations (i.e. number of PDIs; number of PRIs;
and number of BMIs) produced by a firm, during a certain time period. The inputs will be all
activities at the organisational level that may affect the innovation process. These activities
could include R&D effort, hiring innovation talents, organising to enhance a creative climate,
promotion of individual projects, investments etc. The inputs of Innovation Productivity are
beyond the scope of this study and will not be considered. Instead, the outputs and the
facilitating attributes of agility are the critical measures.
The following Figure 2.4 depicts the definition of Innovation Productivity that has been
formulated for this study. In this process, agility will be defined as a catalyst.
Z
(Output)
X+Y
Agility Number of PDIs
(Units of Input)
Number of PRIs
Number of BMIs
21
3 Methodology
The overall purpose of this research study was to test if there exists a relationship between a set
of defined agility attributes and innovation productivity. Innovation was broken down into
three dependent variables: Product Innovation (PDI), Process Innovation (PRI) and Business
Model Innovation (BMI). All these three dependent variables depict the ‘number’ of respective
innovations during the three years 2012 - 2014. This has helped define them as Product,
Process and Business Model Innovation productivity.
The literature review laid down the foundation of the theoretical framework by providing an
exploratory dimension through the review of academic articles, management journals, books,
and dissertations within the field of agility and the three types of innovation. The literature
review helped uncover the variables of agility that could have an effect on the innovation
process.
3.1.2 Strategy
As the research aimed at testing the existence of a relationship between the attributes of agility
and innovation productivity, the primary research was therefore quantitative in nature. The
research has assessed various organisations and multiple industries across the EU. The
explanatory or independent variables were the agility attributes employed by these
organisations, along with a number of control variables to make the study more reliable.
A quantitative study was opted for due to multiple reasons. Firstly, it allowed for a broader
study, with a greater number of observations as compared to a qualitative study. Second,
through a quantitative study it was possible to collect a large amount of data across numerous
cases, which can then help produce generalised results. Third, as the goal of this research work
was to measure the extent of the impact of agility attributes on innovation productivity, a
quantitative study was best suited for it (Bryman and Bell, 2011). Finally, a quantitative study
allowed for more objectivity in the results, as the findings are not based on the subjective
thoughts, opinions or interpretations of the respondent (Bryman and Bell, 2011).
• Accept/reject hypotheses
• Conclusions
Deliverable • Open up to new researches
23
3.2 Research Design
3.2.1 Field Study
The design of the primary research was aimed towards simulating a field study by sending out
questionnaires to individuals working with innovation all over the EU. A field study was opted
for because the study needs to observe the “real world” in order to assess the impact of agility
attributes on innovation productivity.
A cross-sectional design was chosen as the research aims at collecting a body of quantitative
data on numerous cases, at a single point in time, in order to detect a potential relationship
between the agility attributes and innovation productivity.
In order to collect a body of quantitative data on numerous cases, a survey research method was
used. The research instrument that was opted for was a questionnaire. The answers were
collected more or less simultaneously (over a four-week period), which is a characteristic of the
cross-sectional design. However, as there is no time ordering with regards to the variables
(caused by the simultaneity of the collection of data), it might create a problem when
establishing the direction of the causal relationship (Bryman and Bell, 2011). Consequently,
careful attention is required during the interpretation of the results. It is possible to draw certain
inferences regarding causality, however, cross-sectional design lacks internal validity and the
inferences could also lack the credibility that can be found in most experimental studies
(Bryman and Bell, 2011).
Table 3.1 depicts the structure of the cross-sectional design. Such a design includes the
collection of a body of quantitative data on many variables (Variable 1 ; Variable 2, Variable 3 …
Variable n ), at a single point in time whilst considering multiple cases. For each observation (in
this study each observation will be one organisation), data is available for all variables. All
these variables will be collected at time T 1 , through a questionnaire – the research instrument.
For the purpose of this study, a regression analysis was chosen as a quantitative research
method, in order to analyse the data from the multiple variables. The regressions were used to
conduct a statistical analysis and thereby determine the nature of relationships, if any. Such an
24
analysis was required for each type of innovation in order to test the proposed hypotheses. The
details of the regressions and the acceptance or rejection of the hypotheses were examined in
Section 4 of this report.
In order to collect the data necessary for the research, a survey methodology was formulated.
Generally, surveys can be divided into two broad categories: interviews and questionnaires.
One of the main objectives of this research was to collect as many observations as possible in
the given time frame for this study, in order to build statistical inferences. For this reason, the
choice to use a questionnaire as the instrument to collect data seemed to be the optimal course
of action. The questionnaire took the shape of an online form and was sent out via e-mail in
order to garner maximum reach.
Although an online questionnaire will assist in reaching out to more respondents in the limited
timeframe, several issues may arise and these potential issues should be considered during the
design of the questionnaire. One such issue is related to the formulation of questions. As the
contact with the potential respondents will be initiated indirectly through e-mail, it will be
almost impossible to verify if the questions were understood and interpreted in the right
manner. This in turn could lead to biased answers and affect the results of the research.
Therefore, the questions need to be designed in a standard way, using standard definitions.
Another potential issue is concerning the collection of data. The design of the questionnaire
needed to ensure that data regarding all questions can be quantified across the defined
variables. Finally, conveying the usefulness of this research to the potential respondents and
gaining their interest will be a major challenge. Therefore, it is essential to incorporate a short
argumentation communicating the value of the defined research area to not only the given field,
but to the respondents as well. A detailed description of the questionnaire design can be found
in Section 3.4.2.
3.3.2 Questionnaire
The question addressed in this research is: To what extent can agility improve innovation
productivity inside large organisations? The data required in this study can be broken down into
three sets of variables. The questionnaire was designed in such way that it was possible to
collect data on each of the three sets of variables.
Dependent variables: These variables correspond to the three types of innovations, in terms of
their productivity.
Independent variables: These variables correspond to the agility attributes. These attributes can
be summarised under four categories, as done in the literature review. However, in section 3.4
25
(Data Collection), these categories will be broken down into the eleven attributes of agility
contained in these four categories.
Control variables: These variables correspond to the factors that may have a confounding
effect on the results, and their inclusion is critical in order to measure the unbiased effect of the
given independent variables on the dependent variables. The controls usually take into account
an observation’s characteristic and environment.
In order to analyse the collected data, a regression analysis method was used. An empirical
analysis was carried out in order to estimate the relationship to be tested (i.e. possible
relationship between agility and innovation productivity).
The models that were tested took the following generic forms:
The above econometric models represent the generic equations that will be used in order to test
the effect of different agile attributes on the productivity of product, process and business
model innovations. The β values are the coefficients of the independent and control variables
that will reflect the change in the dependent variable given a unit change in the independent
variable. Different models will be created using different variables, along with any interaction
variables if relevant.
As the dependent variables can contain any whole number value from 0 to infinity, depending
on the number of innovations listed, they are count variables and a Poisson or Negative
Binomial regression model will be used. In these equations, the agility attributes are the
independent variables or predictors. Depending on the nature of the variables, they will be
expressed as ranked values, dummy variables or numerical variables. Finally, the regressions
also include control variables in order to ensure that any potential bias in study has been
accounted for.
In order to build these regression models and conduct the subsequent data analysis, the
statistical tool STATA was used. Using STATA made it possible to run and test the model.
A number of measures had to be considered to ensure that the model was not biased. Firstly,
through exploratory data analysis, the presence of outliers was controlled and it was assessed
whether the outliers were to be kept as part of the analysis or not. They were not deleted if they
were not considered as typos or if the value was extremely abnormal; it was however important
to be aware of their presence in the data set. Second, multicollinearity was also accounted for in
26
order to ensure that the effect of a given variable was not overestimated. The problem of
multicollinearity may arise if two variables having a strongly similar effect are both used as
independent variables in the regression equation.
Once the regressions were performed, the attention was put on the β values, their statistical
significance, the number of observations taken into account, and the robustness check to
account for the good fit of the model. Finally, it was possible to determine if there exists a
relationship and if so what the nature of it is. The data analysis can be found in Section 4.
This part will address the sampling methodology formulated, in order to obtain a representative
sample of the population under observation. In order to define a sample that ought to match the
requirements of the study, certain filters were applied whilst selecting the respondents. It was
defined that the respondents must work for an organisation with more than 250 employees, in
the EU, and occupy an influential position in an innovation role.
Given the limited timeframe to conduct this study, the population under observation
corresponded to members belonging to the three LinkedIn groups: Innovation Management
Group, Innovation Excellence Group, and Open Innovation Network Group. In total, these
groups comprise of more than 75,000 members. However, this number should be interpreted
carefully as some members may be present in more than one, or all three groups. Also, not all
of the members were working with innovation in the European Union. After attaining
membership of these groups, it was possible to contact the members through a personalised e-
mail.
The members that were considered for this research comprised of Innovation Managers,
Innovation Directors, Heads of Innovation, and Vice President Innovation. The main reason
behind reaching these type of positions was to increase the reliability of the answers. In fact, it
was assumed that high positions were more inclined towards knowing the overall activities of
their respective organisations. Other positions were also considered, but to a smaller extent.
The selected members represented their respective organisations while responding to the
questionnaire. Therefore, the chosen unit of analysis was the organisation, even though the
observation was collected on an individual level.
27
3.4.2 Questionnaire Design
In order to carry out this quantitative research, a questionnaire was created through the online
tool Qualtrics, with closed-ended questions to quantitatively assess the responses of the
observed sample. Some definitions and questions were adapted from the Community
Innovation Survey (CIS) 2010; it was used as a guideline to formulate questions pertaining to
product innovation and process innovation. The definitions of these two types of innovations
were also incorporated in the survey for this study as they were deemed to be representative of
the broad industry wide understanding of these terms. Other definitions stem from the literature
review that was conducted before.
The average time to complete the questionnaire was estimated to be 10 minutes. It consisted of
both the core research questions and the control questions. The full questionnaire can be found
in Appendix C. The following section builds on the introduction to the questionnaire presented
in the research methods Section 3.3 by providing greater details about each variable.
3.4.2.1 Variables
a) Dependent Variables
Business Model Innovation Productivity: Business model innovation was defined as the
creation of new or significantly improved value offering, by taking into account the customers,
28
infrastructure, and revenue model. The respondents were then asked if their organisation
introduced any new or significantly improved value offering to customers; new or significantly
improved customer relationships, customer segments, or distribution channels; new or
significantly improved key partners, key activities, or key resources; and new or significantly
improved cost structures or revenue streams - during the three years 2012-2014. These four
were contingency questions, and if the respondents answered yes to any of them, they were
required to answer a follow up question. This follow up question asked them how many times
their organisation had re-invented its business model from 2012 to 2014. The respondents were
given a range of numerical choices to answer this question.
b) Independent Variables
Agility and Organisational Behaviour: The respondents were asked to rate a number of agility
attributes on a 7-point likert scale, where 1 meant that the factor had no presence within the
organisation and 7 meant that it had an extremely strong presence within the organisation. 4
meant the factor had a moderate presence. According to the literature review, the agility
attributes incorporated in the questionnaire included the following:
These independent variables were then used to create a regression model in order to assess the
impact on the dependent variables and by doing so, test the proposed hypotheses.
c) Initial Controls
A number of control questions were included in the questionnaire that took into account
variables relative to location of the organisation’s headquarters, geographic markets the
organisation is involved in, type of industry, size of the organisation, exogenous factors, and
technological integration with business processes.
Location of Headquarters: This variable was used in order to ascertain whether the
headquarters of the organisation were located within the EU. A binary (dummy) variable was
introduced if the organisation’s headquarter was inside the EU. Organisations in the EU operate
29
in a common market but with socio-culturally diverse facilities, with the potential to affect
innovation and change (Bock et al., 2012).
Geographic Markets: The geographic markets in which the organisation is involved in were
used to determine the extent of the organisation’s operations and control for the scale of their
products or services. Four separate choices were introduced, which represented an increasing
subset of each other. All four choices were binary (dummy) variables representing a particular
type of geographic market.
Type of Industry: As the research is a cross-industry research, this variable was used to identify
the respective industry of an organisation and control for the effect it may have on the number
and kind of products/services. Multiple choices were included along with an ‘Other’ option. All
industry choices were introduced as binary (dummy) variables representing a particular type of
industry. Manufacturing, Fast Moving Consumer Goods, and Luxury organisations were
grouped under the Manufacturing industry variable. In addition, service, banking/financial and
consulting organisations were grouped under the Service industry variable.
Size of the Organisation: The variable ‘size of the organisation’ was used to control for the
minimum number of employees required to deem an organisation as being ‘large’. The
organisation was considered ‘large’ if and only if it had equal to or greater than 250 employees,
according to the initial definition of a large organisation. In case the employees were stated to
be less than 250, the responses were not considered suitable for the study. The intuition behind
including this variable was the fact that size of the organisation may affect innovation efforts
(Damanpour, 1992), and therefore the size of the organisation was aggregated into six
categories. The categories were: less than 250 employees, between 251 and 500 employees,
between 501 and 1,000 employees, between 1,001 and 5,000 employees, between 5,001 and
10,000 employees and greater than 10,000 employees.
Exogenous Threats: The variable concerning potential exogenous threats to the organisation
was controlled for in order to assess whether the organisation’s innovation activities were
affected by the perceived macroclimate of the future. All choices were introduced as binary
(dummy) variables representing a particular type of exogenous factor. The most important
exogenous factors for this study were considered to be technological advancements and
macroeconomic factors.
d) Core Controls
Innovation Objectives: In order to control for the reasons behind an organisation’s intention to
engage in product, process, or business model innovation; a set of questions regarding each
category were introduced using a 7-point likert scale. Under each category, the respondents
were asked how important were each of the given objectives for their organisation’s
30
product/service, process, business model, or organisational innovations introduced during the
three years 2012 to 2014. With respect to the 7-point likert scale, 1 meant that the objective was
not important at all and 7 meant that the objective was of critical importance. 4 meant that the
objective held average importance. The reason for introducing these core controls was to
account for the confounding effect of these innovation objectives.
The following table summarises all three sets of variables:
Dependent
Independent Variables Control Variables
Variables
The procedure for selecting the respondents, designing the variables to be measured, and
administrating the questionnaire has been detailed in the previous sections. Similarly, in section
4, the process used to carry out the data analysis will be explained in detail as well. This
process will ensure that this study is replicable. It is important to ascertain the replicability of
this research, as it is a pre-requisite to ensure the reliability of the study.
31
3.5.2 Reliability
In this study, reliability is concerned with the consistency of the measures of agility (Bryman
and Bell, 2011). As the attributes of agility were taken from existing research during the
literature review, it can be argued that the stability of the measures was accounted for.
Similarly, the different definitions of innovation were borrowed from the Community
Innovation Survey which therefore ensure the reliability of this concept
Internal reliability, on the other hand, will be tested in Section 4 through the use of Cronbach’s
Alpha after the Principal Component Analysis.
3.5.3 Validity
Validity is concerned with the integrity of the measures employed in the research. The
attributes of agility were borrowed from the literature and have already been validated in terms
of measuring agility. Therefore, content validity was accounted for and all of the attributes of
agility are considered to be valid.
On the other hand, bearing in mind the cross sectional design of the research, internal validity
might be impeded. This is mainly due to the fact that there is no time ordering with respect to
the variables and also, there is no manipulation present. In case a relationship is depicted, the
confidence for drawing a causal relationship and statistical inferences will generally lack the
credibility of the findings derived from an experimental design.
External validity can be questioned due to the non-probability sampling method. However, the
consecutive sampling technique employed strengthens this criterion. Nevertheless, it will be
necessary to take into account the non-randomisation of the sampling while interpreting the
results. This also suggests that any generalisations should be treated carefully.
3.6 Execution
3.6.1 Survey Administration
Before starting the survey, the questionnaire was tested on two individuals in order to assess
whether the questions were understood in the right context and to calculate the average time
needed to complete the questionnaire. The questionnaire was deemed comprehensible and the
average time to complete the questionnaire was determined to be 10 minutes. Following this,
the survey was initiated.
317 questionnaires were sent out to Innovation Managers, Innovation Leaders, Innovation
Directors, and Heads of Innovation, working for organisations within the European Union.
These 317 questionnaires corresponded to all possible observations that matched the control
criteria and could be associated with the sampling population (the sampling methodology can
be found in section 3.4.1). In order to create awareness, arouse interest and increase the
response rate, all correspondence was conducted on an individual level through personalised e-
32
mails sent out to the filtered members of the three LinkedIn groups. After the initial round of
emails in early March, follow up e-mails were sent out after ten days to those individuals who
did not respond to the initial correspondence. This strategy helped increase the response rate by
over 100%. The personalised e-mails can be found in Appendix B.
Table 3.3 shows the practical details of the survey methodology and the survey administration
results.
Figure 3.2 shows the geographic locations of the organisations that were approached. Where it
was not possible to determine the geographic location of an organisation through the target
profile, it was categorised under “Not Available”.
70
60
50
40
30
20
10
33
3.6.2 Final Sample
Out of a possible 317 responses, 103 responses were gathered over a period of four weeks. This
suggests a response rate of 32.49%. The base line target to ensure the reliability of the study
was set at minimum of 40 responses. This target was surpassed by approximately 158%, which
accounted for a significant increase in the reliability of the study. Figure 3.3 shows the
geographic locations of the organisations that responded to the survey questionnaire. Where the
geographic location of an organisation was not specified, it was categorised under “Not
Specified”.
18
16
14
12
10
8
6
4
2
0
Figure 3.3 illustrates that the most number of respondents belong to the UK, France, Sweden
and Netherlands. Out of these four countries, Sweden had the highest response rate of 60%. A
cursory look at both figure 3.2 and 3.3 suggests that Sweden may represent a response bias, as
the response rate in Sweden is significantly higher than the other countries. It may be the case
that Sweden is overweight in the final sample, but as the number of responses from Sweden
account for only 12% of the total responses, the effect can be considered to be insignificant.
Out of the 103 responses collected, 28.2% were completed by Directors of Innovation, Strategy
or R&D, 20.4% were completed by Heads of Innovation, 24.3% were completed by Managers
in Innovation, 7.8% were by Innovation Vice Presidents, and 19.4% were completed by other
positions within innovation or non-specified positions. Figure 3.4 gives a graphical
representation of this distribution.
34
Figure 3.4 Distribution of respondents across positions
Over 50% of the respondents occupy very high position within innovation (VP, Director or
Head of Innovation). This ensures the reliability of the data collected as it was assumed that
people occupying these types of position have a better understanding and overall view of the
activities conducted within their respective organisations. Other positions related in general to
positions within R&D or Marketing.
35
4 Data Analysis
In total, 19 observations with less than 250 employees were present in the dataset and were thus
deleted from the dataset. This reduced the number of valid observations from 103 to 84. In
addition, two further observations were dropped from the dataset. One of these observations
contained missing values across all variables and the other observation was assumed to be a
case of wrong data entry due to the presence of an abnormal outlier.
4.2 Variables
4.2.1 Variable Names and Description
Table 4.1 presents all the variables used in this study along with their descriptions:
36
AND Ind_MANU Firms from Manufacturing Industry (=1 if yes, 0 if not)
EXOGENOUS
Ind_SERV Firms from Service Industry (=1 if yes, 0 if not)
CONTROLS
Ind_Tech Firms from Technology Industry (=1 if yes, 0 if not)
Org_Size Size of the organisation
Importance of technological integration to the firm (1 if no value to business,
Tech_Int
5 if very important)
Exo_Tech Technological advancements (1 if small impact on business, 5 if high)
Exo_Macro Macroeconomic factors (1 if small impact on business, 5 if high)
The descriptive statistics of the variables used in this study can be found in table 4.2. The
statistics show the number of observations, mean, standard deviation and minimum/maximum
values of the given variables. A cursory look at the means of Num_PDI, Num_PRI and
Num_BMI suggests that during 2012-2014, the number of product innovations were
considerably greater than the number of process or business model innovations.
37
Agility_CCOL 80 4.85 1.661934 1 7
Agility_MA 80 4.475 1.792919 1 7
Agility_IT 79 4.088608 1.594782 1 7
Agility_Comm 79 4.556962 1.516971 1 7
Agility_FM 79 4.582278 1.661075 1 7
EU 81 .8148148 .390868 0 1
Geo_EU 82 .6097561 .4908068 0 1
Ind_SERV 82 .4146341 .4956906 0 1
Ind_MANU 82 .5487805 .500677 0 1
Ind_Tech 82 .3536585 .481047 0 1
Org_Size 82 4.95122 1.430725 2 6
Exo_Macro 82 .4268293 .4976609 0 1
Exo_Tech 82 .7560976 .4320773 0 1
Tech_Int 82 3.865854 .8128929 2 5
PDI_CONT_1 81 5.45679 1.351382 1 7
PDI_CONT_2 81 5.666667 1.360147 1 7
PRI_CONT_1 81 5.407407 1.272574 2 7
PRI_CONT_2 81 5.049383 1.650009 1 7
PRI_CONT_3 81 4.901235 1.585914 1 7
PRI_CONT_4 81 5.234568 1.217288 1 7
PRI_CONT_5 81 5.222222 1.431782 2 7
BMI_CONT_1 80 5.1875 1.567823 1 7
BMI_CONT_2 80 5.2625 1.490423 1 7
BMI_CONT_3 79 3.658228 1.745956 1 7
In addition to the descriptive statistics, it was also interesting to see the existing correlation
trend between the agility attributes. Table 4.3 below depicts that most of the variables are
correlated with each other to a certain extent, however, the correlation is not high. The only
exception seems to be Agility_PR and Agility_OP with a correlation of 0.8068. All other
variables possess a correlation of less than 0.7.
38
4.3 Principal Component Analysis
4.3.1 Purpose
For the purpose of the analysis, the 11 agility attributes (independent variables) needed to be
reduced to a smaller number of composite variables due to two main reasons:
i) The total number of observations (N = 82) were not considered sufficient enough to
depict the individual unbiased impact of each variable.
ii) The correlated observed variables could potentially lead to multicollinearity and
thereby give biased results.
In order to achieve this, two methods were considered i.e. Principal Component Analysis
(PCA) that would give the principal component factors, and Factor Analysis (FA) that would
give the iterated principal factors. As the aim of this analysis was to only reduce the
independent variables into a smaller set of composite variables, the PCA was preferred. The FA
would have been preferred if the aim was to test a theoretical model of latent factors causing
observed variables.
4.3.2 Analysis
The command for the PCA was run in STATA in order to extract the principal component
factors using the eigenvalues. As a default rule of thumb derived using the Kaiser criterion,
eigenvalues above 1 were used to create the composite factors (Kaiser, 1960). Table 4.4 shows
the eigenvalues as well as the factor loadings acquired in STATA. The variables displayed in
table 4.5 represent the independent variables i.e. the agility attributes.
39
Factor loadings (pattern matrix) and unique variances
-----------------------------------------------------------
Variable Factor1 Factor2 Factor3 Uniqueness
-------------+ ------------------------------+--------------
Agility_CC 0.8224 -0.1938 -0.2008 0.2458
Agility_MC 0.6927 -0.1603 -0.1046 0.4836
Agility_DM 0.6933 -0.1811 -0.5318 0.2038
Agility_IS 0.7391 -0.0515 -0.3600 0.3215
Agility_OP 0.5548 0.7583 -0.0378 0.1159
Agility_PR 0.7159 0.6093 0.0720 0.1110
Agility_CCOL 0.7471 0.1865 0.1305 0.3901
Agility_MA 0.6958 0.0018 0.3833 0.3689
Agility_IT 0.5597 -0.2348 0.6346 0.2289
Agility_Comm 0.7438 -0.3526 0.1882 0.2871
Agility_FM 0.8079 -0.2101 -0.0218 0.3027
-----------------------------------------------------------
The three eigenvalues highlighted in red in table 4.4 represent the optimal number of factors or
components that can be generated using the 11 independent variables. Figure 4.1 gives a visual
representation. It is important to note that even though the final sample included 82 valid
observations, there are some missing values that reduce the number of observations used in the
factor analysis to 78 (see table 4.4).
The factor loadings were then given an orthogonal varimax rotation (Kaiser on) in order to
ensure that the factors are not correlated with each other and to also assign all rows the same
weight (Horst, 1965).
40
Rotated factor loadings (pattern matrix) and unique variances
-----------------------------------------------------------
Variable Factor1 Factor2 Factor3 Uniqueness
-------------+------------------------------+--------------
Agility_CC 0.7687 0.3496 0.2458
Agility_MC 0.6061 0.3438 0.4836
Agility_DM 0.8809 0.2038
Agility_IS 0.7614 0.3215
Agility_OP 0.9270 0.1159
Agility_PR 0.8743 0.1110
Agility_CCOL 0.3789 0.4464 0.5168 0.3901
Agility_MA 0.6742 0.3394 0.3689
Agility_IT 0.8715 0.2289
Agility_Comm 0.5243 0.6606 0.2871
Agility_FM 0.6521 0.4864 0.3027
-----------------------------------------------------------
(Blanks represent abs(loading)<.3)
The rotated factor loadings that represented a correlation of < 0.3 with the generated factors
were then removed in order to present a less noisy table. Table 4.6 shows the rotated factor
loadings i.e. the correlation between the independent variables and the principal component
factors.
All variables show a low uniqueness of < 0.5, which suggests that all variables belong to at
least one of the factors uniquely and the relevance of the variables is higher for a given factor
model. A very high uniqueness would have suggested that a given variable does not uniquely
belong to any of the factors listed.
Table 4.7 shows how the 11 independent variables are classified into the three factor loadings,
in accordance with Table 4.6. After conducting the literature review, three propositions were
formulated on how the agility attributes can be classified under the agility capabilities (see
Figure 2.2). The factor analysis confirmed the intuition formulated through these propositions.
In fact, all three factors contain the same set of agility attributes as proposed in Figure 2.2.
Therefore, the composite factors containing the independent variables were named as
Flexibility, Responsiveness, and Partnerships and Collaboration based on the understanding
gained through the literature review. These three composite factors, consequently, are identified
as capabilities of agility containing the specific agility attributes.
41
- Modular Approach (Agility_MA)
RESPONSIVENESS
- Efficient IT System (Agility_IT)
(Factor 2)
- Efficient Communication (Agility_Comm)
- Reliance on Partners (Agility_PR)
PARTNERSHIPS
AND
- Customer Collaboration (Agility_CCOL)
COLLABORATION
(Factor 3) - Outsourcing and Partnering Activities (Agility_OP)
The three factors Flexibility, Responsiveness and Partnerships and Collaboration were then
used in three independent regression models as independent variables to assess the impact on
the dependent variables Num_PDI, Num_PRI and Num_BMI.
In order to assess the reliability of the factors formed using the PCA, the Cronbach’s alpha test
was run in STATA to check the internal consistency of the factors (Nunally, 1978). The
following table shows the results:
A Cronbach’s alpha value of > 0.7 depicts strict internal consistency within the factors
(Nunally, 1978), whereas a value of > 0.6 is considered to be the composite reliability
coefficient (Bagozzi and Yi, 1988). In the table above, Flexibility and Partnerships and
Collaboration have a value of > 0.7, thereby depicting strict internal consistency.
Responsiveness on the other hand has an alpha value of < 0.7, however, the value is > 0.6 and
the factor can therefore be considered as reliable.
42
4.4 Regression Analysis
4.4.1 Poisson Regression vs. Negative Binomial Regression
Before conducting the respective regressions, there was a need to identify whether a Poisson
regression was more appropriate or a Negative Binomial regression. In order to ascertain this,
the descriptive statistics of Num_PDI, Num_PRI and Num_BMI were analysed in order to
compare the means and the variances of the dependent variables. Table X shows the results.
The rule of thumb suggests that if the Mean < Variance, the data is over dispersed and a
Negative Binomial regression is the appropriate choice as the Poisson assumption is violated.
The table above shows that in all three cases the Mean < Variance, therefore indicating that a
Negative Binomial regression is the optimal choice in all three cases. Another method to
choose the right model is to look at the Pearson goodness‐of‐fit test results. If the Prob > chi2
value lies below the threshold of 0.05, it is more appropriate to use a Negative Binomial
regression. The following table shows the Pearson goodness-of-fit test results, all of which
indicated a Prob > chi2 value of < 0.005.
Poisson = Num_PDI
Pearson goodness-of-fit = 2669.014
Prob > chi2(74) = 0.0000
Poisson = Num_PRI
Pearson goodness-of-fit = 740.6172
Poisson = Num_BMI
Pearson goodness-of-fit = 174.9379
43
4.4.2 Hypotheses
The factors created using the PCA were then used to generate the hypotheses to be tested using
the Negative Binomial regression models. As the agility attributes needed to be grouped before
proceeding with the regression analysis, the hypotheses were formulated after the PCA. The
following hypotheses will be tested:
4.4.3 Analysis
Three independent regressions were run for the three dependent variables Num_PDI, Num_PRI
and Num_BMI using the three composite factors of Flexibility, Responsiveness, and
Partner_Collab as independent variables.
All regressions also included controls such as EU, Geo_EU, Ind_MANU, Ind_SERV,
Ind_Tech, Org_Size, Tech_Int, Exo_Tech, and Exo_Macro.
In addition, the regression concerning Num_PDI included the controls PDI_CONT_1 and
PDI_CONT_2; the regression concerning Num_PRI included the controls PRI_CONT_1,
PRI_CONT_2, PRI_CONT_3, PRI_CONT_4, and PRI_CONT_5; and the regression
concerning Num_BMI included the controls BMI_CONT_1, BMI_CONT_2, and
BMI_CONT_3.
44
Product Innovation Process Innovation Business Model
Productivity Productivity Innovation Productivity
Flexibility
Independent Variables
Responsiveness
(Composite Factors)
Partner_Collab
EU
Geo_EU
Ind_MANU
Ind_SERV
Common Control Ind_Tech
Variables Org_Size,
Tech_Int
Exo_Tech,
Exo_Macro.
PRI_CONT_1 BMI_CONT_1
Additional Controls PDI_CONT_1 PRI_CONT_2
PRI_CONT_3 BMI_CONT_2
PDI_CONT_2 PRI_CONT_4
PRI_CONT_5 BMI_CONT_3
Table 4.11 Three independent regressions and their respective independent variables
The first model included the independent variables of Flexibility, Responsiveness and
Partner_Collab. The second model further included the controls EU, Geo_EU, Ind_MANU,
Ind_SERV, and Ind_Tech. The third model further included Org_Size, Tech_Int, Exo_Tech,
and Exo_Macro. Lastly, the fourth model included the specific PDI, PRI and BMI controls as
well.
By creating four separate models and running step-wise regressions, it was possible to see the
impact on the coefficients and the significance levels as more variables were added to the
equations.
45
4.4.3.1 Product Innovation Productivity
The results for the Negative Binomial regression for Num_PDI can be found in Table 4.12.
In order to assess the impact of the independent variables on Num_PDI, model (4) was used as
it gives the most unbiased results due the presence of relevant controls. The coefficients of the
three independent variables can be interpreted as follows:
• For a one unit change in Flexibility, the difference in the logs of expected counts of
Num_PDI is expected to change by 0.148, given all other variables in the model are
held constant. However the result is insignificant as the p-value depicts a value higher
than 0.05.
• For a one unit change in Responsiveness, the difference in the logs of expected counts
of Num_PDI is expected to change by 0.300, given all other variables in the model are
held constant. The result is significant as the p-value depicts a value lower than 0.05.
• For a one unit change in Partner_Collab, the difference in the logs of expected counts of
Num_PDI is expected to change by 0.341, given all other variables in the model are
held constant. The result is highly significant as the p-value depicts a value lower than
0.01.
46
(1) (2) (3) (4)
Num_PDI Num_PDI Num_PDI Num_PDI
Num_PDI
Flexibility -0.0642 -0.0675 0.0619 0.148
(-0.50) (-0.55) (0.50) (1.12)
PDI_CONT_1 -0.130
(-1.43)
PDI_CONT_2 -0.0408
(-0.45)
47
As the coefficients given in the Negative Binomial regression are interpreted as the difference
between the log of expected counts, the incident rate ratios must be computed in order to assess
the impact in terms of percentages. The following table shows the values of the coefficients in
incident rate ratio form with respect to model (4).
(4)
Num_PDI
Num_PDI
Flexibility 1.160
(1.12)
Responsiveness 1.350*
(2.33)
Partner_Collab 1.406**
(2.75)
In addition to the independent variables above, the control variable Exo_Macro gives a
significant result and suggests that exogenous macroeconomic factors may have a significant
impact on product innovation productivity. The result suggests a positive relationship.
In order to test for the robustness and goodness of fit of the four models, the Akaike
Information Criterion (AIC) values were used. The AIC values were multiplied with the
number of observations N. As a rule of thumb, a lower AIC*N value represents a better
goodness of fit of the model.
The regression table depicts these values for the four models independently with model (1)
having the highest value of 561.0, and model (3) and (4) having the lowest value of 535.4 and
536.6 respectively. Therefore, the decision to choose model (4) as the optimal model is a
reasonable one.
48
4.4.3.2 Process Innovation Productivity
The results for the Negative Binomial regression for Num_PRI can be found in Table 4.13.
In order to assess the impact of the independent variables on Num_PRI, model (4) was used as
it gives the most unbiased results due the presence of relevant controls. The coefficients of the
three independent variables can be interpreted as follows:
• For a one unit change in Flexibility, the difference in the logs of expected counts of
Num_PRI is expected to change by 0.213, given all other variables in the model are
held constant. However the result is insignificant as the p-value depicts a value higher
than 0.05.
• For a one unit change in Responsiveness, the difference in the logs of expected counts
of Num_PRI is expected to change by 0.363, given all other variables in the model are
held constant. The result is highly significant as the p-value depicts a value lower than
0.01.
• For a one unit change in Partner_Collab, the difference in the logs of expected counts of
Num_PRI is expected to change by 0.235, given all other variables in the model are
held constant. The result is significant as the p-value depicts a value lower than 0.05.
49
(1) (2) (3) (4)
PRI_CONT_1 -0.215*
(-2.03)
PRI_CONT_2 -0.302***
(-3.47)
PRI_CONT_3 0.281**
(2.99)
PRI_CONT_4 -0.218*
(-2.27)
PRI_CONT_5 0.292**
(2.68)
50
As the coefficients given in the Negative Binomial regression are interpreted as the difference
between the log of expected counts, the incident rate ratios must be computed in order to assess
the impact in terms of percentages. The following table shows the values of the coefficients in
incident rate ratio form with respect to model (4):
(4)
Num_PRI
Num_PRI
Flexibility 1.237
(1.92)
Responsiveness 1.438**
(2.88)
Partner_Collab 1.265*
(2.00)
Apart from the three independent variables above, the controls of Org_Size, Exo_Macro,
PRI_CONT_1, PRI_CONT_2, PRI_CONT_3, PRI_CONT_4, and PRI_CONT_5 are also
significant. This suggests that these variables have a significant impact on process innovation
productivity and it was important to take their impact into account.
In order to test for the goodness of fit of the four models, the AIC*N values were used from the
regression table. The regression table depicts these values for the four models independently
with model (1) having the highest value of 424.2, and model (4) having the lowest value of
404.5 respectively. Therefore, the decision to choose model (4) as the optimal model is a
reasonable one.
51
4.4.3.3 Business Model Innovation Productivity
The results for the Negative Binomial regression for Num_BMI can be found in Table 4.14.
In order to assess the impact of the independent variables on Num_BMI, model (4) will be used
as it gives the most unbiased results due the presence of relevant controls. The coefficients of
the three independent variables can be interpreted as follows:
• For a one unit change in Flexibility, the difference in the logs of expected counts of
Num_BMI is expected to change by 0.264, given all other variables in the model are
held constant. The result is significant as the p-value depicts a value lower than 0.05.
• For a one unit change in Responsiveness, the difference in the logs of expected counts
of Num_BMI is expected to change by 0.138, given all other variables in the model are
held constant. However, the result is not significant as the p-value depicts a value higher
than 0.05.
• For a one unit change in Partner_Collab, the difference in the logs of expected counts of
Num_BMI is expected to change by 0.099, given all other variables in the model are
held constant. However, the result is not significant as the p-value depicts a value higher
than 0.05.
52
(1) (2) (3) (4)
Num_BMI Num_BMI Num_BMI Num_BMI
Num_BMI
Flexibility 0.105 0.117 0.230* 0.264*
(1.01) (1.14) (2.19) (2.54)
BMI_CONT_1 -0.0948
(-1.37)
BMI_CONT_2 0.0694
(0.95)
BMI_CONT_3 0.145**
(2.70)
53
As the coefficients given in the Negative Binomial regression are interpreted as the difference
between the log of expected counts, the incident rate ratios must be computed in order to assess
the impact in terms of percentages. The following table shows the values of the coefficients in
incident rate ratio form with respect to model (4):
(4)
Num_BMI
Num_BMI
Flexibility 1.302*
(2.54)
Responsiveness 1.148
(1.32)
Partner_Collab 1.104
(0.98)
Other than the three independent variables above, the controls of Ind_Tech, Org_Size,
Exo_Macro and BMI_CONT_3 are also significant and their impact on business model
innovation cannot be disregarded.
Once again, in order to test for the goodness of fit of the four models, the AIC*N values were
used from the regression table. The regression table depicts these values for the four models
independently with model (1) having the highest value of 320.5, and model (4) having the
lowest value of 313.9 respectively. Therefore, the decision to choose model (4) as the optimal
model is a reasonable one.
The main analysis in the previous section was across multiple industries that were controlled
for in order to account for their impact on the three types of innovations. However, as the
54
industries vary in nature and strategy, it was considered important to conduct an industry
specific analysis as well.
For this reason, two different datasets were created. The first dataset contained all
organisations that belonged to the Manufacturing industry (Ind_MANU == 1). The second
dataset contained all organisations that had operations in the Service industry (Ind_SERV ==
1). After filtering the two datasets, a total of 37 observations were deleted for Ind_MANU and
only 45 valid observations remained. Similarly, a total of 48 observations were deleted for
Ind_SERV and only 34 observations remained.
Independent regressions were then run using both datasets to ascertain the industry specific
impact of Flexibility, Responsiveness and Partner_Collab on Num_PDI, Num_PRI and
Num_BMI. The model used in this analysis takes the form of the same regression equation as
that of model (4) in the previous section, but without the industry controls. The results of the
regressions are grouped together in the tables below and only the coefficients of the three
independent variables are shown for the purpose of clarity. In addition, the coefficients have
been converted to their incident rate ratios in order to interpret the results in percentage form.
Ind_MANU==1
55
Ind_SERV==1
The results for Ind_MANU suggest that Responsiveness and Partner_Collab are significant for
Num_PDI; Responsiveness is significant for Num_PRI; and Flexibility is highly significant for
Num_BMI. The coefficients of the significant results can thus be interpreted as:
The results for Ind_SERV, on the other hand, suggest that only Responsiveness is significant
for Num_PRI and there are no significant results for either Num_PDI or Num_BMI. The
coefficients of the significant result can be interpreted as:
The results of the industry specific analysis above point towards some similarities as well as
differences with the cross-industry analysis conducted previously. The table below illustrates
the results from both the cross-industry and industry specific analysis:
56
Cross-Industry (N = 82) Ind_MANU (N = 45)
Flexibility Responsiveness Partner_Collab Flexibility Responsiveness Partner_Collab
Num_PDI NS S S NS S S
Num_PRI NS S S NS S NS
Num_BMI S NS NS S NS NS
S = Significant NS = Not significant
Num_PDI NS S S NS NS NS
Num_PRI NS S S NS S NS
Num_BMI S NS NS NS NS NS
S = Significant NS = Not significant
Table 4.17 Summary of results
The tables above depict that the only difference between cross-industry and Ind_MANU is that
Partner_Collab is significant for Num_PRI in the former and not significant in the latter. On the
other hand, when it comes to Ind_SERV, only Responsiveness in Num_PRI is significant with
everything else being insignificant.
At this point it must be noted that the number of observations were reduced drastically during
the industry specific analysis, which makes it harder to rely on the results as stated above. For
the case of Ind_SERV, the results should be viewed with greater scepticism due to the fact that
the number of observations (N = 34) was well below the minimum number of observations set
at N = 40. Anything below N = 40 cannot be given enough weight to consider the results as
unbiased and reliable. In short, both Ind_MANU and Ind_SERV require more observations in
order for the results to be considered reliable. This presents an important area for future
research.
4.5 Results
4.5.1 Product Innovation Productivity
When it comes to PDI productivity, Responsiveness and Partnerships and Collaboration are
understood as having a significant impact. The Responsiveness factor includes the agility
attributes of having a modular approach towards products or services, an efficient IT System
(knowledge management), and an efficient communication setup (both inside the organisation
and with external clients). Together these attributes play an important role in bringing about an
increase in PDI productivity and their combined effect is positive.
Similarly, the Partnerships and Collaboration factor includes the agility variables of reliance on
partners, customer collaboration, and outsourcing and partnering activities. The combined
effect of these agility attributes is understood to be positive, even though reliance on partners is
57
assumed to have a negative impact on product innovation productivity. One explanation of this
could be that the other two agility attributes crowd out the effect of reliance on partners and the
combined result suggests a positive impact. The other explanation could be that the respondents
understood reliance on partners as simply a collaborative effort and not as a limiting factor.
Finally, the factor of Flexibility was not found to have any significant impact on PDI
productivity. This factor included the attributes of creative climate, iterative strategies,
promotion of a fast and flexible mind-set, simplification of products/operations to manage
complexity, and delegation of power and decentralised decision-making. Although some of
these individual attributes may have a significant impact, the combined effect suggests that the
impact is not significant.
The results suggest that PRI productivity is also positively affected by the factors of
Responsiveness and Partnerships and Collaboration. The agility attributes contained within
these factors are the same as those explained in the PRI productivity section above. However,
in the case of PRI productivity, Flexibility is almost significant with a p-value of 0.055. This
value is very close to the p-value of 0.05 taken as the threshold for measuring the significance
of the factors. This almost significant p-value suggests that Flexibility should not be completely
disregarded when thinking about PRI productivity. Some of the agility attributes in the
Flexibility factor will be more significant than the others, but Flexibility may still be taken as
potentially having a significant impact on PRI productivity. However, in order to ascertain this
claim, more observations may be required.
As far as BMI productivity is concerned, the factor of Flexibility comes across as the only
significant result. The relationship between Flexibility and BMI productivity is positive. All
agility attributes in this factor seem to have a direct link with BMI, as suggested by the
literature. On the other hand, Responsiveness and Partnerships and Collaboration do not have a
significant impact on BMI productivity. Through this, it can be ascertained that when it comes
to BMI productivity, the emphasis is more towards the abstract organisational behaviour and
environment. It is widely understood that organisations do not innovate their BMs on a regular
basis, however, through this research it was also ascertained that organisations do tend to
innovate specific parts of their BMs on a regular basis. It can be argued that innovating a
specific portion of the business model inherently implies innovating the business model
incrementally. Since the research only took into account three years (2012 – 2014), it can be
argued that most BMIs were incremental by nature. Therefore, it can be concluded that
Flexibility helps organisations innovate specific parts of their BM, eventually leading to a new
BM over time.
58
5 Conclusion
The overall purpose of this research study was to test if there exists a relationship between a set
of defined agility attributes and innovation productivity. The following research question
guided the study:
To what extent can agility impact the innovation productivity of large organisations?
In total, eleven agility attributes were defined through a thorough review of the literature. These
eleven agility attributes were then used to create three composite factors using the PCA, in
order to serve as the independent variables for this study. Based on the literature review, the
three composite factors were defined as Flexibility, Responsiveness, and Partnerships and
Collaboration; these factors correspond to the agility capabilities. On the other hand, Innovation
was broken down into three dependent variables: Num_PDI, Num_PRI and Num_BMI.
Finally, a number of controls were added as well in order to account for a biased study.
Based on the findings from the three independent regressions, the status of the proposed
hypotheses and whether they were accepted or rejected with regards to the cross-industry
analysis are shown below:
Hypotheses Status
59
A cursory look at the table above suggests that H2, H3, H5, H6, and H7 have been accepted,
whereas H1, H4, H8 and H9 have been rejected. It must be noted that the factors of Flexibility,
Responsiveness, and Partnerships and Collaboration depict a composite impact of the relevant
agility attributes contained within these factors. The individual impact of the agility attributes
cannot be ascertained through these results. It may very well be the case that in a given factor, a
particular agility attribute has a much more significant impact than another. Due to this reason,
the factor needs to be taken into account as a whole and the individual agility attributes should
all be considered together when assessing the impact of the relevant factor.
There is a possibility to try and determine the individual impact of the agility attributes as well,
by using the factor loadings generated during the PCA, however, that method does not provide
a reliable way of measuring the individual impact of each agility attribute. As the accuracy of
results cannot be verified through this method, the individual impact of each agility attribute
was not explored.
The three types of innovation productivity and their results will now be interpreted in further
detail.
5.1 Interpretation
Product Innovation Productivity
Partnerships and Collaboration showed significant results for PDI Productivity. When
customers collaborate with organisations, they co-create value by expressing their
requirements, sharing their knowledge, or by participating in engineering or manufacturing
activities. Such activities may in turn help boost the innovation activities of an organisation
(Darmody, 2009). Partnering and collaborating with external partners or suppliers can also
enhance PDI Productivity. Accessing the knowledge of other firms improves strategic
decisions. Also, networking and collaboration improves both innovation and performance
(Bock et al., 2012). A firm engaging in partnering activities looks for new knowledge outside
its own boundaries, thereby enhancing its PDI Productivity.
Responsiveness also depicted a significant and positive relationship with PDI Productivity.
Responsiveness was previously defined as the capability of a firm to identify future changes
and to cope with them when unforeseen. In fast-paced business environments such as Fast
Moving Consumer Goods or Technology industries, the market is most likely to be disrupted
by novel products that cannot always be foreseen due to secrecy. Being responsive in such
environments helps increase PDI Productivity by generating a quicker reaction to change than
other competitors i.e. being a fast follower; but most importantly by being able to anticipate
future trends and being the first to market i.e. being a first mover.
It was expected that Flexibility would depict a positive relationship with PDI Productivity. It
comprises of all the cultural aspects of a young organisation (see section 2.2.3). Such cultural
aspects were expected to enhance creativity and therefore have a positive impact on PDI
productivity. However, the results did not depict such a relationship. The results, however, need
to be interpreted cautiously. Even though Responsiveness and Partnerships and Collaboration
60
showed positive results, organisations should not discard the last composite factor of
Flexibility, primarily because certain agility attributes in the Flexibility factor could have a
potentially significant impact on PDI productivity.
In the case of PRI, Partnerships and Collaboration can be understood in terms of outsourcing
activities. Recently, organisations have engaged in Business-Process Outsourcing, not only as
part of an operational effort strategy to reduce costs, but more specifically with the objective to
constantly innovate (Lacity and Willcocks, 2013). In order to foster PRI in outsourcing
activities, it is crucial for an organisation to create strong incentives for the service provider.
Therefore, the outsourcing work is all about collaborating with the service provider and
creating a win-win environment, which in turn will lead to a higher PRI productivity.
Considering the importance of partnering activities in BMs, it was expected that Partnerships
and Collaboration would depict a positive relationship with BMI Productivity. Even though the
results did not suggest it to be the case, it should be noted that the key factor for BMI is access
to knowledge (Bock, 2012). Therefore, the significance level of Partnerships and Collaboration
has to be interpreted carefully.
5.2 Limitations
As with all research studies, there are always certain limitations present. This research is
subject to three main limitations.
The first limitation is with regards to the research design itself and can be broken down into
three sub-limitations:
61
• Firstly, this research followed a cross-sectional design, with a non-probability sample.
Such a sampling technique allowed the application of specific controls to our sample (i.e.
EU location, large organisations, employees with specific titles such as “Innovation
Director”). However, such a design could have had an adverse impact on internal validity
as the observations were not completely randomly assigned and were subject to self-
selection due to additional controls.
• Second, as the data was collected through a survey methodology, some of the responses
might have been based on the subjective interpretation of the respondents – in particular
questions containing the likert-scale.
• Third, it is hard to ascertain through the design of the study whether the number of
innovations truly represent the effect of the agility attributes. As the study evaluates the
number of innovations per organisation during the time period 2012 - 2014, it may be the
case that a given organisation was pursuing intensive R&D and innovation research in the
years preceding the observed period, and then launched their innovations in the time period
that was observed. This situation also applies vice versa.
The second limitation is concerned with the sampling method and can be fragmented into three
sub-limitations as well:
• Firstly, the number of observations used in the data analysis (82) cannot be deemed
adequate for completely unbiased statistical inferences. For optimal results, at least 300
observations should be used as part of the analysis.
• Second, the study is cross-industry in nature. Even though the study design is rigorously
controlling for these different industries, it should be kept in mind that different innovation
strategies and a different pace of innovation could affect the number of innovations across
industries.
• Third, the study is concerned with large organisations, which were defined as organisations
with more than 250 employees. The statistical models developed in this research control
for the size of the organisation, but it is important to note that the number of innovations
might differ significantly between a firm with 250 employees and one with more than
5,000 employees.
The third limitation is concerned with how innovation is measured. In fact, innovation was
defined as a new or significantly improved product (good or service), process or business
model. However, it may be the case that the respondents interpreted innovation simply as being
something new, a novelty, rather than an innovation. Innovation can be understood as having
the quality of something new, but most importantly innovation should have an impact and
create value.
62
5.3 Future Research
The study of agility attributes and their impact on innovation is still in its neophyte stage and
requires extensive research. A high number of participants in this study expressed great interest
in the aforementioned field of research, primarily due to it being a novel domain. However, in
order to strengthen the validity of this study and improve this research, the following areas
should be considered:
• Focus on one type of industry, one country or one type of innovation, in order to
improve the internal validity of the research and to increase the possibility of making
sound statistical inferences with regards to the observed population. It could be interesting
for instance to analyse the impact of agility attributes on innovation in the service industry
or to assess the impact of agility on just Business Model Innovation.
• Carry out a case study research, by observing one or more organisations internally. The
researchers could use hard data from these organisations through official documents,
especially regarding how innovation is dealt with. Researchers could also assess if the firm
is promoting a culture of agility or not by observing the internal environment and
disseminating surveys throughout the organisation. Such an approach would help decrease
the problem of subjectivity and therefore increase the reliability and validity of the study.
63
6 References
Abrahamsson, P., Salo, O., Ronkainen, J. and Warsta, J. (2002) “Agile Software Development
Methods – Review and Analysis”, VTT Publications, Issue 478, Pp. 107.
Amabile, T. M. and Khaire, M. (2008) “Creativity and the Role of the Leader”, Harvard
Business Review, 86, Pp. 100-9.
Bock, A. J., Opsahl, T., George, G. and Gann, D. M. (2012) “The Effects of Culture and
Structure on Strategic Flexibility during Business Model Innovation”, Journal of Management
Studies, (49:2), Pp. 279-305.
Bodje, N. R. and Balaji, V. (2012) “Evaluation of the Most Used Agile Methods (XP, Lean,
Scrum)”, International Journal of Engineering Science and Technology, Vol. 4, No.01, Pp. 23-
29.
Breu, K., Hemingway, C. J., Strathern, M., and Bridger, D. (2001) “Workforce Agility: The
New Employee Strategy for the Knowledge Economy”, Journal of Information Technology,
17(1), 21-31.
Bryman, A. and Bell, E. (2011) “Business Research Methods”, 3rd Edition, Oxford University
Press, USA.
Chesbrough, H. W. (2007) “Business Model Innovation: It’s Not Just About Technology
Anymore”, Strategy and Leadership, 35, Pp. 12-17.
Christopher, M. and Towill, D. (2001) “An Integrated Model for the Design of Agile Supply
Chains”, International Journal of Physical Distribution & Logistics Management, Vol. 31 No.
4, Pp. 235-246.
64
Clark, K. B. and Fujimoto, T. (1991) “Product Development Performance: Strategy,
Organisation and Management in the World Auto Industry”, Harvard Business School Press,
Cambridge.
Damanpour, F. (1992) “Organization Size and Innovation” Organization Studies, 13 (3): 375-
402.
Darmory, A. (2009) “Value Co-Creation and New Marketing”, TIM Technology Innovation
Management Review, November 2009, Article No 302.
Dodgson, M., Gann, D. and Salter, A. (2008) “The Management of Technological Innovation:
Strategy and Practice”, Oxford: Oxford University Press.
Dove, R. (1999) “Knowledge Management, Response Ability and the Agile Enterprise”,
Journal of Knowledge Management, Vol.3, No.1, Pp. 18-35.
Dove, R. (2001) “Response Ability: The Language, Structure, and Culture of the Agile
Enterprise”. New York: Wiley.
Doz, Y. and M. Kosonen, “The Dynamics of Strategic Agility: Nokia’s Roller Coaster
Experience,” California Management Review, 50/3 (2009), Pp. 95-118.
Ganguly, A, Nilchiani, R. and Farr, J.V. (2009) “Evaluating Agility in Corporate Enterprises”,
International Journal of Production Economics, 118, Pp. 410-423.
Gassmann, O., Sandmeier, P. and Wecht, C. (2006) “Extreme Customer Innovation in the
Front-End: Learning From A New Software Paradigm”, International Journal of Technology
Management, Vol. 33, No. 1, Pp.46-66.
Goldman S. L., Nagel R. N., and Preiss K. (1995) “Agile Competitors and Virtual
Organisations: Strategies For Enriching The Customer”, New York: Van Nostrand Reinhold,
Cop. 1995.
Goodstein, J., Boeker, W. and Stephan, J. (1996) “Professional Interests and Strategic
Flexibility: A Political Perspective on Organizational Contracting” Strategic Management
Journal, 17, Pp. 577-6.
Hage, J. and Dewar, R. (1973) "Elite Values versus Organizational Structure in Predicting
Innovation." Administrative Science Quarterly 18(3), Pp. 279-290.
65
Hannola, L., Friman, J. and Niemimuuko, J., (2013) “Application of Agile Methods in the
Innovation Process”, International Journal of Business Innovation and Research, Vol. 7, Issue
1, Pp. 84.
Hartley, J., (2006) “Innovation and Its Contribution to Improvement, a Review for Policy
Makers, Policy Advisers, Managers and Researchers”, London: CLG.
Horst, P. (1965) “Factor Analysis of Data Matrices” Holt, Rinehart and Winston, New York,
1965.
Kettunen, P. (2009) ‘Adopting Key Lessons From Agile Manufacturing To Agile Software
Product Development – A Comparative Study’, Technovation, Vol. 29, Nos. 6–7, Pp.408-422.
Lindvall, M., Basili. V., Boehm, B., Costa, P., Dangle, K., Shull, F., Tesoriero, R., Williams, L.
and Zelkowitz, M. (2002) “Empirical Findings in Agile Methods”, In Proceedings of the
XP/Agile University 2002: Second XP Universe and First Agile Univere Conference, Chicago:
Springer-Verlag, Pp. 197-207.
Lin, C. T., Chiu, H., and Chu, P. Y. (2006) “Agility Index in the Supply Chain”, International
Journal of Production Economics, 100 (2), Pp. 285-299.
Morris, L., Ma, M. and Wu, P. C. (2014) “Agile Innovation: The Revolutionary Approach To
Accelerate Success, Inspire Engagement, and Ignite Creativity”, New York: Wiley.
Nirmal, P. and Pantaleo, D. (2005) “The Agile Enterprise: Reinventing Your Organization for
Success in an On-Demand World”, Springer.
Nunnally, J. C. (1978) “Psychometric Theory” (2nd Ed.). New York, NY: Mcgraw-Hill.
66
Overby, E. Bharadawaj, A, and Sambamurthy. (2006). Enterprise Agility and the Enabling
Role of Information Technology. European Journal of Information Systems, 15 (2), Pp. 120-
132.
Ren, J., Yusuf, Y. Y. and Burns, N. D. (2003) “The Effects of Agility Attributes On
Competitive Priorities: A Neural Network Approach”, Integrated Manufacturing Systems, 14
(6), Pp. 489-497.
Rigby, C., Day, M., Forrester, P. and Burnett, J. (2000) “Agile Supply: Rethinking Systems
Thinking, Systems Practice”, International Journal Of Agile Management Systems, Vol.2,
No.3, Pp.178-186.
Rothaermel, F. T., Hitt, M. A. and Jobe, L. A. (2006) “Balancing Vertical Integration and
Strategic Outsourcing: Effects on Product Portfolio, Product Success, and Firm Performance”,
Strategic Management Journal, 27, Pp. 1033-56.
Sharifi, H., Barclay, I., Colqhoun, G., and Dann, Z. (2001) “Agile Manufacturing: A
Management and Operational Framework” Journal of Engineering Manufacture, 215(6), Pp.
857-869.
Sharp, J. M., Irani, Z. and Desai, S. (1999) “Working towards Agile Manufacturing in the UK
Industry”, International Journal of Production Economics, 62, Pp. 155-169.
Schneider, S. and Spietch, P. (2014) “Business Model Innovation and Strategic Flexibility:
Insights from an Experimental Research Design”, International Journal of Innovation
Management, Vol. 18, No. 6, Dec 2014, 1440009 (21 Pages).
Sull, D. (2009) How To Thrive In Turbulent Markets. Harvard Business Review, (February),
Pp. 17-32.
67
Tichy, N. and Charan, R. (1989) “Speed, Simplicity, Self-Confidence: An Interview with Jack
Welch”, Harvard Business Review, 67(5), Pp. 112-120.
Tidd, J., Bessant, J. and Pavitt, K. (2005) “Managing Innovation: Integrating Technological,
Market and Organizational Change”, 3rd Ed., John Wiley & Sons Ltd., Chichester, England.
Trott, P. (2005) “Innovation Management and New Product Development”, Prentice Hall.
Tseng, Y. H., Lin, C. T. (2011) “Enhancing Enterprise Agility By Deploying Agile Drivers,
Capabilities and Providers”, Information Sciences, Vol. 181, Issue 17.
Vecchiato, R. (2014) “Creating Value through Foresight: First Mover Advantages and Strategic
Flexibillity, Technological Forecasting and Social Change, September 2014.
Vokurka, R. J. and Fliedner, G. (1998) “The Journey toward Agility", Industrial Management
& Data Systems, Vol. 98 Iss: 4, Pp.165-171
Youssef, M. A. (Ed) (1994) “Agile Manufacturing: The Battleground for Competition in the
19990s and Beyond”, International Journal of Operations & Production Management, Vol. 14
No. 11.
Yusuf, Y.Y., Sarhadi, M. and Gunasekaran A., (1999) ‘‘Agile Manufacturing: The Drivers,
Concepts and Attributes’’, International Journal of Production Economics, 62, Pp. 33-43
Wadhwa, S. and Rao, K. S. (2003) “Flexibility and Agility for Enterprise Synchronisation:
Knowledge and Innovation Management towards Flexagility”, Studies in Informatics and
Control, Vol. 12, No.2, June 2003, Pp. 111-128.
Wilson, K. and Doz, Y.L. (2011), “Agile Innovation: A Footprint Balancing Distance And
Immersion”, California Management Review, Vol. 53, No. 2, Pp. 6-26.
Worren, N., Moore, K. and Cardona, P. (2002) “Modularity, Strategic Flexibility, and Firm
Performance: A Study of the Home Appliance Industry”, Strategic Management Journal, 23,
Pp. 1123-40.
Internet Sources:
Fowler, 2000, http://www.martinfowler.com/articles/newMethodologyOriginal.html
[Accessed 03/15/2015]
Highsmith, 2012, “Adaptive Leadership: Accelerating Enterprise
Agility”, http://fr.slideshare.net/dleyanlin/adaptive-leadership-wp-us-single-pages
[Accessed 04/20/2015]
The Community Innovation Survey, 2010, http://www.dst.dk/~/media/Kontorer/XX-
Direktionssekretariatet/brugerudvalg/vidensamfundet/cis_2010_final_9july2010_25112010-
pdf.pdf
[Accessed 02/06/2015]
The IBM Global CEO Survey, 2006, http://www-07.ibm.com/sg/pdf/global_ceo_study.pdf
[Accessed 11/15/2014]
68
7 Appendices
In the last decade of the 20th century emerged – in parallel of business agility – the concept of
Agile methods, as a response to the rather inflexible, bureaucratic, slow and formal software
development process (Gasmann et al., 2006). These two concepts – business agility and Agile
software development methods – were rarely brought together in the literature. Here, certain
methods employed in the SD process that could be applied to the Product Development process
are presented.
As for the literature on business agility, there is no general definition of Agile software
development. (Hannola et al, 2013). Nevertheless, the “Manifesto for Agile Software
Development”, published in 2001 by software consultants and leaders software development,
reports the extolled values of Agility:
It is interesting to denote the common traits between these four principles and the literature on
business agility. In fact, the literature revealed the role of the individual and the importance of
collaboration in order for the firm to be agile and this corresponds to value 1. The customer-
focused behaviour and the customer relationship attributes of agility refer to value 3. Value 4
“responding to change” is an integral part of business agility as defined by Zhang and Sharifi
when presenting the proactive side of agility.
Kettunen (2009) described the foundation of Agile SD as a small co-located team, working
closely together with customers, therefore creating a high-value product cost-effectively and
69
with continuous short iterations. Abrahamsson (2002) identified an Agile SD when it is
incremental (small releases and rapid cycles); cooperative (integration of customers in the
process); straightforward (the method is easy to understand); and adaptive (adapt to change).
Similarly, Lindvall et al. (2002) recognised Agile methods as being incremental, iterative, self-
organising and emergent.
Two of the most popular Agile method will be adressed: Scrum (the most popular method as
revealed by a survey conducted by Forrester, 2012), eXtreme Programming – commonly called
XP – (Gassmann et al., 2006).
The XP software development method focuses on iterative work and rapid cycles of
development. It consists of six distinct phases: Exploration; Planning; Iterations to first release;
Productionising; Maintenance; and Death (Beck, 2000, as cited by Bodje et al., 2012). Teams
are co-located, usually organised in pairs; they are coding in a clean and simple style and
working in a collaborative and cooperative environment, where quick feedback are required, in
order to deliver small and simple software packages in short two-week intervals i.e.
incremental work.
Schwaber (1995, 2002) described Scrum as being a flexible, adaptable, empirical, productive
and iterative method. The main idea of the Scrum method is that SD process involves several
variables, environmental and technical (requirements, time frame, resources and technology)
which are likely to change during the development process (Abrahamsson et al., 2002).
Schwaber (1995) distinguished three phases which composed Scrum: pre-game; game or
development; and post-game. The pre-game phase has two sub-phases: planning and high level
design. In the planning phase, a product backlog record listing the entire current requirements is
created and a definition of the system being developed is properly articulated. Based on that,
the architecture of the system is planned during the high level design sub-phase. In the
development phase, the system is developed in Sprints – a one to two-week chunk of work
where functionality is developed or enhanced to produce new increments. Finally, in the post-
game phase, the system is ready for release. Actions such as testing and documentation take
place here.
70
Source: Beck, 2000 and Schwaber & Beedle, 2002
71
Appendix B: Contact Message Draft
Thank you very much for your time (name of respondent). Have a nice day!
Hi (name of respondent),
I reached out to you last week regarding a questionnaire pertaining to my current Masters
Thesis research.
As I’ve not heard back from you, I am writing to check if you are still willing to participate in
this study. As of now, I have received over 80 responses from Innovation Leaders, Innovation
Directors and VPs from all over Europe. As you are a great fit for this study, I believe your
participation will add great value to the findings of this study.
Average time to take the survey is 10 minutes. The results of this survey will be completely
anonymous, and no identities will be revealed in the final report.
72
Appendix C: Questionnaire
73
74
75
Appendix D: Respondent Sample – Organisations
76