The relationship between information system strategy, accidental innovation and contextual ambidexterity in organizations : an exploratory multiple case study research

(1)

Master Thesis Business Administration – specialization Strategy

The relationship between Information System Strategy,

Accidental Innovation and Contextual Ambidexterity in organizations:

an exploratory multiple case study research

Author:

Hasse Jansen

Student ID:

10121501

Supervisor: Dr. Sebastian Kortmann

Date:

January 2015

(2)

Abstract

This research investigates the relationships between information system (IS) strategy, contexts that invite accidental innovation, and contextual ambidexterity in organizations. Although there is a general consensus on the importance for organizations to become ambidextrous, by finding a balance between performing exploitational and explorational activities, varying conceptions of how to effectuate such a balance exist. While attempting to find a suitable balance between exploration and exploitation, firms have to cope with the ever-increasing influence of ISs in their business conduct. Using ISs as efficiency enhancers alone will not suffice in today’s dynamic business environments; managers need to find a balance between using ISs to facilitate - and to transform their business in order for their organizations to remain ambidextrous. Becoming ambidextrous, whether information systems are used to do so or not, requires firms to periodically take a big leap forward in their development. Remarkable innovations oftentimes come to existence serendipitously so organizations need to create a context that increases the likelihood of such valuable accidents occurring. Although the concepts of contextual ambidexterity, IS strategy and accidental innovation may appear to be connected in some way, their relationships have remained largely unclear. To explore the relationship between these variables more thoroughly, data has been collected interviewing 18 individuals from an intentionally diverse array of organizations. The empirical findings suggest a positive relation between all three variables. Also, a combination of existing literature and empirical findings provide a basis to propose environmental dynamism as a predictor to organizations’ IS strategy and accidental innovation scores. Directions of causality between IS strategy, accidental innovation and contextual ambidexterity are proposed with caution.

(3)

1. Introduction

The future is approaching businesses at an ever-increasing pace. But, while keeping a keen eye on the future in order not to overlook threats or opportunities, the exploitation of current day-to-day activities cannot become any less of a priority. Therefore, it is necessary for organizations to become ambidextrous, focusing on both the efficiency of current operations and the long-term effectiveness of their firm with the same dexterity. The creation of ambidexterity leads to superior business unit performance (Gibson & Birkinshaw, 2004) and a superior sales growth rate (He & Wong, 2004). However, our understanding of how ambidexterity can be realized is limited (Kortmann, 2012; Gupta et al., 2006).

Today, nearly all business executives believe that the information systems (ISs) in their organizations are critical to commercial success (Worthen, 2007). Traditionally, ISs and other digital technologies were constructed mainly to support business strategy and to make its execution more efficient. But, the importance of technology in business conduct has grown steadily over the last decades, to the point that technology should no longer be considered supportive of business strategy, according to Chen et al (2010). The effective use of ISs can now be considered equally important to commercial success as business strategy (Chen et al, 2010). This power shift requires organizations to adapt their internal structure. IS agents and business strategists are to engage in dialogue from a hierarchically equal position, working together in search of innovation to construct an innovative ‘information system strategy’, according to Chen at al (2010).

Organizations may attempt to become ambidextrous by inviting employees from the same business unit to perform both exploitational and explorational tasks, shifting activities depending on current requirements. Such employees are called contextually ambidextrous and have been proven to have a positive effect on firm performance (e.g. He & Wong, 2004; Gibson & Birkinshaw, 2004). The positive effect of a suitable IS strategy on firm performance has also been scientifically proven (Chen et al, 2010). But, the relationship between IS strategy and contextual ambidexterity under employees has remained unclear. Therefore, the first objective of this exploratory research is to investigate whether there is a relationship between IS strategy and contextual ambidexterity. Also, the suspicion that IS strategy positively influences contextual ambidexterity under employees is discussed. Among other elements, the increased reciprocity between IT department and business strategists, and efforts to stimulate both effectiveness and efficiency using IT that are indicative of organizations executing IS strategy, are expected to positively influence the diversity of knowledge and capabilities workers possess, which increases contextual ambidexterity under employees.

Aside of incremental improvements to existing products, periodically, firms need large leaps forward, creatively destroying the old to make room for the new. Interestingly, history has proven that many of the most revolutionary inventions were not meticulously sought after by their creators but were discovered serendipitously. The second objective of this investigation is to research whether there is a relationship between IS strategy and accidental innovation. Also, the suspicion that IS strategy positively influences organizations’ ability to effectuate accidental innovations is discussed. Along with other elements, the improved rate and quality of communication between IT department and business strategists, which characterizes organizations executing an IS strategy, is expected to increase the number and variety of ideas generated and lower the cost of experimentation, which characterizes organizations where accidental innovations are likely to occur.

The third objective is to research whether there is a relationship between accidental innovation and contextual ambidexterity. Also, the suspicion that the context required to effectuate accidental innovation positively influences contextual ambidexterity in organizations is discussed. Among other elements, the bringing together of people holding

(5)

different sets of expertise to discuss loosely defined problems, which characterizes organizations where accidental innovations are likely to occur, is expected to increase the ability of employees to perform both effectiveness and efficiency related tasks in response the context they operate in, which characterizes contextually ambidextrous employees.

Following the introduction above, the main research question of this thesis is:

“What is the relationship between Information System Strategy, Accidental Innovation and Contextual Ambidexterity?”

From the empirical evidence gathered to answer the above research question, an indication of the predictive capacity of environmental dynamism for organizations’ IS strategy and accidental innovation score emerged. The researcher thereon decided to also consult literature on environmental dynamism, and further analyzed its relation to IS strategy and accidental innovation.

This thesis is structured as follows: first, a literature review on contextual ambidexterity, IS strategy and accidental innovation, including the effects of environmental dynamism on IS strategy and accidental innovation, is performed. Thereafter, three working propositions describing the supposed relationships between IS strategy and contextual ambidexterity, IS strategy and accidental innovation, and accidental innovation and contextual ambidexterity are described. Because the relationships between the three main variables have only sparsely been researched before, the working propositions have been formulated with more subjective input from the researcher than is suitable for the formulation of hypotheses. Next, the methodology section will describe the methodological foundations of this investigation, including descriptions of the research design, interviewee selection criteria and the categories on which the data analysis is based. Thereafter, in the results section, within case and cross-case results are presented, as well as two elaborate testimonials on increasingly important IS strategy and accidental innovation related topics. In the discussion section, the results will be discussed from different, unusual viewpoints. Key findings will be described in the conclusion, and theoretical and practical implications will be provided after. At the end of this text, limitations to this investigation as well as suggestions for future research are presented.

(6)

2. Literature Review

In this part of this thesis, the current status of scientific knowledge surrounding three main concepts is described. Firstly, the body of literature on organizational ambidexterity is described, secondly, there will be elaborated upon the topic of information system strategy, and thirdly, the current knowledge on accidental innovation is described. Also, it is explained how organizations’ IS strategy and accidental innovation score could be influenced by environmental dynamism.

2.1 Organizational Ambidexterity

In the search for innovation, where innovation is defined as “intricate knowledge management processes of identifying and utilizing ideas, tools, and opportunities to create new or enhanced products or services” (Subramaniam & Youndt, 2005 In: Andriopoulos & Lewis, 2009), tensions can exist between parties looking for different levels of novelty, which can actually be opposing to each other. Androipoulos and Lewis (2009) have built a theoretical framework around their research on tension caused by pursuing innovation. Three aggregate dimensions to tensions were found, being (1) strategic intent, (2) customer orientation and (3) personal drivers, for which a suitable balance between integration and differentiation is to be found for al three dimensions. For the strategic intent dimension, profit emphasis and breakthrough emphasis need to be balanced; for customer orientation, a balance needs to be found between tight coupling and loose coupling of registered client requirements and firm decision making; and for personal drivers, a balance between allowing for passion in the workplace and imposing discipline needs to be found. Organizational ambidexterity is prized as a means to create the right balance between the powers at work in realizing innovation.

Duncan (1976) formulated the first conception of organizational ambidexterity as “an organization’s ability to be aligned and efficient in its management of today’s business demands while simultaneously being adaptive to changes in the environment”. In this definition, alignment refers to “coherence among all the patterns of activities in the business unit”, to coordinate and streamline activities and to deliver value by taking out costs of existing operations. Contrarily, adaptability is defined as “the capability to reconfigure activities in the business unit quickly to meet changing demands in the task environment” (Gibson & Birkinshaw, 2004), in other words, to have the ability to quickly move to new opportunities appearing in volatile markets.

The concept of ambidexterity introduced by March in 1991 managed to capture interest on larger scale. In this new article, March (1991) proposed a distinction between exploitation and exploration as different learning activities to which firms can allocate resources. In this conception, exploitation refers to activities associated with “selection, refinement, efficiency and implementation”, and is defined as “the refinement and extension of existing competencies, technologies, and paradigms” while exploration refers to activities such as “variation, search, experimentation and discovery” and is defined as “experimentation with new alternatives that have returns that are uncertain, distant, and often negative” (March, 1991). But, although quite many researchers have drawn a distinction between different sorts of activities and have stressed that organizations need to find a balance between those activities, an overarching theory explaining organizational ambidexterity has not been generated (Adler et al, 1999). Moreover, “the level of ambidexterity shown by an individual, apart from the organizational contexts faced by the individual, depends on the individual’s personal characteristics to great extent” (Raisch et al, 2009) which, among other reasons, has caused a lack of clarity around how organizations can achieve ambidexterity (Gupta et al, 2006, In: Kortmann, 2012).

(7)

In an attempt to manage trade-offs between conflicting demands firms may create a dual structure. In such a structure, certain business units focus on alignment while other units’ attention is centred around adaptation. This dual structure of business units taking an ‘either/or’ approach to their conduct is referred to as structural ambidexterity (Gibson & Birkinshaw, 2004). Alternatively, as indicated by Raisch et al (2009), firms can take a more dynamic approach to becoming ambidextrous and use temporal sequencing. Herein, business units temporarily switch their focus between performing exploitation related activities and performing exploration related activities.

However, in order to avoid isolation of R&D departments from the core business activities (Birkinshaw & Gibson, 2004), firms have started to progressively develop a more paradoxical ‘both/and’ approach to ambidexterity (Gibson & Birkinshaw, 2004) to end up with what Andriopoulos and Lewis (2009) call case of ‘synergistic and interwoven polarities’. Viewed as the opposite to structural ambidexterity (Birkinshaw & Gibson, 2004), contextual ambidexterity is far more complex and more difficult to achieve. Contextual ambidexterity is defined as the “behavioural capacity to simultaneously demonstrate alignment and adaptability across an entire business unit” (Gibson & Birkinshaw, 2004). Herein, individual employees are expected to choose between alignment-oriented activities and adaption-oriented activities in the context of their day-to-day work. Birkinshaw and Gibson (2004) indicate that individuals should satisfy four requirements to be called contextual ambidextrous: they need to be alert to opportunities and show initiative beyond the confines of their jobs, they need to be cooperative and combine their efforts with others, they are brokers looking to build internal linkages, and they are multitaskers feeling comfortable to fulfil a dual role.

To elicit these behaviours under employees, who therewith can be considered contextually ambidextrous, managers are to create a context where employees experience a suitable amount of discipline (1), stretch (2), support (3) and trust (4) in their daily operations, Gibson and Birkinshaw (2004) state, borrowing attributes from Ghoshal and Bartlett (1994). Discipline (1) “induces members to voluntary strive to meet all expectation generated by their explicit and implicit commitments”. Clear performance and behavioural standards, open, rapid and candid feedback, as well as consistence in the application of sanctions improve discipline under employees (Ghoashal and Barlett, 1994). Stretch (2) “induces members to voluntary strive for more, rather than less ambitious objectives”. Stretch can be established when members share a collective identity and ambitions, and when employees “give personal meaning” to the overall purpose of the organization. Support (3) “induces members to lend assistance and countenance to others”. Employees should have access to resources available to others, lower level employees should be allowed to show initiative, and senior functionaries should guide subordinates more than to exercise their authority over them in order establish support in the organization. Lastly, trust (4) “induces members to rely on the commitments of each other”. When employees have equity in the organization’s decision-making process of which the outcome affects them, when they experience the process as fair, and when people are seen to possess the necessary capabilities for the role they fulfil, a high degree of trust is likely to be present in organizations (Ghoshal and Bartlett, 1994).

2.2 Information System Strategy

In 2007, 87 percent of executives believed IT to be critical to their strategic success (Worthen, 2007), which has caused “an increasing appreciation for the potential of emerging technologies” (McAfee and Brynjolfsson 2008). IT is often indicated to be the anchor for the creation of both exploration and exploitation in organizations (Davenport and Prusak, 1998; Scott, 2000; Alavi and Leidner, 2001; Gold et al., 2001), and as such should be able to facilitate organizational ambidexterity, Van den Brink (2003) and Prieto et al (2007) argue. Although some attempts have been made to describe how organizations can achieve ambidexterity, the total body of knowledge surrounding this issue is still limited (Gupta et al,

(8)

2006). This section describes how managers can utilize IT to its full potential by executing an IS strategy, potentially increasing organizational ambidexterity.

Organizations’ information systems consist of the information technology infrastructure, application systems, data, and personnel that employ IT to deliver information and communications services in an organization (Davis 2000). Strategy can be defined as ‘an organizational perspective on setting and meeting organizational goal’s’ (Mintzberg 1987). Defining the term ‘information systems strategy’, however, has proven to be more difficult, Chen et al (2010) state. Disagreement exists as to what the scope of IS strategy should be, whether it should focus on the functional level, (Ragu-Nathan et al. 2001; Adler et al. 1992), the strategic business unit level (Chan et al. 1997), the organizational level as a whole (Earl 1989), or across organizational boundaries (Finnegan et al. 1998), Chen et al (2010) state. Also, some researchers articulate IS strategy to be something that the management of an organization plans in advance (e.g., Chan et al. 1997), while others consider it an emergent pattern (e.g., Ciborra 1994, 2004). According to yet others IS strategy can be a combination of these two, meaning it can be both planned in advance and can emerge without such planning (Benner and Tushman 2003; Galliers 2004). This conviction matches strategy as something that “reflects the collective mind of all the organizational members through their intentions and/or by their actions.”

Chen et al (2010) emphasize that an IS strategy does not have to be an ‘ex post’ only, actually realized IS strategy. They consider an IS strategy to be something that exists as an intention of creating something, independent of whether the intention has already yielded tangible fruits. Thus, the mere existence of IS in an organization, without the right intention, does not constitute the existence of an IS strategy (Chen et al, 2010). In agreement with Ashurst et al (2012) and Chen et al (2010), Ashurst et al (2008) state that the realization of benefits when implementing IT can only achieved if ongoing commitment throughout the development and implementation of the IT system can be established. Focusing on the technological solution is necessary but not sufficient for the successful realization of benefit. The focus should be on the benefits the technology is to create rather than the technology itself, therefore the undertaking cannot be successful if it is led by IT processionals only. IT professionals should operate in close collaboration with business stakeholders and ensure that the final system is aligned with end-users’ needs, Ashurst et al (2008) conclude.

Traditionally, the deployment of ISs in organizations is dependent on pull from business strategy (e.g. Earl, 1993; Chan et al. 1997; Sabherwal and Chan, 2001). In this conception, the aim is to find ways is which IS can support the business to sustain competitive advantage. If business strategy demands the organization to focus on market development or aspires cost leadership, ISs are to be designed to support these goals. In this view, IS strategy is a derivation of business strategy and does not have a strategy of its own (Chen et al 2010). Alternatively, some firms may engage in a technology push strategy, when business strategy follows IT strategy (Herstatt & Lettl, 2004). In this conception, the main concern is the effective management of the IS function to best allocate and utilize IS resources (Earl 1989, 1993). IS strategy exists separate from business strategy, and can even be considered a business within a business (Adler et al. 1992; Ragu-Nathan et al. 2001), providing services to customers within the organization (Ragu-Nathan et al., 2001, pp. 277). In this conception, IS strategy operates only on a functional level. But, Chen et al (2010) note that “an emphasis on the IS function might create difficulty for business functions to understand IS strategy and thereby build possible barriers to aligning IS strategy with business strategy” (Agarwal and Sambamurthy 2002; Chan 2002).

As the influence of IT on firm performance grows, so does the need to grow out of the push-pull dichotomy. The traditional relationship between business strategy and information systems needs to change dramatically. Chen et al (2010) state that IS strategy should not be examined as part of the business strategy. It should support as well as question the existing

(9)

business strategy (Earl, 1989). This standpoint provides an answer to the questions asked previously as to on which organization level IS strategy should be applied. Rather than on functional level, it should be applied to the organizational level (Chen et al, 2010). Although individual IS managers can have their own thoughts about IS, organizational IS strategy should effectuate a collective view in the upper echelons of the organization (Mintzberg, 1987). As such, Chen et al (2010) define IS strategy as ‘an organizational perspective on the investment in, deployment, use and management of IS’, in strategic management literature. However, not all organizations are required to execute an IS strategy in order to be successful. Having an IS strategy wherein IT department has a organization wide perspective and wherein the IT department operates on hierarchically equal level to business strategists is only required when organizations aspire to use IT in innovative ways. Chen et al (2010) have distinguished between two types of organizations and how they utilize IT: IS innovators and IS conservatives. IS innovators focus on exploration in IT usage by introducing new technologies. They strive to be the leader in their industry by responding to opportunities first. In contrast, IS conservatives focus on exploitation of IT by refining existing technologies, who “seek to exploit IS innovations only after they are carefully scrutinized.” To operationalize the definition, provided at the end of the previous paragraph, Chen at al (2010) defined IS strategy as ‘the degree to which the organization has a shared perspective to seek innovation through IS’ (Chen et al, 2010).

2.2.1 Environmental dynamism and IS strategy

The previous section described which internal dynamic, according to Chen et al (2010) is required for organizations to successfully pursue an innovative IS strategy. Referring to a large number of researchers who have investigated topics related to IS strategy, this section describes in which situations it is appropriate or even necessary for organizations to innovatively use IT.

The structure-conduct-performance paradigm (e.g., Domowitz et al. 1986; Porter 1985) suggests that “an industry’s structure influences the behaviors and performance of the firms in that industry.” One of the most influential characteristics of environments firms operate in is industry dynamism. Referring to (Keats and Hitt 1988), Xue et al (2012) explain that dynamism refers to the “volatility and unpredictability of the changes in various environmental variables with which a firm has to deal”. Dynamic industries are those where consumer preferences fluctuate and technologies alter quickly, and where, according to Tallon and Pinsonneault (2011), product life cycles are short, demand is volatile and customer turnover is extremely high. Contrarily, in non-dynamic environments, product life-cycles are longer, due to a slow rate of product innovation, allowing for homogenous products to be sold on large scale. Meanwhile, demand in non-dynamic markets is static and customer turnover is low.

Although the importance of matching firm behavior to suit its environment is well known under researchers and business executives, this knowledge has only been applied to the field of information technology to limited extent. Knowledge on how certain IT portfolios perform in different industry environments is limited, according to Xue et al (2012). Aral and Weill (2007) state that there is a convincing positive relationship among IT investments, economic productivity, and business value across distinct measures (Brynjolfsson and Hitt 1996, Dewan and Min 1997, Bharadwaj et al. 1999). But although total IT capital stock does improve productivity, it does not always lead to profitability (Hitt and Brynjolfsson 1996) and thus has little advantage for firms in itself (Bakos 1991, Clemons and Row 1991). Firms holding a comparable total IT capital stock demonstrate large differences in performance. “Some firms simply derive greater value per IT dollar even when controlling for industry-level variation”, Aral and Weill (2007) state.

(10)

In an attempt to relate industry characteristics to firms’ IT portfolios, Mithas et al (2013) have found that when industry turbulence is low, firms tend to converge their IT portfolios toward the industry standard. Conversely, when industry turbulence is high, the propensity for organizations to diverge from industry standards and look for valuable deviations increases (Mithas et al 2013). According to Aral and Weill (2007) “investments in a particular IT asset class deliver higher performance only along dimensions consistent with the strategic purpose of that asset”, but only few researchers have studied IT investments distinguishing between IT asset types (Aral and Weill, 2007). When firms decide upon investing in a transactional IT application, the costs for the execution of repetitive processes are bound to go down, but it will not influence product innovation. In turn, the more strategic IT applications are likely to stimulate product innovation but will not reduce operation cost. Moreover, Aral and Weill (2007) mention that different strategies can even have mutually exclusive performance implications.

Xue et al (2012) have researched the influence of environmental dynamism and complexity on firms’ IT asset portfolios. They state that when firms operate in an environment where dynamism is low, firms are likely to pursue relatively enduring strategies and compete through operational efficiencies and incremental innovations, referring to Tushman and Anderson, (1986). In more stable environments, firms that “sustain incremental innovation and focus on efficiency” rather than to seek innovation are likely to perform best, Benner and Tushman (2003) state. Such firms “build information systems for maintaining control over existing operations and seek to grow through incremental improvements in products and processes rather than through radical innovations to explore new product and/or market opportunities”, Xue et al (2012) state. They, furthermore, add that such firms’ IT asset portfolios are geared toward increasing operational efficiency by effectuating an increased payables, receivables and inventory turnover, and a decrease in selling and administrative costs. In contrast, turbulent environments should increase the likelihood of organizations to utilize IT more innovatively because “the ability of an organization to rapidly develop new technological capabilities is essential in an environment characterized by dynamic change” (Benner and Tushman 2003; Brown and Eisenhardt 1997). Firms operating in dynamic environments compete by “introducing new products and services and by identifying and expanding into new areas of opportunity” (Xue et al, 2012). Here, firms generally attempt to survey trends and events to be able to respond quickly to new product and market opportunities to stay ahead of the competition. Such firms’ IT assets should support exploration activities such as new product development and process innovations. Xue et al (2012) refer to Kleis et al (2012) stating that IT assets are to facilitate knowledge management, innovation production, and inter-organizational coordination.

Aral and Weill (2007), however, have developed a more precise, further disaggregated framework of IT asset qualifications, moving away from the often held “monolithic conceptualizations of IT toward a disaggregated view of IT assets”. They have used a framework developed by Weill (1992) and extended by Weill and Broadbent (1998), resulting in four IT asset categories: infrastructure (1), transactional (2), informational (3), and strategic (4) assets. IT infrastructure (1) assets are “made to provide a flexible base for future business initiatives and thus are made in anticipation of future business needs”, Aral and Weill (2007) state. Although large infrastructure implementations are costly, they lay the groundwork for performance improvements in the long-term (Duncan 1995, Broadbent et al. 1999). Transactional IT (2) investments are initiated to cut costs by automating processes, or to increase the output volume per unit cost. Informational IT assets (3) “provide information for managing, accounting, reporting, and communicating internally and with customers, suppliers, and regulators”. Informational IT can increase firm effectiveness by supporting responsiveness, reliability, control, adaptability and decision making, as well as increase efficiency by supporting sales analysis and mining of consumer behavior related data to optimize pricing and product offerings. Strategic IT (4) investments can drive innovation by supporting firms in entering a new market or developing radically new products or services,

(11)

therewith sometimes even changing the nature of service delivery or organizational processes in the entire industry.

Once executives have acknowledged the importance of a good fit between their organization and the environment it is in, and thus are likely allow their behavior to be guided by, for instance, environmental dynamism, the question of how firms can remain attune their environment for longer periods of time arises. For firms operating in an environment where dynamism is low, firms are likely to ‘get away’ with incremental, mainly efficiency related improvements such as improving transactional IT (2) systems and, arguably, information IT (3) asset investments. However, a greater challenge is reserved for firms operating in dynamic environments which are required to consider information IT assets (3), IT infrastructure (1) and strategic IT (4) investments in order to stay competitive in the long term. To remain attune their environment, according to many scientists, organizations need agility, defined as the ability to “detect and respond to opportunities and threats with ease, speed and dexterity” (Tallon and Pinsonneault, 2011) or, alternatively, the firm’s ability to “cope with rapid, relentless and uncertain changes and thrive in a competitive environment of continually and unpredictably changing opportunities” (Dove 2001; Goldman et al 1995). Tallon and Pinsonneault (2011) state that agility is “less of a necessity” for firms operating in stable environments because there is “less to gain from agility or less to lose from being slow to react.” In stable environments, occasions that require a quick response occur less frequently. In volatile settings, “the same degree of agility may have a far higher effect on firm performance due to the higher degree of market uncertainty” (Miller and Chen 1996; Miller et al. 1996; Sambamurthy et al. 2003), and there is more “downside risk to firm performance from failing to respond in time, while there is a greater chance that performance will improve if firms can react faster than their rivals” (Meyer 1982), Tallon and Pinsonneault (2011) state. Some researchers have stated that IT does oftentimes not contribute to firm agility or even inhibits it, stating that there could therefore be too much investment in restructuring internal processes (Lu and Ramamurthy, 2011). If IT investment is not done right, IT systems can become inflexible and rigid (Oosterhout et al, 2006), hindering firm agility. Embedding IT deeply within business activities can also lead to excessively automated, routinized, simplified, and rigid activities that hurt agility by limiting a firm’s strategic choices“ (Bharadwaj 2000; Henderson and Clark 1990; Sanchez 1995, In Tallon and Pinsoneault, 2011).“More IT spending and reliance on business intelligence tools may reinforce environmental scanning in the current domains, yet they may ignore signals in new domains” (Lu and Ramamurthy, 2011), thereby unintentionally increasing rigidity in managerial routines. IT’s apparent inhabitance of agility could incite researchers to believe that IT is, in many cases, mainly an efficiency enhancer relying on transactional IT to cut cost in the execution of everyday tasks.

However, as has become clear in the sections on strategic purpose of IT assets and disaggregated IT portfolios, different sorts of IT systems can exist to support a diverse array of tasks. As said, transactional IT systems are likely to support cost reductions by streamlining and/or standardizing processes presumably making processes more rigid, hindering innovation. But, as Aral and Weill (2007) explained, researchers have often taken IT on aggregate as their subject of study, without distinguishing between different types of IT systems serving different strategic objectives. IT can inhibit agility but it seems probable that such researchers have in these cases mainly focused on transactional IT systems while largely ignoring informational IT systems and strategic IT assets.

According to an increasing number of researchers, firm agility can actually be improved by IT when there is strategic IT alignment, defined as “the extent of fit between information technology and business strategy” (Tallon and Pinsonneault (2011). Tallon and Pinsonneault (2011) argue that literature on organizational agility has always evolved separately from the organizational alignment literature, “hence, the literature has largely overlooked agility as a

(12)

potential outcome of alignment”. According to Tallon & Pinsonneault (2011) researchers have repeatedly proven that strategic IT alignment affects “profit, productivity, sales growth, and reputation” (Chan et al. 1997; Chan et al. 2006; Oh and Pinsonneault 2007; Preston and Karahanna 2009; Tallon 2008). But, the effect of [strategic IT] alignment on firm performance is fully mediated by agility, proving that “the ultimate value of [strategic IT] alignment lies in how [strategic IT] alignment prepares firms for change” (Tallon and Pinsonneault, 2011).

Formulated in different ways, Xeu et al (2012), Mithas et al (2013) and Aral and Weill (2007), as well as the other authors mentioned, all emphasize that the effectiveness of organizations’ IT portfolios is affected by the degree of dynamism in organizations’ environment, and that therefore firms need to match their IT portfolios to fit the degree of environmental dynamism in order to perform well. Highly dynamic environments should motivate organizations to pursue a more innovative IS strategy (Chen et al, 2010) and install IT portfolios associated with such a strategy, while less dynamic environments allow organizations to ‘get away’ with a more conservative IS strategy. Therefore, the researcher expects environmental dynamism to be an important factor influencing organizations’ likelihood of employing an innovative IS strategy.

2.3 Accidental Innovation

For long-term success firms need to create a dual structure to both initiate radical innovation and execute incremental innovation (Duncan 1976, In: O’Reilly & Tushman, 2007 ). However, many organizations attempt to reduce uncertainty in innovation. Firms tend to take a rational approach to innovation and kill unproductive innovation oriented projects as the first signs of an impending undesired project pay-off in the future begin to appear (Austin et al, 2012). However, historical accounts of human achievement prove a significant positive role of accidents in innovation. Some even consider unexpected problems to be a key impetus for continuous improvement (Krafcik 1988, Shah and Ward 2003). Accidental inventions or serendipitous discoveries can turn out to be of great value to humanity, examples of which include dynamite, photography, stainless steel, smallpox vaccine, Viagra and many more (Austin et al, 2012). Therefore, a number of firms have actively attempted to create working conditions and surroundings that invite unpredictability, causing valuable accidents to occur more often.

As has also become clear in the previous chapter of this thesis, some researchers have discovered that technology may lead to process rigidity, dampening the ability to innovate (Licklider and Taylor 1968; Hedberg and Jönsson 1982; Kivijärvi and Zmud 1993; Henderson 1991, 1998; Davenport et al. 1998; Shneiderman 1998). Organizations’ IT and other digital technologies designed to support innovation have been criticized for not taking into account emergent value (Austin et al, 2012). Hedberg and Jönsson (1982) state that many IT systems “lead to organizational rigidity” and “tether organizations to yesterday’s perceptions.” When the problem space and design objectives cannot be defined in advance, automated procedures do not perform well, Kivijärvi and Zmud (1993) argue. Therefore, Boland and Collopy (2004) argue that a new “design attitude” is to be embraced, acknowledging that “the best outcomes will not be predefined and that innovation processes must support the emergence of value from unexpected sources” (Austin et al, 2012).

Digital technologies can be developed in a way that allows people from different departments or backgrounds to extract different meanings from the same materials (Carlile, 2002). Hedberg and Jonsson (1982) suggest ‘semiconfusing information systems’ that “encourage experimental behavior, variety in communication, perception, and evaluation, and systems that counteract organizational stability”. Markus et al (2002) emphasize the importance of an “emergent process of deliberations with no best structure or sequence” to support creativity. Austin et al (2012) refer to Avital and Te’eni (2009, p. 346) when they speak of the necessity

(13)

of presence for two parts; on the one hand there should be an IT systems present that evokes and enhances generative capacity of it’s users, and on the other hand, the users of the IT system are to “reframe reality to produce something ingenious or at least new in a particular context”.

For some businesses, efficiency is sought after, and therefore planning based process structures are most beneficial. For other businesses, where creativity is paramount, a more iterative approach is needed. Austin et al (2012) have identified the factors supporting accidental innovation. Whether the planning-based approach or an iterative approach is most beneficial dependent on two factors; the benefit and cost of original outcomes (Austin et al, 2012). For fast food chains, taking an iterative approach to stimulate creativity is very expensive and hardly beneficial. On the other end of the spectrum, visual artists do greatly benefit from creativity while the costs of prototyping with clay or digitally, or rehearsing a play are very low. Austin et al (2012) labeled the ratio of cost and benefit of experimentation ‘conduciveness to innovation’. As the ratio of benefits to costs becomes more favorable, conduciveness to innovation increases.

Austin et al (2012) found that as firms’ conduciveness to innovation increases, so does their openness to accident. Furthermore, they found that cost matters more than benefit as “benefit motivates openness to accident, but cost controls the degree to which it appears” (Austin et al, 2012). Also, the rate at which ideas are processed and converge into a final outcome is found to be important in the realization of an innovation (Austin et al, 2012). In some cases, a significant amount of time is needed to exhaustively extract the value of an innovation. If the convergence of a (rough) idea to its execution in practice is too rapid, a suboptimal result is inevitable.

What seems to be the most significant of firms’ features to stimulate innovation however, is the intentional search for accident. It was found that conduciveness to accident increased for firms collecting ideas without having a specific purpose in mind, rather than reusing ideas, for which a specific purpose has been determined in advance. The storage of collected ideas needs to be loosely organized, so that the retrieval process becomes more random and inefficient. This forces users to encounter many ideas they are not specifically looking for, which will all be stored in their “back head” (Austin et al, 2012) to support creative thought in the future. Austin et al (2012) refer to Dew (2009, p.739) who states that serendipitous discoveries happen “when the purposeful search encounters contingency in the presence of prior knowledge”. However, Austin et al (2009) go a step further and mention that innovative firms intentionally search for contingency. Accidents can occur when two people juxtapose wildly differing ideas, after which serendipitous synergies can be discovered.

2.3.1 Environmental dynamism and Accidental Innovation

The next section introduces a means called ‘improvisational capabilities’, deemed adjacent to the ‘capability’ to invite accidental innovation in organizations. Firstly, the concept of improvisational capabilities is described, and secondly, similarities between improvisational capabilities and accidental innovation are indicated. This description should install environmental dynamism as a plausible antecedent to accidental innovation.

One of the most well known means for managing organizations in moderately turbulent environments, showing “relatively predictable patterns of change” is dynamic capabilities (Eisenhardt and Martin 2000). Dynamic capabilities are defined as “the ability to integrate, build, and reconfigure internal and external competencies to address rapidly changing environments” (Teece et al. 1997) while turbulent environments are unpredictable environments characterised by “rapidly changing customer needs, frequent technological breakthroughs, and unpredictable new product introductions by aggressive competitors” (Pavlou and El Sawy, 2010). Dynamic capabilities are said to only thrive in moderately

(14)

dynamic and - turbulent environments.

Many researchers have produced works on how organizations should cope with moderately turbulent environments where patterns of change are relatively predictable (Eisenhardt and Martin 2000). When environments become very unpredictable, however, sense making collapses (Weick 1993) and past experiences offer little help. In such situations, reconfiguring existing operational capabilities may not be sufficient (Winter, 2003). However, it has remained “less obvious” how organizations should operate in such very turbulent environments (Pavlou and El Sawy, 2010). To manage organizations operating in very turbulent environments, Pavlou and El Sawy (2010) have introduced ‘improvisational capability’. Improvisational capability should be considered distinct from dynamic capabilities as the two capabilities are most effective when operating in different degrees of environmental turbulence. Dynamic capabilities is suited to “predict, sense, and ‘ride’ quasi-predictable patterns or ‘waves’ in the environment” while improvisational capabilities are likely to be effective when the environment is so turbulent with unanticipated ‘storms’. Improvisation can take place when conditions are too novel and unique for existing plans and capabilities to apply (Crossan, 1998), and /or when there is too little time or when it is too costly to engage in formal planning (Crossan et al. 2005), Pavlou and El Sawy (2010) argue. Improvisation is a way to cope with novel events and other challenges rising from environmental turbulence, but it is not always an involuntary last resort to cope with the situation at hand. Organizations can intentionally forego formal planning (Pavlou and El Sawy, 2010) and deliberately employ improvisation “because the anticipated outcome of spontaneous actions is expected to be superior.” Furthermore, although the word improvisational might invite the suspicion that it is a chaotic measure, according to Winter (2003) improvisational capabilities are, in this respect, similar to dynamic capabilities because they qualify as a capability, for which “the set of actions must be collective, repeatable or patterned, and purposeful”, not individual, ad hoc, or random (Pavlou and El Sawy, 2010). Although there are no strict rules, there is a pattern to expedite innovation (Cheng and Van de Ven, 1996). This, however, does not mean that improvisations always yield the desired effect, as “it is possible for them to be poor competitive actions that would result in poor outcomes” (Pavlou and El Sawy, 2010). The improvisation process is contingent, highly experiential and iterative, making outcomes unpredictable, Pavlou and El Sawy (2010) argue.

Practicing improvisational capabilities is in many ways similar to creating a context that facilitates the occurrence of accidental innovation. Provoking accidental innovations is, similar to improvisational capabilities, different from dynamic capabilities in that it is not about reconfiguring existing internal and external competencies. According to Ashurst et al (2012), inviting people from different backgrounds to extract different meanings from the same materials can be very fruitful for the creation of accidental innovations. Although it might look similar to the reconfiguring of existing competencies the dynamic capabilities approach is know for, the vital difference lies in the degree of newness that can be expected when using either one of the approaches. Reconfiguring existing competencies can yield creations make previous versions obsolete, but they are unlikely to be something nobody could have envisioned. In contrast, inviting people from different backgrounds combines two sets of expertise with the goal of creating something entirely new, unrelated and sometimes nearly impossible to trace back to either one of the two participants’ sets of expertise. The resulting creations will come as a surprise to the participants involved in the creation of the idea.

As stated earlie, in creating a context for accidental innovation, the cost of experimentation has to be low (Ashurts et al, 2012). In doing this, many wildly differing possibilities on product development or any other organizational process can be tested for effectiveness in the turbulent environments organizations need to operate in. Having the ability to experiment

(15)

freely at minimal cost, organizations improvise and discard the ideas that do not seem to be effective in the environment. Where improvisational capabilities rely upon employees’ wildly creative hunches, accidental innovations rely on the power of numbers and coincidence, when (nearly) all possible scenarios are tested for effectiveness against the environment. Although both approaches differ, they are similar in that both methods become more attractive, or even become the only effective measure for organizations operating in very turbulent environments. Therefore, the researcher expects environmental dynamism to be an important factor influencing the degree to which organizations invite accidental innovations.

(16)

3. Working Propositions

In this part of the text, expectations on how the topics introduced in the literature review relate to each other are described. Because the relations between the topics treated in the literature review have never been explicitly researched, hypotheses cannot be stated. An alternative, fitting the exploratory nature of this research, is the formulation of working propositions. Working propositions are provisionally accepted hypotheses serving as a basis for further research (ref 1). Even ultimately unsupported working propositions can contribute to the construction of a theory (ref 2).

3.1 IS strategy and Contextual Ambidexterity

As described in the literature review of this thesis, whether an organization can be considered to have an innovative IS strategy is determined by “the degree to which the organization has a shared perspective to seek innovation through IS” (Chen, 2010). Contextual ambidexterity is defined as the “behavioural capacity to simultaneously demonstrate alignment and adaptability across an entire business unit” (Gibson & Birkinshaw, 2004).

IS agents have traditionally focused on the ‘how’ of facilitative, efficiency and alignment related issues, operating on department level. When using an IS strategy, IS agents are to widen their scope to organizational level to propose explorative activities. Keeping in mind the technical possibilities of IS (how), IS agents should focus more on explorative activities (what). Conversely, business strategists are to narrow their scope to ‘department level’ and learn about the ample technical possibilities of IS (how), as a starting point for explorative activities (what). In other words, IS agents and business strategists take part in, what is traditionally the others’ job. Both parties need to start their reasoning from what is traditionally the other party’s starting point. Neither a push or pull strategy is used, and none of the parties is to lead the other. The one party is to think about the exploitational side to explorative activities, while the other is to think about the explorational side to exploitative activities, so to speak. On top of their respective traditional activities, the increase in adaptability is to come from IS agents, whereas business strategists will now reason from an technological, traditionally alignment focused domain as a starting point for adapting the firm to its environment.

Referring to the organizational attributes that encourage contextual ambidexterity under employees formulated by Ghoshal and Bartlett (1994), the close collaboration between IT department and business strategists present in organizations pursuing an innovative IS strategy is expected to increase ‘support’ among IT agents and business strategists. Increased ‘support’ should be the result of the sharing of resources between IS agents and business strategists, reduced authoritative stance of business strategists and invitation of initiative under IT agents. ‘Discipline’ is expected to increase as the rate of timely feedback IT agents and business strategists is considered likely to go up of the collaboration required for the effective execution of an innovative IS strategy. According to the theory, this will in turn invite employees to live up to their explicit and implicit commitments. The organization wide scope IT agents are to consider in search of innovation is expected to strive for more ambitious objectives, increasing ‘stretch’. Moreover, increased collaboration between IT agents and business strategists is expected to improve ‘trust’ under employees of both parties. Pursuing an innovative IS strategy can be expected to have positive influence on all four of attributes improving contextual ambidexterity under employees.

In organizations where the emphasis is on exploitation through the use of IT, such as at fast food chains, the highly influential IT agents are encouraged to reason more from an explorational perspective, like business strategists. Conversely, in organizations where the emphasis is on exploration, such as SpaceX, the highly influential business strategists are to

(17)

reason from a more exploitational, technological perspective, like IS agents. For firms relying on highly efficient processes, such as KFC, the increased explorational activities initiated by IS agents can be efficiency increasing measures. KFC, for example, can become explorational in their search for efficiency, by realizing revolutionary process improvements to cut costs. Conversely, firms relying on creative activities, such as SpaceX, can become exploitational in their creative activities.

Independent of the emphasis on either exploration or exploitation in a particular firm, IS strategy is likely to facilitate the increase in whichever of the two activities is least dominant. Increased interaction between equally ranked IS agents and business strategists strengthens the subordinate party and/or least practiced activity, causing them to perform a more balanced spread between explorational and exploitational activities. In realizing this, the organization will become more contextually ambidextrous.

Therefore, it can be expected that:

WP1: Innovative IS strategy has a positive influence on Contextual Ambidexterity

3.2 IS strategy and Accidental Innovation

As described in the literature review of this thesis, conduciveness to innovation positively influences openness to accident, which in turn increases the likelihood of the realization of accidental innovation in organizations. The primary driver of conduciveness to innovation is the cost of experimentation, and when this condition is satisfied, the next prerequisite in line is the potential benefit from innovation (Austin et al, 2012). Apart from these two requirements, the rate at which ideas are to be processed into final product, and the disorganized collection and storage of ideas are of significant influence to the likelihood of accidental innovation to be realized.

The researcher argues that the collaboration between IS agents and business strategists that is a requirement to execute an innovative IS strategy, is likely to reduce the cost of experimentation, thereby satisfying the first requirement for establishing conduciveness to innovation. When IS agents and business strategists are to engage in dialogue together, the lines of communication are bound to become shorter. Individuals from the separate departments are likely to knit together more closely, either online or offline. The cost of running an idea past someone from the other department then becomes smaller. The researcher does not foresee the cost of, for instance prototyping, to become smaller for a typical firm, but the cost of the very early development of an idea that exists only in the heads of a hand full of individuals does become smaller.

The researcher furthermore argues that when IS agents and business strategists communicate with each other as equals, as is the case when neither a push or pull strategy is present, neither one of the parties involved can cause serious harm to the other party. Therefore, it is more likely for both parties to run unorthodox ideas past one another. By doing so, the involved parties can inspire one another to look further than their traditional field of expertise. Creative ideas will not be withheld by fear of losing face, or getting fired. Therefore, the more revolutionary ideas are bound to be expressed and perhaps developed into a final product. When IS strategy is used, ideas are expected to become more revolutionary and more plentiful, increasing the possibility of realizing large benefit. This satisfies the second driver of conduciveness to innovation.

Also resulting from the absence of a push or pull strategy and the equality between business strategists and IS agents, the speed by which ideas are processed into final products the researcher expects to become lower when an IS strategy is used. Because none of the two parties can command and set unreasonable deadlines for the other party, the lead times of

(18)

ideas are likely to increase. This will, according to Austin et al (2012), promote accidental innovation.

The researcher further argues that the disorganization of the collection and storage of ideas is likely to increase when both IS agents and business strategists are encouraged to collect and store ideas. Both parties most likely have different habits and possess a different frame of reference of how ideas are to be collected and stored. The unintentional chaos resulting from an array of inevitable miscommunications and disputes between the two parties the researcher expects to strongly promote the disorganized collection and storage of ideas, increasing the likelihood of valuable accidents to occur.

WP2: Innovative IS strategy has a positive influence on Accidental Innovation

3.3 Accidental Innovation and Contextual Ambidexterity

If firms have managed to create a context wherein valuable accidents are likely to occur (by keeping the cost of experimentation low, the benefit of possible innovation considerable, the collection and storage of ideas unorganized, and the lead time of ideas to final product at least moderate (Ashurst et al, 2012)), the researcher expects firm adaptability to increase. Accidents can in theory happen at any place at any time, without considering ethical boundaries or social norms. If valuable accidents become more likely and thus in the long term inevitably more frequent, outmoded traditions and sacred cows are likely to be abolished. When only a planned approach to innovation is practiced, firms are more likely to intentionally or even unintentionally stay within the boundaries of tradition and social norms. The new products that came to existence through accident can seem useless at some point in time, but if those ideas are stored, they might become useful as time goes by and external business environments change. Firms’ stored collection of ideas presently considered useless can be viewed as a treasure chest for which the environment is only to change for managers to recognize its value. When the external business environment changes and a particular firm’s modes become outdated, this firm does not necessarily need to create new ideas to become competitive again. Employees simply need to go over the stored ideas and identify which idea can be of value in the changed environment. Because this firm has already completed the stage of generation of ideas, employees will be able to respond more quickly to changes in the market and/or business priorities, the researcher argues.

Although the researcher argues that the benefits of accidental innovation for increased firm adaptability are considerable and evident, its advantages to firm alignment are harder to identify. Deliberately disorganized communication between workers holding different sets of expertise, long lead times and unstructured storage of ideas are things rarely associated with high efficiency and do not appear to support coherence and efficiency in management systems. In fact, the drivers of accidental innovation seem similar to the reversed alignment characteristics (Birkinshaw & Gibson, 2004 ; Im & Rai, 2008 ; and Kortmann, 2011) presented in the next section. The strength of accidental innovation however, is that organizations can influence and provoke the occurrence of valuable accidents. The characteristics of organizations that have created an environment in which valuable accidents are more likely to occur, are opposite to alignment oriented firms, but intentionally so. In contrast to firms that aspire to be perfectly aligned but cannot manage to do so, accidental innovation oriented firms collect and store the fruits of misalignment (accidental innovations) and use them to their advantage. Unintentionally misaligned firms are more likely to leave accidental innovations unidentified and dispose of them, therewith destroying potential value. Accidental innovation oriented firms can utilize the fruits of misalignment to increase

(19)

adaptability of the firm, as the researcher has previously argued, but those fruits can also be used to increase alignment and efficiency. Although increases in efficiency are often realized by intentional, minor, evolutionary improvements, the major, revolutionary efficiency gains are more likely to come from less obvious and surprising sources. Therefore, the researcher argues that significant increases in firm alignment can sprout from the intentionally misaligned internal environment of the accidental innovation oriented firm. The researcher argues that, paradoxically, misalignment can cause alignment if the firm’s managers are conscious about misalignment and prepared to use it to their advantage. Accidental innovation will dominantly cause an increase in firm adaptability, but might simultaneously cause significant improvements in firm alignment. The researcher therefore considers accidental innovation to positively influence organizational ambidexterity.

Referring to the organizational attributes needed to create contextual ambidexterity under employees formulated by Ghoshal and Bartlett (1994), operating in an environment that invites accidental innovation can positively influence ‘stretch’ under employees, the researcher argues. Chasing potentially revolutionary accidents can easily be considered a “more ambitious objective” while potentially ‘hitting the jackpot’ in search of innovation, the researcher argues can also increase the “personal meaning” of work experienced by employees. The increased stretch under employees will stimulate their contextual ambidexterity.

Getting workers from different disciplines together to share ideas is bound to increase one person’s knowledge about the other person’s disciple. Furthermore, when workers are allowed to experiment more frequently and more freely, they are more likely to get in contact with ideas that are not usually considered to be related to their expertise. When ideas are stored in unorganized manner, and workers periodically glance over the stored ideas to gain inspiration, they are sure to get in contact with, again, ideas that are not usually considered to be related to their expertise. When lead times for ideas to develop into a final product are less lean and take longer, people that are the less obvious choice to shine their light on the project because they hold adjacent sets of expertise, are more likely to get involved in the project and can be involved for a longer time. The increased variety of people brought together should, over time, increase the variety of knowledge held by all participants. Gaining knowledge on other workers’ expertise can make it more likely for workers to be able to perform both exploitation and exploration related tasks when needed, the researcher argues.

(20)

4. Methodology

This section describes the research methods used in this investigation. A justification for the chosen research design, data gathering approach and data analysis process is provided. This section also contains a list of interviewees and interviewee selection requirements.

4.1 Research Design

The aim of this research is threefold. The first objective is to find empirical evidence to expose the relationship between IS Strategy and contextual ambidexterity. The second objective is to discover the relationship between IS strategy and accidental innovation, and the third objective is to investigate the relationship between accidental innovation and contextual ambidexterity. Information system strategy or comparable concepts have been researched fairly extensively. Similarly, quite many articles on organizational ambidexterity have been published but no consensus on how organizational ambidexterity can best be achieved has been reached (Gupta, 2006). Due to an apparent lack of interest in the relationship between IS strategy (especially the unusual internal organization needed to execute an innovative IS strategy) and contextual ambidexterity, a gap in the literature exists. As for the second and third research objective, very little is known about accidental innovation, with only one scientific article published on the specific topic. The relationship between having created a context that invites accidental innovation and IS strategy, and the relationship between such a context and the contextual ambidexterity of employees has never been researched. According to Brown (2006) “exploratory research tends to tackle new problems on which little or no previous research has been done”. Furthermore, the conceptual distinctions between the variables measured in this thesis are not know in detail, in which case, according to Shields and Rangarjan (2013), exploratory research is most suitable. Therefore, the researcher has chosen to conduct exploratory research.

Because some knowledge exists on the three separate main variables but knowledge on the relationship between the topics is scarce, the researcher decided to use different modes of generalization; both an inductive and deductive approach. The existing knowledge on IS strategy, accidental innovation and contextual ambidexterity was used to as a scaffold to guide the collection of new data. Themes have been identified and questions have been formulated based on existing theories, which is a deductive process. Thereafter, the data gathered was used to construct a tentative general explanation or theory. The resulting theory does not necessarily follow from the gathered data but reasons to accept the inference are outlined in the working propositions and discussion section.

Although the relationship between the three main variables was investigated using inductive and deductive methods, an explanation for organizations’ IS strategy score and (an inviting context for) accidental innovation score was reached though abductive reasoning. In abductive reasoning, a cause is inferred from an observed effect. An inadequate amount of empirical evidence to suggest environmental dynamism as a predictor to IS strategy and accidental innovation was gathered. Consequently, to be able to propose a casual relationship, the researcher consulted additional literature on environmental dynamism and literature relating environmental dynamism to IT portfolios and improvisational capabilities, which can be considered to share similarities with accidental innovation as explained in the literature review. Backed by the additional literature, the precondition of environmental dynamism was abducted from the empirically observed IS strategy and accidental innovation score of organizations. According to Dubois and Gadde (2002) “the abductive approach is to be seen as different from a mixture of deductive and inductive approaches”. Where the effect is a formal logical consequence of the cause in the process of deduction, in abduction, the cause is abducted from the observed effect.

(21)

4.2 Gathering Data

Qualitative data has been gathered by conducting a literature review as well as 18 semi-structured interviews. In the first 16 interviews conducted, interviewees were asked to express their opinion on how the three main topics treated in this thesis play a role in their respective organizations. The 16 case studies performed can be considered to serve “as a distinct experiment that stands on its own as an analytic unit” (Yin, 1994). Thereafter, 2 more interviews were performed, either one focussing on a different but distinct topic related to the three main topics treated in the thesis. Two highly experienced and knowledgeable individuals operating in the field of IT outsourcing, and idea generation/ innovation consulting, were asked to elaborate on the service their organization performs, and how these services relate to the topics treated in this thesis.

Using both existing literature about the topics as well as data acquired during 18 interviews allowed for data triangulation. Multiple sources of evidence provide different measures for the same problem, serving as “replications, contrasts and extensions to existing theory” (Yin, 1994), Eisenhardt (1989) states. According to Yin (2009 p115-116) “the most important advantage presented by using multiple sources of evidence is the development of converging lines of inquiry; a process of triangulation and corroboration”. Consequently, the 18 case studies performed for this research, according to Yin, (2009, p117), would be “rated more highly in terms of their overall quality, than those that relied on single sources of information” (Yin 2009, p117). Although triangulation is mostly done combining quantitative and qualitative sources, in this research the topics treated are viewed from multiple perspectives using different qualitative sources of information; existing literature and interviews.

Interviewing quite a large number of organizations instead of single cases, comparisons between cases can be made so that it can be discovered whether “an emergent finding is simply idiosyncratic to a single case or consistently replicated by several cases” (Eisenhardt, 1991). Furthermore, the multiple-case character is better suited to the inductive, theory building character of this study as “multiple-case studies typically provide a stronger base for theory building” (Yin, 1994), Eisenhardt and Graebner (2007) state, and are more deeply grounded than single-case studies.

Questions asked pertained to either one of the three topics, not to the relationship between the topics treated. In interviewing the three topics in relative isolation, the researcher gained insight in to which degree the topics play a role in the interviewees’ organizations and how their organizations score on these topics. This approach allowed the researcher to gain insight in the correlations between the distinct phenomena but do not explain the direction of causality. Inductive reasoning was used to propose a direction of the causal relationships between the three main topics treated. The researcher initially did not specifically ask questions about how interviewees experience the influence of environmental dynamism on the three main topics. However, about half of the interviewees explicitly mentioned environmental dynamism as a determinant for the degree to which any of the three main topics plays a role in their organization. The researcher argues that this incident should not necessarily be considered to negatively influence the quality of investigation, as according to Saunders et al (2007), when conducting exploratory research, researchers should be willing to change the direction of their investigation as new data is acquired and new insights are gained.

As more interviews were conducted, the rate by which new phenomena were introduced by interviewees slowed down significantly. Nearing the end of the series of 16 interviews, the researcher experienced a high degree of repetitiveness in the interviews signalling the point of theoretical saturation. This is the point where researchers should stop adding cases, according

(22)

to Eisenhardt (1989). Each interview was audio recorded and transcribed word for word, yielding a 77.000 word document.

4.3 Interviewee Selection

In contrast to the often-used statistical sampling where the chosen statistical population can represent an entire population, the researcher has performed a theoretical sampling approach of “polar types”. In this approach, extreme cases such as (expectedly) very successful and unsuccessful organizations, are chosen to discover contrasting patterns in the data (Eisenhardt, 1989). The researcher has attempted to select theoretically useful cases, “those that replicate or extend theory by filling conceptual categories” (Eisenhardt, 1989). According to Eisenhardt and Graebner (2007), selecting polar types could lead to “very clear pattern recognition in the central constructs, relationships and logic of the focal phenomenon.” Although estimating firm performance for the three main variables, prior to the interviews, can be difficult, about half of the selected firms did indeed perform either very poorly or very well and can in contrast be considered polar types.

The relationship between information system strategy, accidental innovation and contextual ambidexterity in organizations : an exploratory multiple case study research

Master Thesis Business Administration – specialization Strategy