The Influence of Cross-Functional Collaboration and Knowledge
Integration Mechanisms on Contextual Ambidexterity and the
Moderating Effect of Environmental Dynamism
Master Thesis
Executive Programme in Management Studies Strategy Track
Student: Carolus Johannes Bart Bulsing Student number: 10681884
Date: January 2016
Statement of Originality
This document is written by Carolus Johannes Bart Bulsing who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Content
Abstract ... 4
Introduction ... 5
Theoretical background and hypothesis... 10
Ambidexterity ... 10
Organizational supportive context and contextual ambidexterity ... 12
Knowledge based view on ambidexterity ... 16
Supportive organizational context and cross-functional collaboration ... 17
Environmental dynamism ... 20
Cross-functional collaboration and knowledge integration mechanisms ... 21
Cross-functional collaboration, KIMs and ambidexterity ... 22
Methods... 24
Sample... 24
Survey ... 25
Measurement of variables ... 25
Dependent variable ambidexterity ... 25
Organizational supportive context ... 26
Cross-functional collaboration ... 27
Knowledge integration mechanisms ... 28
Environmental dynamism ... 28 Cross-functional collaboration ... 28 Analytic Strategy ... 30 Missing data ... 30 Recoding ... 30 Reliability ... 31 Scale means ... 31
Analysis and results ... 33
Discussion and conclusion ... 38
Limitations and further research ... 43
Appendix ... 45
Abstract
There is a wide range of studies on contextual ambidexterity, and it is argued that a supportive context positively influence the achievement of ambidexterity. Yet there is little understanding of the role of knowledge transfer and integration for achieving contextual ambidexterity. This research investigate this role and at also examines how managers can influence it. It is argued that knowledge transfer (i.e., cross-functional collaboration) and knowledge integration mechanisms are important antecedents for exploitative and
exploratory innovation. Further it examines how environmental dynamisms influence the level of cross-functional collaboration. Using a questionnaire under 88 respondents from 25 different industries, the empirical analysis indicate that an organizational supportive context positively affect functional collaboration. It also provide evidence that the use of cross-functional collaboration has a positive effect on the use of knowledge integration
mechanisms. And it shows that the interaction of cross-functional collaboration and
knowledge integration mechanisms appears to stimulate the achievement of ambidexterity. Interestingly the findings reveal that a lower level of environmental dynamism is negatively influencing the use of cross-functional collaboration. Through the expanded explanation and empirical assessment, this study contributes to the contextual ambidexterity literature by providing more evidence on the relation between the organizational context and
ambidexterity. Second, it extends prior literature by examining the influence of cross-functional collaboration and knowledge integration mechanisms on achieving contextual ambidexterity. Third, this research contributes to the existing literature by investigating the effect of a dynamic environment on the use of cross-functional collaboration.
Key words: contextual ambidexterity, cross-functional collaboration, knowledge integration mechanisms, organizational learning
Introduction
Some organizations manage to reinvent themselves in times of hard competition. For example, Philips was able to switch from lightning and consumer electronics to become a specialist in the medical healthcare market. More recently, Microsoft changed its structure to emphasize hardware with their brand Surface and to service products with products like their platform Cloud. How are these companies able to adapt to new environmental developments, or become a player in new markets, while they still have to maintain their position in the ‘old’ market to secure short-term returns? And why were other companies like Kodak or Polaroid not able to adapt to the new digital technology developments?
The ability of an organization to succeed over time in face of environmental and technological change derives from their ability to create an organizational context that enables the organization to adapt (Tushman and O’Reilly, 2002). An organization therefore has to make a trade-off between efficiency and flexibility. Put differently, it is the adaptive challenge for organizations to simultaneously exploit existing assets and capabilities, and to provide enough resources for sufficient exploration to keep an advantage over their
competitors despite changing markets and technologies. Search, variation, risk taking, experimentation, flexibility, discovery and innovation characterize exploration; whereas exploitation refers to subjects such as refinement, choice, production, efficiency, selection, implementation, and execution (March, 1991). Exploitation is focused on the daily business from which most of the returns of the organization arise at that particular moment.
Exploration, however, is less linked to returns on the short term but relates more to future returns (Benner and Tushman, 2003).
There is a natural habit of organizations to focus on exploitative activities, because that is what they are already good at. Besides, these activities cost less energy to pursue compared to exploratory activities they still know little about. This tendency is enhanced by what prior literature called the competency trap (Levinthal and March, 1993). That is, focus is kept on what the organization is very competent in, and therefore the assets and resources are used to increase the returns from these competencies. As a consequence, though, the focus on new innovation for future returns is neglected (Levitt and March, 1988; Gupta, Smith and Shalley, 2006). The shortage of exploratory activities makes the adaptive processes of an organization self-destructive, because it is not able to build or acquire the new knowledge that is necessary to be adaptive and to secure future returns.
Contextual ambidexterity is the ability of an organization to create a supportive context so individuals will simultaneously pursue both exploratory (discontinuous) and exploitative (incremental) innovation within an organization. By creating the right context and achieving ambidexterity an organization is less vulnerable for the competency trap and the lack of adaptive capacity, and in better conditions to gain competitive advantage in new areas (Benner and Tushman, 2003; O’Reilly and Tushman, 2013). But although it is known that innovation is built on new knowledge that has to be acquired, shared and integrated in the organization (Grant, 1996), what is less well understood is the role of sharing and integrating knowledge in achieving contextual ambidexterity.
First, there is still little empirical evidence on the contextual antecedents that positively influence the achievement of ambidexterity. Prior literature has emphasized the antecedents that can be created to positively affect alignment and adaptability like the organization
and Doerr, 2013), and “hard” and “soft” elements that should be combined by top management to stimulate individuals (Mom, Van Den Bosch, Volberda, 2009). The advantage of creating a context that contributes to contextual ambidexterity is that such an organizational design helps fast knowledge transmission between exploratory and
exploitative learning domains, which is necessary for the simultaneous development of any innovative and applicable solutions (Güttel and Konlechner, 2009, Andriopoulos and Lewis, 2009). This research contributes to the contextual ambidexterity literature by providing more evidence on the relation between the organization’s context and ambidexterity.
Second, it not clear how the search for and integration of knowledge is stimulated within a unit and how it influences exploitative and exploratory innovation. There are studies that emphasize knowledge transfer as a subject for achieving an ambidextrous state (Im and Rai, 2008; Jansen, Van Den Bosch and Volberda, 2006), but as Kauppila (2010, p.286) describes, there is a shortcoming in the contextual ambidexterity literature because it ‘assumes that exploratory knowledge is produced somewhere and that it is then selectively adapted to the organization’s purposes’. Moreover, prior literature has tended to focus on the relation of knowledge sharing and exploration, thereby ignoring how knowledge sharing can be influenced. What is also neglected is that knowledge sharing alone is not enough, but that knowledge should also be integrated in the organization to use it (Grant, 1996). As Simsek, Heavey, Veiga and Souder (2009) state in their study, exploitative and exploratory activities can aid the acquisition of new knowledge, but more research has to be done on the specific mediating variables that further explain the link with ambidexterity. Luca and Atuahene-Gima (2007) state that the application of knowledge for innovation performance, whether exploratory or exploitative, is preceded by processes of cross-functional collaboration and knowledge integration mechanisms (KIMs). So although we know that an organizational
context can positively influence ambidexterity, it is important to know whether this influence is mediated by cross-functional collaboration and KIMs. Because it gives managers more understanding how ambidextrous behavior in the organization is developed and it provides managers a tool to achieve more ambidextrous behavior. This study contributes to the contextual ambidexterity research by examining what the influence of cross-functional collaboration and KIMs is on the effect of an organizational supportive context on achieving ambidexterity.
Third, there is more research needed to further understand the role of environmental dynamics in achieving organizational ambidexterity. In their meta-analysis on organizational ambidexterity Junni, Sarala, Taras and Tarba (2013) state that several studies have mentioned the organization’s environment as possible moderator of ambidexterity but most empirical studies used the industry effects as control variable instead of a moderator. They propose that more research is needed to further understand the role of industry dynamics in achieving organizational ambidexterity. It is widely researched that in a dynamic environment the integration of new knowledge is very important (Wang and Ahmed, 2007; Eisenhardt and Martin, 2000). March (1991) argued that a dynamic environment allows the firm more opportunities for exploration and exploitation based on experiential knowledge. Lavie, Stettner and Tushman (2010) already argued that the initial need for balancing exploratory and exploitative activities was due to environmental influences. But there is no clear understanding of the effects of environmental dynamism on the sharing of new knowledge within the organization as antecedent of exploratory innovation. This research studies the effect of a dynamic environment on the relation between a supportive organizational context and the level of cross-functional collaboration in order to share knowledge.
From theories of organization learning, strategic management and the knowledge-based view, this research asserts that cross-functional collaboration and KIMs are determinant factors through which an organizational supportive context can positively influence
ambidexterity. The main contributions of this research are threefold. First it contributes to the ambidexterity literature by providing more empirical evidence on the relation between the organizational supportive context and achieving ambidexterity. Second this research gives insights in how a supportive context can aid achieving ambidexterity by stating that the interaction of cross-functional collaboration and KIMs is important for achieving
ambidexterity and that these processes can be positively influenced by a supportive context. Third this research provides more clarity about the role of the environment on the antecedents of ambidexterity. Moreover, a highly dynamic environment will lead to higher levels cross-functional collaboration to share knowledge that may result in exploratory innovation.
In the next section the literature background and hypothesis are presented. I examine the relation between organizational supportive context, cross-functional collaboration, KIMs and ambidexterity, as well as the moderating effect of environmental dynamism on cross-functional collaboration. After describing the research method, the empirical findings are presented, using data of 88 employees from 25 different industries. I conclude with the discussion and conclusion of the results and the limitations and issues for further research.
Theoretical background and hypothesis
The following section look into ways an organization supportive context can contribute to achieving ambidexterity and what the role of a dynamic environment is. A theoretical background is given on the concept of ambidexterity and on the relation between an organizational supportive context and contextual ambidexterity. After this a theoretical framework is provided from the knowledge-based view on ambidexterity. In the last section four hypotheses are stated that clarify: (1) the relation between an organization supportive context and cross-functional collaboration, (2) the moderating influence of environmental dynamism on such a relation, (3) the link between cross-functional collaboration and KIMs, and (4) the mediating effect of KIMs on the relation between cross-functional collaboration and ambidexterity.
Ambidexterity
In the literature the tension from simultaneous work on exploitative and exploratory activities is widely studied. Burns and Stalker (1961) already described how on one side mechanistic structures that rely on standardization, centralization and hierarchy, support efficiency and are linked to exploitative innovations or activities. And how on the other side organic structures that have a high level of decentralization and autonomy, support flexibility which is necessary for the creation of exploratory activities. The capability of simultaneously exploiting the existing capabilities and creating new ones is a key strategic challenge in creating competitive advantage (Hamel and Prahalad, 1992). Research on ambidexterity is trying to get more insights on how organization should cope with that challenge. Tushman and O’Reilly (1996) state that ambidextrous organizations have the ability to simultaneously pursue exploitative and exploratory innovation and argue that ambidextrous organizations are
likely to achieve superior performance compared to firms emphasizing at exploitation at the expense of exploration or vice versa. Exploitative innovations are incremental innovations that are developed to meet the needs from existing customers and markets (Benner and Tushman, 2003), while exploratory innovations are discontinuous or radical innovations developed for new customers or new markets (Lubatkin, Simsek, Ling and Veiga, 2006). Exploitative innovations arise from existing knowledge and reinforce existing skills, processes, and structures (Jansen et al., 2006, Levinthal and March, 1993). Whereas exploratory innovations require new knowledge or departure from existing knowledge (Jansen et al., 2006; Levinthal and March, 1993; Lubatkin et al., 2006).
From prior literature three mayor ways of realizing ambidexterity can be distinguished. The first is sequential ambidexterity that discusses how organizations can realign their
structures in line with changing environments. Organizations alternate between periods with a focus on exploitation capabilities, when seeking efficiency, and periods with a focus on exploration capabilities, when there is an emphasis on acquiring new knowledge for new innovations (O’Reilly and Tushman, 2013). The second way is through structural
ambidexterity where the emphasis is put on the balance between exploitation and exploration which is achieved simultaneously by separating both activities in different units (He & Wong, 2004; Lubatkin et al., 2006;Jansen, Tempelaar, Van Den Bosch, & Volberda, 2009). The last way which is widely discussed in ambidexterity studies, is contextual ambidexterity. Here the focus is on achieving the right balance within a unit by creating an organizational context for employees that supports them to divide their time in the best way between exploitative and exploratory activities (Gibson and Birkinshaw, 2004; O’Reilly and Tushman, 2013).
In effort to achieve ambidexterity prior literature endorses the role of top management as key because they can create and renew the context (Birkinshaw and Gupta, 2013;
Kortmann, 2014). Contextual ambidexterity helps to understand how organizations function. It provides the circumstances for a manager to make choices and trade-offs among different and competing objective, as well as to avoid the competency trap. That is the risk of
following the path of the least resistance and focus too much on exploitation with a neglect of exploration as negative result (Levinthal and March, 1993). Different studies have found evidence that a relative balance of exploratory and exploitative innovation strategies is positively related to performance (He and Wong, 2004; Wang and Rafiq, 2014; Junni et al., 2013). This research focus on the balance of exploitation and exploration in a business unit, thus on contextual ambidexterity. Following prior literature, in this research ambidexterity is measured as the sum of exploitative innovation and exploratory innovation within a unit (Gibson and Birkinshaw 2004; Gupta et al., 2006; He and Wong 2004; Jansen et al., 2006).
Organizational supportive context and contextual ambidexterity
For organizations to gain competitive advantage, or to survive for that matter, it is thus necessary to acquire a right balance between exploitation and exploration. From a resource-based view, the performance of an organization is influenced by its position in the market and is stock of valuable, rare, inimitability and non-substitutable resources (Barney, 1991). The ability of an organization to develop these resources and to obtain the best market position is influenced by its unique organizational capabilities. Both strategic positions, as for
organizational capabilities, managerial choices are important. Managerial choices are the decisions and actions that managers within the firm have taken over time (Ghosal and Bartlett, 1994). They are important because decisions and actions for the choices that arise
limited by the decisions that were made in the past (Porter, 1991). But what factors influence those decisions and actions, and why do some individuals do much more ‘for the good of the organization’ than their personal economic or political reward would justify (Barnard, 1938)? In search for answers to those questions, Ghoshal and Bartlett (1994) suggest that the
organization's context has an influence on the behaviors and actions of its members. Context refers to systems, processes, and beliefs that shape individual-level behaviors in an
organization. The organizational context must help all individuals to judge for themselves how they must divide their time between the conflicting needs for exploitation and
exploration.
Ghoshal and Bartlett (1994) describe the context as four behavior-framing attributes,
discipline, stretch, trust and support. The attributes are created and can also be reinforced or
renewed by management. They build their model on the assumptions that the four behavior-framing attributes create higher levels in initiative, cooperation and learning on an individual level. Creating an organizational design from these four attributes can thus enhance
individual entrepreneurship within the organization. In the following section the four attributes will be described.
Discipline creates the condition for members of an organization to voluntarily pursue
all expectations generated by their own explicit or implicit commitment. “It represents a way of life, a norm applicable to all tasks, rather than compliance with a well-defined set of contracts embodied in a company's strategic and operational control tools” (Ghosal and Bartlett 1994; p. 97). There are three important subjects that contribute to the establishment of discipline, the first is to have clear standards on expectations for everyone, i.e., not only the development of a set of standards but also the acceptance and commitment to these
standards by all individuals. The second is to have a system in which feedback can be easily provided in an open and trustful way. The third entails that the application of sanctions is done with consistency to create perceived fairness among all members of the organization.
Stretch is an organization state in which individuals voluntarily stretch their own
standards and expectations, doing more than they initially intended. To stretch their subordinates, managers stimulate them to strive for more rather than less ambitious
objectives. Moreover, a shared ambition and a collective identity are elements that contribute to the establishment of stretch as being part of group makes individuals pursue organizational goals. Furthermore the ability of individuals to see the relation between one’s own work and the overall priorities of the organization give personal meaning to them and creates
motivation for stretch at the individual level. With stretch as key element, managers can influence the aspiration levels of individuals; it can be a tool for promoting entrepreneurship.
Trust motivates employees to rely on the commitment of others. For instance a higher
level of perceived fairness and equity in the decision-making process increases the level of trust with individuals in the organization. Involvement in the outcomes of important
decisions, by making these outcomes transparent to everyone, can also support trust because it will contribute to the level of perceived fairness as well. Another element of trust is the increase in the level of competence at all levels of the organization, e.g. a CEOs with strong specific knowledge about the organizations products or markets are likely to be seen as more competent than CEOs who are more generalists.
Support is the last behavioral attribute of the contextual framework described by
rather than constrain lower-level initiatives and entrepreneurship. Important aspects that increase support are the availability and the accessibility of resources to members of the organization. Further, the freedom of initiative given to members of lower levels increased autonomy. As well as the guidance and help of senior management (as opposed to exercising authority) support the increase support for lower-level initiatives.
Ghoshal and Bartlett (1994) do not argue the above described four contextual attributes establish the capacity for ambidexterity. But they do see them as stimulation for individual-level behaviors to stimulate initiative, cooperation, and learning. In the search of finding antecedents that explain achieving contextual ambidexterity, Gibson and Birkinshaw (2004) build on this framework by suggesting that superior business unit performance is not
achieved primarily through charismatic leadership, a formal structure, or a “strong culture”. Instead it is achieved by “building a carefully selected set of systems and processes that collectively define a context that allows the meta-capabilities of alignment and adaptability to simultaneously flourish, and thereby sustain business-unit performance” (Gibson and
Birkinshaw, 2004; p. 210). Alignment activities are geared toward improving performance in the short term and have for that reason a focus on exploitation. Activities for adaptability on the other hand, are geared toward improving performance in the long term and have a focus on exploration (Benner and Tushman, 2003). Gibson and Birkinshaw (2004) found evidence that truly ambidextrous units outperform units which had a focus on either adaptability or alignment.
Knowledge based view on ambidexterity
Raisch and Birkinshaw (2008) argue that aligning organizations to a good balance of exploitation and exploration is dynamic than static. The question that follows is how to implement a dynamic balance in an organization’s strategy? An organization must have the capabilities to continuous scale exploitation and exploration to adapt to the dynamic
environmental changes. From the resourced based view this means that an organization has to develop dynamic capabilities to keep advantage over competitors. A dynamic capability is the capability of an organization to adapt its resources so new value-creating strategies for exploitation and exploration can consistently be generated (Eisenhardt and Martin, 2000, Wang and Ahmed, 2007).
From the knowledge based view we can derive that when knowledge integration leads to a continuous process of interaction of new information with existing knowledge, dynamic capabilities are developed (Grant, 1996; Kogut and Zander, 1992; Eisenhardt and Martin, 2000). Ambidexterity is about finding the best balance of using existing knowledge with new knowledge for exploitation and exploration. Thus ambidexterity becomes a dynamic
capability if these knowledge integration processes are strategically integrated so they can form a continuous process of using new knowledge with existing knowledge to align the balance of exploitation and exploration continuously (O’Reilly and Tushman, 2008;
Schreyögg and Kliesch-Eberl, 2007). This means an organization needs the right context with a common set of values, a shared vision, and clear governance processes and systems to encourage a continuous process of search, transfer and integration of knowledge (O’Reilly and Tushman, 2008).
The context create and stimulate what Eisenhardt, Furr and Bingham (2010) describe as “microfoundations” (i.e., the underlying individual-level and group level actions), which shape strategy, organization, and dynamic capabilities, and lead to greater organization-level performance. So the microfoundations are critical in finding the balance between efficiency and flexibility. Managing the paradoxes of exploitation and exploration is thus a shared responsibility, not only of top management, but across organizational levels (Andriopoulos and Lewis, 2009). In this shared responsibility the individual actions focused on new knowledge integration, define the extent to which the organization is able to adapt to environmental changes. Therefore an organizational supportive context not only support search, transfer and integrate knowledge for innovation, it also provides the microfoundations to create dynamic capabilities. By dynamic capabilities an organization can continuously renew value-creating strategies for exploitation and exploration (Eisenhardt and Martin, 2000, Wang and Ahmed, 2007). The process finding the best balance between exploitation and exploration will become an organizational perpetuum mobile. An supportive
organizational context is therefore not only important for achieving ambidexterity, it also important for remaining in a state of ambidexterity.
Supportive organizational context and cross-functional collaboration
In their research Gibson and Birkinshaw (2004) found support that the contextual attributes, discipline, stretch, trust and support may contribute to the development of
ambidexterity. Yet there is little understanding how an organizational context can positively influence ambidexterity. The role of existing knowledge and new knowledge is key in the development of exploitative and exploratory innovation. An organization needs knowledge integration for more efficiency and thus better exploitation, but it also needs new knowledge
to develop new innovations for spurring exploratory capabilities. Internal knowledge is linked to exploitative knowledge and external knowledge acquisition is often related to exploratory activities (Kang and Snell, 2009; Im and Rai, 2008). Ambidexterity may therefore depend on the firm’s ability to integrate internal and external knowledge bases (Raisch, Birkinshaw, Probst and Tushman, 2009). As Grant (1996, p. 375) states, “the primary role of the firm, and the essence of organizational capability, is the integration of knowledge”.
The specialist knowledge that is needed for innovation resides within individuals; the primary role of the organization is therefore knowledge application rather than knowledge creation (Grant, 1996). Prior research showed that top-down knowledge inflows are positively related to exploitation where horizontal and bottom-up knowledge inflows are positively related to exploratory activities (Mom, Van Den Bosch, and Volberda, 2007). Therefore, individuals in lower levels of the organization have to be stimulated to acquire and share new knowledge which is useful for new product innovations. Thus new knowledge applicable in the organization always starts with knowledge held by an individual.
Individual knowledge is transferred into the organization to be valuable for the organization as a whole. Knowledge transfer is ambiguous, but the primary reason for the existence of an organization is transferring information and knowledge through the
organization. The organizations systems and processes encompass the routines, knowledge sharing and knowledge itself. For instance when an individual leaves the organization, the organization does not stop performing, nor is the knowledge that provides production gone. The lack that appeared is simply filled in an organic way. Because knowledge transfer is such an ambiguous process, an organization can be seen as a living organism rather than a
about ideas (Nonaka, 1991). Questions reproducing these ideals are ‘where do we want to go’ or ‘what do we want to be’ give direction for individuals within the organization in their search for knowledge. The context of the organization is made on these ideals of what an organization should be. And the context must stimulate individuals to transfer knowledge.
Information and knowledge is needed for innovation, whether exploitative or
exploratory. Cross-functional collaboration (CFC) help to transfers specialist knowledge from individuals. CFC is about how often people from different departments (e.g. marketing, R&D and other functional units) work together in the product innovation process, how well they are represented in this process and how much they contribute to this innovation process (Kahn, 1996; Li and Calantone, 1998; Ruekert and Walker, 1987; Song, Montoya-Weiss and Schmidt, 1997). People contribute to CFC because they realize that different departments are strategically interrelated and they have to cooperate with each other for the benefit of the organization (Kahn, 1996; Luca and Atuahene-Gima, 2007). Characterization of CFC is that is intangible, unstructured and often comes forth as the result of individual will. In this regard the context of the organization can stimulate individuals to be more entrepreneurial and to better divide their time between exploitative and exploratory activities. An organizational supportive context creates higher levels of initiative, cooperation and learning on individual level (Ghoshal and Bartlett, 1994). Therefore individuals that are reinforced by an
organizational supportive context will engage in more CFC, so they can contribute to new innovation processes for the benefits of the organization.
Hypothesis 1: An organizational supportive context has a positive influence on the degree of cross-functional collaboration.
Environmental dynamism
Environmental dynamism refers to the rate of change and the degree of instability of the environment (Jansen et al., 2006).A dynamic environment may be caused by changes in technologies, variations in customer preferences, and fluctuations in product demand or supply of materials. Dynamic environments make current products and services obsolete and require that new ones be developed (Jansen et al. 2005). Hence, dynamic environments may stimulate exploration. Previous literature already argued that environmental dynamism is likely to moderate the impact of exploratory innovation on performance (Levinthal and March, 1993). Exploration involves the use of tacit knowledge from which new innovations can be developed to enter new markets or new customer needs (Lubatkin et al., 2006). As argued before the use of CFC stimulates new information processes to share and use new knowledge for innovation processes. When the environmental dynamism increases, an organization will favor flexibility, and thus less structure, to be able to give more emphasis on exploratory innovation (Eisenhardt et. al, 2010). Because the characterization of CFC is that it is intangible, unstructured and often a result of individual will (Luca and Atuahene-Gima, 2007), it favours from situations with less structure and more flexibility. A highly dynamic environment will provide the flexibility for individuals to engage in CFC, so they search and share new knowledge that increase exploratory innovation. I therefore expect that environmental dynamism will positively moderate the relation between the organizational supportive context and CFC, as depicted in figure 1.
Hypothesis 2: Environmental dynamism positively moderates the effect between an organizational supportive context and cross-functional collaboration.
Cross-functional collaboration and knowledge integration mechanisms
Knowledge that resides within individuals does not become available for innovation purposes if there are no adequate mechanisms for sharing and integrating the knowledge throughout the organization (Marinova, 2004). Higher use of CFC increases the sharing of knowledge, whereas KIMs help to integrate and exploit the acquired knowledge for
competitive advantage (Ordanini and Parasuraman, 2010). Where CFC is seen as intangible, unstructured and not imposed by higher level managers, KIMs are characterized by their regular, planned or mandated pattern of interactions that enable the transfer, recombination and use of knowledge from different functions. Thus KIMs are needed to exploit acquired knowledge for competitive advantage (Zahra and Nielsen, 2002).
Luca and Atuahene-Gima (2007) define KIMs as structures and processes, such as the use of documentation, information-sharing meetings, analysis of successful and failed projects and project reviews. KIMs enable organizations to internalize and reorganize what they have learned, and to decide how to use new knowledge in carrying out product
development projects (Tsai and Hsu, 2013). For that reason they are helpful in solving the problem of coordinating different knowledge to execute complex organizational tasks, like product innovation. There are two factors that define the relatedness between CFC and the use of KIMs. First, the use of CFC leads to more information processes between functional teams, increasing the need for coordination mechanisms to use the acquired knowledge or information (Kahn and Mentzer, 1998). Second, the transfer of knowledge among
interdependent units is ambiguous and uncertain. Therefore if there is only shared or transferred knowledge by the use of CFC and nothing is done with it, it remains only potential knowledge. It is the integration of this specialized knowledge that makes the knowledge applicable for innovation (Grant, 1996). Mechanisms are needed to capture the
value from collaborations between functions and to help to translate the collective knowledge into knowledge embodied in new products (Luca and Atuahene-Gima, 2007). Thus an
increase of CFC leads to a greater use of KIMs to coordinate information and knowledge flows for the use of product innovation.
Hypothesis 3: The greater the use of cross-functional collaboration, the greater the use of knowledge integration mechanisms.
Cross-functional collaboration, KIMs and ambidexterity
Exploration involves the use of tacit knowledge from which new innovations can be developed to enter new markets or meet new customer needs (Lubatkin et al., 2006). When an organization needs a higher level of exploration, higher use of tacit knowledge is needed. As argued aboveCFC is used to share tacit knowledge, and the use of CFC result in increased need for information processing demands to use the knowledge in the best and most efficient way. An increased level of KIMs will be developed to fulfill these information-processing demands (Birkinshaw, Nobel and Ridderstråle, 2002). KIMs are mandatory processes that try to ‘catch’ the knowledge; they bring order and structure in knowledge sharing, the use of knowledge, and in learning. Because of its structured property KIMs provide regular feedback possibilities that enable individuals to learn from past experience for new
innovations (Zahra and Nielsen, 2002). They provide the capture and analysis of knowledge and make it possible to synthesize and distribute knowledge among different functional units. Therefore KIMs facilitate combining firm capabilities (Luca and Atuahene-Gima, 2007) to reduce inefficiencies during the innovation process (Sheremata, 2000), and help exploit the acquired knowledge for competitive advantage (Zahra and Nielsen 2002). Formal processes
such as KIMs are especially important for exploiting the potential of complex and tacit knowledge into high-quality solutions and thus higher product innovation performance (Tsai and Hsu, 2013). KIMs thus provide the formal processes that help organization to achieve exploratory and exploitative innovation. Which can be used to adapt the organization by entering new markets or meet new customer needs (Lubatkin et al., 2006). Therefore KIMs positively influence the relation between CFC and exploitative and exploratory innovation, hence ambidexterity (see figure 1).
Hypothesis 4: Knowledge integration mechanisms positively influence the effect of cross-functional collaboration on ambidexterity.
Figure 1
Conceptual model of the role CFC and KIMs in the relation between OSC and achieving ambidexterity, and the moderating effect of environmental dynamism between OSC and CFC.
Methods
In this section the structure and design of the research is outlined. First, information is provided how the sample was determined and how the data was collected. Second, the measures reliability and appropriateness are discussed. The last section describes how the collected data was analyzed. For the complete questionnaire I refer to the appendix.
Sample
For the research a sample was needed with respondents whom have a link in their job to product innovation e.g. marketing, R&D, sales, etc. Also the sample should represent a wide variety of industries to minimize industry bias. To meet these requirements, the questionnaire was distributed in different modes. First it was distributed to employees of 4 business units within one publishing company by sending a link for the questionnaire by email. To get as many respondents from a broad range of organizations and industries, the link to the questionnaire was then also distributed by the LinkedIn network to get respondents from a broad variety of industries. Last, respondents that completed the questionnaire were asked to forward the link to the questionnaire within their business network. The sample finally
consisted of 103 respondents, 88 gave answers to all questions in the survey. The respondents that received a direct invitation via e-mail or LinkedIn were provided with a short explanation about the research topic in the invitation e-mail which also contained the link to the
questionnaire. To inform the respondents who did not receive an invitation e-mail but e.g. received the link to the questionnaire out of their network, the brief explanation was also displayed as the start of questionnaire. The explanation clearly described that the questions in the survey were on the business unit level for employees working in organizations with different business units. For respondents working in a single unit organization, it was
clarified that the concept ‘business unit’ could be taken as synonym for a ‘single unit organization’.
Survey
All constructs were researched in one questionnaire. The survey consisted of 8 constructs and 4 control variables. Together there were 48 questions in the survey, all
question are presented in the appendix. The constructs were assumed from prior literature, all original scales and questions of the constructs were used. The translation, back-translation procedure was used to translate all questions to Dutch. First a person translated all questions from English to Dutch, after that a second person translated the Dutch questions back to English. If the results of this back translation varied significantly, the Dutch translation was adjusted to avoid ambiguities and the potential for miscomprehension.
Measurement of variables
All measures were collected from prior literature. In the next section I will state the choice for the measure and scale used for the dependent variable ambidexterity and all other, independent variables that are used in the research. See also the conceptual model as
described on page 23.
Dependent variable ambidexterity
Following prior literature I used a two-step approach on exploratory innovation and exploitative innovation to measure ambidexterity (Gibson and Birkinshaw 2004; Gupta et al., 2006; He and Wong 2004; Jansen et al., 2006). The measure of exploratory innovation and
exploitative innovation were acquired from Jansen et al. (2006). Both constructs were
measured with a four item seven-point scale and are used in different researches, which subscribe the validity of the construct (Jansen et al., 2009; Kamasak and Bulutlar, 2010; Kortmann, Gelhard, Zimmermann and Piller, 2014). The four items consistent with
exploratory innovation described the business-unit as one which searches for and develops new products, example items are We commercialize products and services that are
completely new to our organization, and We frequently utilize new opportunities in new markets. The four items used to measure exploitative innovation were formed around the
more incremental product innovation that result in more efficiency and or economies of scale. Example items used are We frequently make small adjustments to our existing products and
services, and Our organization expands services for existing clients.
In prior research ambidexterity has been measured in distinctive ways the construct of exploratory innovation and exploitative innovation. He and Wong (2004) subtracted
exploitation from exploration, Gibson and Birkinshaw (2004) use the method of multiplying both constructs. I followed the method of adding exploratory and exploitative outcomes to measure ambidexterity in line with research of Lubatkin et al. (2006) and Jansen et al. (2009) who followed both the different procedures recommended by Edwards (1994) to measure ambidexterity and both found that adding both constructs result in the most explanatory power.
Organizational supportive context
Ghosal and Bartlett (1994) developed a model of the four attributes that engender individual behavior, which results in initiative, cooperation and learning. Gibson and Birkinshaw (2004) used this framework for their study on the mediating effect of
ambidexterity. But Wang and Rafiq (2014) state that the use of the model in their study has conceptual limitations because although the framework results in initiative, cooperation and learning, it is not conceptualized to understand the antecedents of exploration or exploitation behavior. Contrary to this limitation I used the constructs of Gibson and Birkinshaws (2004) study in which they found that performance management context and social contexts have a direct relation with ambidexterity. The first reason for this choice is because with this
research I focus on initiative, learning and cooperation within a business unit, all of which are needed to acquire and integrate new knowledge for exploitative or exploratory innovations. The second reason is because although they did not find enough results for the four individual factors, they identified the two factors. The first was performance management context as a seven item combination of discipline and stretch (α = 0.89) that used questions like ‘to what
extent does your management encourage to set challenging/aggressive goals’ and ‘to what extent does your management hold people accountable for their performance’. The second
factor was social context, that was conducted as a nine item combination of support and trust (α = 0.93) with question like ‘to what extent does your management give everyone sufficient
authority to do their jobs well’ and ‘to what extent does your management support you to be willing and to be able to take prudent risks’. For both factors a seven-point scale was applied
for all items.
Cross-functional collaboration
For this research I used the concept of cross-functional collaboration as used by Luca and Atuaheme-Gima (2007), who adapted the items from the study of Li and Calatone (1998). Three items with a five-point scale that focus on the extent of collaboration across different functions and departments measure the construct. Although the measure entails only three items, I decided not to adapt the measure for reliability reasons - Luca and
Atuaheme-Gima (2007) found a good fit (α = 0.75) based on these three items, e.g. ‘in this organization
different departments fully cooperate in establishing goals and priorities for our strategies’.
Knowlegde integration mechanisms
This measure of five items with a five-point scale, was adopted by the study of Luca and Atuaheme-Gima (2007), to measure to what extend an organization uses formal processes (e.g., information sharing meetings, formal analysis of projects) to distribute, interpret and integrate knowledge, e.g ‘to what extent does your business unit use regular
formal reports and memos that summarize learning to capture, interpret, and integrate knowledge and information about market and technology conditions’. Luca and
Atuaheme-Gima (2007) based the measure from research of Zahra, Ireland and Hitt (2000), and Zahra and Nielsen (2002) and they found a reliability of α = 0.83.
Environmental dynamism
To measure environmental dynamism, questions were used that were based on prior literature of Dill (1958), Volberda and Van Bruggen (1997) and Jansen et al. (2006). From this study, the five items on a 7-point scale (α = 0.87) were taken to measure the change and the instability of the environment. Question used for this measure were e.g. ‘Our
organizational unit has relatively strong competitors’.
Control variables
The results of the study were controlled by the variables business unit structural
item measure with a seven-point scale to indicate whether business units are structural
divided with strictly different tasks to perform, e.g. focus on innovation, long-term strategy or exploitation. Other control variables were organization size, business unit size, because large organizations often have more assets and therefore it may be easier to allocate assets for exploratory activities. Organization age was taken as a control variable because older
organizations have the tendency to focus on exploitation where young organizations are often too busy with exploratory activities to gain a position in an existing or new market. Also the simultaneous focus on exploratory and exploitative activities may be embedded by
institutionalization; rules, norms and values of the industry can have a high impact on
organizations culture, norms, values which influence individual behavior. Therefore industry was also attached as control variable.
Analytic strategy
All data was acquired by a survey through the webserivce of Qualtrics. Different steps were used to prepare the data for further analysis. In the next section all steps are described and the choices that were made are argued.
Missing data
All items that were unanswered but were seen by the respondent were valued as -99. All items that were not been seen by the respondent were kept empty. For all SPSS analyses the missing data were list-wise excluded.
Recoding
There was one reversed item in the construct of environmental dynamism; that item was recoded and saved as new variable. After analyzing the normality of the distribution for the variables, 3 items within the construct environmental dynamisms (ENVDYN_1,
ENVDYN_3, ENVDYN_4) indicated negative skewness (between -1.239 and <-.070) and positive kurtosis (between 1.206 and 6.966). Most respondents agreed firmly with the items, noticing that the ‘changes in the local market are intense’, ‘changes are taking place
continuously’ and ‘nothing has changed in the market last year’ (reversed item). Respondents came from 25 different industries with the general book publishing industry as highest
represented (36%) followed by the financial industry as second with 12% respondents. Although the respondents came from a wide variety of industries, there was little variance measured in the perception of the dynamism of the environment. A possible explanation could be that because of the high technological changes in society, people are more aware of
changes. This could also be supported by the negative economic situation in the Netherlands last years. An outcome therefore could be that independently of the industry, people perceive a lot of change in their environment. Therefore, and because there was a high number of respondents (84) that indicated the working of the central limit theorem, we assumed normal distribution for the measure.
Reliability
To check the consistency of the measurements, reliability checks were run for
performance management context (PMC), social context (SC), environmental dynamism
(ENVDYN), cross-functional collaboration (CFC), knowledge integration mechanisms (KIM), exploitative innovation (EXOI) and exploratory innovation (EXOR).On the two measurements exploitative innovation and exploratory innovation the Cronbach’s alpha, which represents the estimator of the internal consistency, was below 0.70 (0.64 and 0.65). There were no items that could be deleted to increase the internal consistency. Also further factor analysis did not reveal that there were different factors within the construct. Former research that used the same construct showed high consistency within the constructs, therefore the data was used with no further adaption. The difference in internal consistency could be result of the translation, although there were no large dissimilarities noticed from the translation – back translation process. All outcomes on the reliability tests can be found in table I.
Scale means
To use all items for further analysis, new variables were created for all constructs by computing the mean of all items within a construct. The construct organizational supportive
context (OSC) represents the performance management context (PMC) and social context (SC), which are both measured in a seven-point scale. Because both constructs are supportive for ambidexterity, the new variable OSC was computed that added both means of both
constructs. For the construct ambidexterity (AMB), the means for exploitative innovation (EXOI) and exploratory innovation (EXOR) were added together to the represent the balance between these construct. Former research by Lubatkin et al. (2006), and et al. (2009),
followed both the different procedures recommended by Edwards (1994) to measure ambidexterity and both found that adding both constructs result in the most explanatory power. The means and standard deviations of all variables are displayed in table I.
Analysis and results
To investigate if an organizational supportive context (OSC) can predict the level of ambidexterity after controlling for the different control variables, the data was analyzed with a hierarchical multiple regression. Respondents that worked in a single entity organization did not provided data for the control variable business unit structural differentiation. Therefore a first test was performed showed that the structural differentiation of units for exploitative or exploratory activities had no statistically significant influence. Then a second test was performed with a hierarchical multiple regression with the remaining four control variables as predictors of ambidexterity in the first model, but also this model was not
statistically significant. After adding OSC the second model was statically significant and the total model explained 14.0 % of the variance F(5, 82) = 2.66; p < .05
In this final model, the introduction of OSC explained 6.7% of the variance in ambidexterity F(1, 67) = 6.01; p < .05. With an increase of 1 unit of OSC (scale 1 to 5), the level of
ambidexterity will increase with B = 0.40 (scale 1 to 7). There were no other variables that showed statistically significant prediction for ambidexterity. These results further underline prior empirical research that an OSC positive influence Ambidexterity.
The PROCES regression functionality (Hayes, 2012) was used to test a serial mediation regression model (figure 2). Table II presents the results of this regression analysis. The total model has a significant coefficient of 0.454, t (86) = 2.888, p<0.005, indicating that that there is an effect on ambidexterity by the antecedents OSC, CFC, and KIM, in the model.
Although we found evidence for the positive influence of OSC on ambidexterity through the hierarchical multiple regression test, this model shows a positive direction
between organizational supportive context and ambidexterity, but this result is not significant, c1’ = 0.497, t(84) = 0.273, p=0.78. The positive effect on ambidexterity is thus not made only by the context of the organization, the indirect effects of CFC and KIMs are important
influencers.
As we expected from hypothesis 1, OSC has a positive and significant effect on CFC, a1= 0.365, t(86) = 5.284, p<0.01. This indicates that a higher degree of the combination of
performance management context and social context increases the amount of CFC. Therefore hypothesis 1 is supported. The effect of CFC is further positive related to KIMs, a3 = 0.265, t
(85) = 2.676, p<0.01. Hence more collaboration between teams indicates a higher use of knowledge integrations mechanisms. Thus hypothesis 3 is supported. Besides the effect from CFF, the model shows that the use of KIMs is also related to the supportive context of the business-unit, a2= 0.255, t(85) = 3.612, p<0.01. Which makes sense, because the context also
determine the systems and processes that could be used for KIMs. The influence of the serial mediation effect of CFC combined with KIMs can be interpreted as significant positive because the bootstrap confidence interval is entirely above zero (indirect effect = .055, SE = 0.039, CI: 0.002 to 0.181). From the results we can thus see the combined positive influence of CFC and KIMs on AMB, indicating that hypothesis 4 is supported. Although there is combined influence of CFD and KIMs, taken apart the positive influence does not give a significant indirect effect, b1 = 0.540, t(84) = 2.262, p<0.05 and b2 = 0.592, t(84) = 2.284,
ambidexterity via the influence of CFC, neither is there an extra significant effect on ambidexterity by the supportive context through the KIMs.
Figure 2: Conceptual model serial mediation
Table II: Serial mediation regression results
A model 1 PROCES test (Hayes, 2012) was used to test the moderating effect of the environmental dynamism (ENVDYN) on the relation between OSC and CFC (see figure 3 and figure 4 for a visualization of the conceptual and statistical model). The results of the
of the environment has a significant effect (c2, p<0.05) in a negative direction. The
environmental dynamism declares 5.06% of the variance of the extent to which people work together in cross-functional teams to share information and knowledge for product
innovation.
A closer inspection of the results indicates that the lower the dynamism of the environment the larger the negative effect on CFC in combination with an organizational supportive context (see table IV for conditional effects). The results indicate also that when dynamism is high the effect of the interaction with OSC has less effect and it is not
significant anymore, hence we cannot state that in that situation the interaction has a negative impact on CFC. We did not found support that high environmental dynamism has a positive effect on the relation of OSC on CFC, but we did find support that low environmental dynamism has a negative effect on the relation of OSC on CFC. Hypothesis 2 was not supported.
Figure 4: Statistical model moderation
Table III: Moderation regression results
Discussions and conclusion
Although there is a wide range of studies on contextual ambidexterity, the
understanding of the role of knowledge sharing and integration on achieving contextual ambidexterity was still unclear (Kauppila, 2010). This research contributes to the
ambidexterity literature by providing more empirical evidence about the relation between the organizational supportive context and achieving ambidexterity. It also presents more
understanding in how an organizational supportive context can be positive for achieving ambidexterity. It indicates that the mediation and interaction of CFC and KIMs is highly important for achieving contextual ambidexterity. Further the results indicate that a low dynamic environment has a negative impact on CFC, which is contrary to the expected outcome. Although this was not expected, it provides more insights about the role of the environment on the levels of CFC as antecedent of ambidexterity. Below the different results are discussed.
This research provides support for the positive relation between an organizational supportive context and achieving ambidexterity. Thereby providing extra empirical evidence to the contextual ambidexterity literature (Güttel and Konlechner, 2009, Andriopoulos and Lewis, 2009). It emphasize the important of managerial choices that create the organizational context (Porter, 1991; Birkinshaw and Gupta, 2013; Kortmann, 2014). It are these choices that create the attributes of discipline, stretch, trust and support, which are important to stimulate individuals to avoid the competency trap by following the path of the least resistance and focus too much on exploitation (Levinthal and March, 1993). This research underline that when individuals are supported to voluntarily pursue for all expectations generated by management, and when they feel the support to stretch their own standards and expectations, and when they can rely on the commitment of others, and they are supported to
take initiatives and show entrepreneurship, they will be able to divide their time in the best way between exploitative and exploratory activities. Hence, it will make the achievement of ambidexterity within the unit possible. And a relative balance between exploitative and exploratory innovation has a positive influence for organizational performance (He and Wong, 2004; Wang and Rafiq, 2014; Junni et al., 2013).
The tested serial mediation model not only provides evidence that an organizational supportive context build around the attributes discipline, stretch, trust and support has a positive influence on ambidexterity. It also provides understanding of how an organizational context can positively influence ambidexterity. The results of this research show that when a supportive context is created, individuals will act more as an entrepreneur. They will
voluntarily search, share and integrate knowledge for new innovations, which is a key element for achieving ambidexterity. Knowledge is sticky and tacit and therefore ambiguous in its transfer (Kogut and Zander, 1992). So it is impossible to create processes like
organizational routine to transfer tacit knowledge. The results of this research provide evidence that an organizational supportive context positively influence individuals to engage more in collaboration with other functional departments. They do so out of their own free will and for the good of the organization (Kahn, 1996; Luca and Atuahene-Gima, 2007). By this collaboration they share existing and new knowledge, both are required in the development of exploitative and exploratory innovation. This research thus reinforce the importance of
sharing individual knowledge, in lower-level of the organization, through collaboration between different departments so that knowledge can be used for exploitative and exploratory innovation (Grant, 1996).
Empirical evidence is provided that greater use of CFC leads to greater use of KIMs. The results demonstrate that KIMs are necessary to capture the value CFC and to help to translate the collective knowledge into knowledge embodied in new products (Kahn and Mentzer, 1998; Luca and Atuahene-Gima, 2007). Thereby there is support for the positive mediating effect of KIMs on the relation of CFC and ambidexterity. Which makes sense, if knowledge is only shared and nothing is done with it, it remains only potential knowledge. It is the integration of knowledge that brings innovations that lead to competitive advantage (Grant, 1996). So we can state that an organizational supportive context creates higher levels in initiative, cooperation and learning on individual level (Ghoshal and Bartlett, 1994), that is expressed by the use CFC and KIMs for innovations.
In regard to the achievement of ambidexterity, clarity is provided about the interrelated role of CFC and KIM. The results indicate it is the interaction of CFC and KIMs which lead to higher levels of exploitative and exploratory innovation, thus ambidexterity. This research did not find a significant effect from CFC as sole influencer on the relation of organizational supportive context on ambidexterity. Neither is there an extra significant effect on
ambidexterity by the supportive context influenced only by KIMs. This makes sense because both constructs, CFC and KIMs, are important for the integration of knowledge (Grant, 1996). One can fully cooperate in generating new ideas for innovation with cross-functional teams, but when this knowledge is not used or made tacit by KIMs not much is done with it. It also indicates that by a higher use of CFC and thus a higher level of the use of KIMs, there is a better balance between exploitation and exploration. This means there is particularly a higher level of exploratory innovation, since organization have a natural tendency to focus on exploitation which would indicate an oblique balance towards exploitation (Levinthal and March, 1993).
This is interesting, because understanding this process, gives managers the contextual tools to (1) actively support and stimulate individuals to engage in collaboration between different departments, and (2) to provide the best context for the use of KIMs (e.g. structure meetings, formalize reports and evaluate projects). As a result individuals from lower level functions and higher management together are stimulating ambidexterity (Andriopoulos and Lewis, 2009; Eisenhardt et al., 2010). Hackman (2002, p. 133) describes it in his book with a beautiful metaphor: ‘if well-designed work team is a seedling, then the organizational context is the soil in which it is planted, the milieu that provide the nutrients needed for it to grow and bear fruit.’ It are these microfoundations on group and individual-level that shape the
strategy, organization, and dynamic capabilities, and lead to greater organization-level performance (Eisenhardt et al., 2010), by which a reciprocal process of adaptation and alignment is captured. So top management better integrate the elements of an organizational supportive context in the development of the strategy so a state of continuous ambidexterity can be created.
The last contribution is that this study provides understanding in the role of
environmental dynamism on ambidexterity. This research contributes also to the literature because former studies used environmental dynamism as moderator on the effect of
ambidexterity, or as control variable (Junni et al., 2013), where this study the focus is on the moderating effect on the transfer of knowledge (i.e., CFC). Prior literature has argued that a dynamic environment stimulate exploration because it makes current products and services obsolete and require that new ones be developed (Jansen et al. 2005). And although most former studies indicate a positive effect of dynamic environment on ambidexterity, in this study we found the opposite to what was expected. The results of the research indicate a
negative moderating effect of environmental dynamism on CFC. More specific, the lower the dynamics of the market, the less people are willing to engage in collaborative behavior with other functional teams out of their own free will. A possible explanation could be that
because of the high technological changes in society, people are more aware of changes. This could also be supported by the negative economic situation in the Netherlands last years. An outcome therefore could be that independently of the industry, people perceive a lot of change in their environment. The results support former literature that with little use of CFC, fewer knowledge is shared which finally leads to lower levels of exploratory innovation (Lubatkin et al., 2006). It concludes that when there is little dynamism of the environment it is even more important for top-management to create a supportive context. So individuals will be more willing, supported, and able, to spend time and effort in the search for and sharing of new knowledge. As a result it will reduce the negative impact of the environment on CFC.
All together this study provides theoretical and managerial implications on achieving ambidexterity within a single unit. It provides more insight how an organizational supportive context can aid the achievement of ambidexterity. In contrast with most former literature about knowledge integration this study not only examines the mediating roles of CFC and KIMs in search for ambidexterity, but it also reveals when a dynamic environmental has influence on the level of shared knowledge by CFC.
Limitations and further research
There are different limitations which can be indicated and to provide direction for future research. First there is a limitation on the outcome of the variable environmental dynamism. Although prior research of Dill (1958), Volberda and Van Bruggen (1997) and Jansen et al. (2006) found good results on this measure, this research found only little variance between industries. That means that although there were respondents from 25 different industries, most of them perceived the organization’s environment as highly dynamic. Outcomes could be biased by personal perceiving of environmental environment because of the economics fluctuations as a whole (caused by the economic crises from 2008 and further), and because people perceive the world around them as a constant changing environment by the fast technological developments that colors our personal world. It would be interesting for further research is to get more clearly when high dynamics have a negative impact on ambidexterity or the antecedents like CFC, or a positive impact (Jansen et al., 2006).
Second, the sample for this research was also very broad and it was hard to determine if all respondents had a clear function link with product innovation. Future research could get better understanding with a more narrow determined target group that is clearly linked to product innovation.
Third, future research could investigate more thoroughly what items of the performance management and social context are most important for this influence on CFC and KIMs. Especially the influence of the context on exploratory innovation part of ambidexterity is not made clear, there might be items that are important in this relation but that were not taken into account for this research.
Fourth, this research was on a business unit level of analysis. We did not take into consideration the ambidextrous capacity of the individual, neither the influence of the network of the organization. As innovation arise more often from the community and although there is literature on inter-organizational knowledge transfer (Lavikka, Smeds and Jaatinen, 2015), a future meaningful direction for future research on ambidexterity would be research with the organization’s larger eco-system as unit of analysis.
Fifth, It might be necessary in order to align to environmental changes to make radical decisions in the allocation of resources. Retailers for instance might make a choice to invest heavily in a new digital strategy instead of investing in the brick shops. It could be that the best solution in these circumstances is to use structural separation to simultaneously conduct radical forms of exploration and exploitation as well as focus on contextual ambidexterity (O’Reilly and Tushman, 2013; Kauppila, 2010). In relative old industries like the publishing industry for instance, new ideas for exploration may arise from within the organization in order to adapt to the technologic innovations that the industry faces. But because these ideas are often radical different from their current exploiting activities, it may be better to exploit them in a structural separated unit. It is an interesting direction to study the impact of different levels (structural and contextual) of ambidexterity within one organization.
