The ambidextrous top-management team; the effect of proactive
members and the mediating effect of a well-developed transactive
memory system
MASTER THESIS
Author: Máté Molnár
Student number: 10828206
Date of submission: 29-06-2015
Qualification: MSc. in Business Administration – Strategy Track
Institution: Amsterdam Business School, Universiteit van Amsterdam
Statement of originality
This document is written by Student Mate Molnar 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.
Abstract
Ambidexterity in general has been previously widely researched and its importance for an organization’s long term survival is unquestioned. Although, top-management teams play an essential role throughout the entire organization to overcome the paradoxical tensions between exploitation and exploration, little is known about the necessarily demonstrated personal characteristics of top managers and the mechanisms that facilitate top-management teams to become ambidextrous. In this study I propose that being proactive allows top managers to overcome the paradoxical tensions between exploitation and exploration through different ways of learning and positively relates to an organization to become ambidextrous. Furthermore, I propose that the above relationship is partially mediated through a transactive memory system, which enhances efficient and effective information processing among its members. Drawing upon data from Dutch and German organizations, I found support that proactivity indeed promotes an ambidextrous top-management team. Furthermore, it is also supported that transactive memory system partially mediates the relationship between proactivity and top-management team ambidexterity. With my study I contribute to the ambidexterity literature through examining its antecedents and gaining further insights on the relationship between transactive memory system and ambidexterity.
Key words: ambidexterity, team ambidexterity, top-management team, proactivity, transactive memory system
Table of Contents
1. Introduction ... 1
2. Theory and hypotheses ... 7
2.1. Ambidexterity... 7
2.1.1. Organizational Ambidexterity ... 7
2.1.2. Ambidextrous TMT ... 10
2.2. Proactivity ... 13
2.2.1. Proactivity and learning ... 14
2.3. Transactive Memory System ... 17
3. Methodology ... 22
3.1. Measurement and Validation of Constructs ... 23
3.2. Analysis ... 25
3.3. Analytical procedure ... 26
3.4. Results ... 28
4. Discussion ... 33
4.1. Theoretical and Practical Implications ... 34
4.2. Limitations and Future Research... 35
References ... 38
1. Introduction
Companies throughout many industries, especially in high tech and knowledge intensive industries, operate in environments which are characterized by high industry clockspeeds and frequent changes in product technologies (Carrillo, 2005; Bourgeois and Eisenhardt, 1988). In order for companies to stay ahead of their competition and not to be forced to close down their business, they continuously have to adapt to the changing environment they are competing in (O’Reilly and Tushman, 2008). At the same time companies also need to think about the future and try to think ahead, but not just in terms of their competition, but as well as for the future demands of their customers (He and Wong, 2004; Smith and Tushman, 2005). This ambidextrous challenge faced by companies - to exploit their current expertise, but also constantly explore possible future solutions (Lewis, 2000) - has been a popular topic for press, and organizational and management research literature resulting in the publication of numerous research papers (Birkinshaw and Gupta, 2013).
Referring to ambidexterity at organizational level, an ambidextrous organization is one that is capable of addressing “two organizationally incompatible objectives equally well.” (Birkinshaw and Gupta, 2013, p. 291). Because of the various streams of research in the field of organizational ambidexterity, a universally accepted definition has not been set up yet (O’Reilly and Tushman, 2013). However, most of the researchers agree on the fact that “ambidexterity is an organizational capability that makes it possible to resolve different tensions that arise within an organization (Nosella et al., 2012, p. 450)”. These different tensions are mostly embodied in the exploitation vs. exploration duality, but there are also other types of tensions, such as alignment vs. adaptability, comfort of the past vs. uncertainty of the future (Birkinshaw and Gupta, 2013; Nosella et al. 2012).
influences the future direction of an organization (Hambrick and Mason, 1984). The top-management of an organization can influence the realization of organizational ambidexterity in various ways. TMT directly has an impact on it by allocating resources and integrating (O’Reilly and Tushman, 2008). It also has an indirect effect as they are the decision-making body in terms of structures, systems and processes (Teece, 2007). Having a deeper understanding about TMTs has dual importance. Firstly, every organization can only have one TMT, thus in case referring to the team of top managers one indirectly refers to the entire organization. Secondly, the top-managers form a team, thus we have to take into consideration the dynamics of a team as well to fully understand the implications of a TMT. These insights are especially important to see the processes how TMT is able to meet its decisions. First, in order to make their decisions they need to be able to process all the available information to them, which is given their vast size and quantity can be a big challenge. In my thesis I suggest possible solutions to overcome this issue. Second, top managers also need to resolve the paradoxical tensions between exploration and exploitation in order to be able to meet decisions to become ambidextrous.
Gaining more insight about how TMTs become ambidextrous further improves our understanding of ambidexterity on the organizational level (Haas, 2010). Despite of the above, the number of academic articles focusing on team level ambidexterity is lagging behind significantly compared to research on organizational ambidexterity (Birkinshaw and Gupta, 2013). Previous articles on team level ambidexterity included research on team compositional factors (Taylor and Greve, 2006), effects of the founding team composition (Beckman, 2006), introduction of newcomers to a team (Perretti and Negro, 2006), role of transformational leadership (Nemanich and Vera, 2009) and team effectiveness (Haas, 2010). At the same time there is also much doubt about the capabilities that managers need to demonstrate to be able to make decisions towards an ambidextrous organization (Birkinshaw
and Gupta, 2013; O’Reilly and Tushman, 2013). Less is know how managers are able to effectively make decisions regarding tensions between the opposite goals of exploration and exploitation. For this they need to be able to effectively process the available information. There are numerous conditions and circumstances that might or might not influence a team to become ambidextrously successful. So far, previous research has not convincingly presented the evidences for setting up an ambidextrous team within organizations. The individuals themselves play an important role in any team, however the necessary personal characteristics of these individuals has not been researched yet extensively. Every role requires various personal characteristics that make an individual successful in a specific position. Having a better understanding of beneficial and necessary personal characteristics has implications both for academic researchers as well as for practitioners and draws up potential avenues for future research.
Personality traits of team members within a TMT have not been widely researched before. Such a personality trait is proactiveness, which - according to Bateman and Crant (1993) - is type of behavior where individuals influence their environments and vice versa. They described proactive behavior - among others - as rather foreactive than counteractive. Foreactive behavior suggests that proactive team members would continuously search for new opportunities (explore) while making use of their past experiences as efficiently as possible (exploit). Proactive individuals seek information and opportunities instead of waiting until they get to them (Crant, 2000) and proactively change and adapt to their environment and focus on the future (Becherer and Maurer, 1999). Through their proactive approach they end up operating differently in their positions as others and can more effectively search for and process information. Proactiveness can lead to successful exploitation and exploration through single- and double-loop learning (Argyris and Schön, 1978; Argyris, 1991), which
also helps to transcend the paradoxical tensions within teams faced by ambidextrous challenges (Lewis, 2000).
Having proactive individuals in a team can have positive effect on the level of ambidexterity of a team. At the same time, having the right group of people put together might not be enough by itself. Rather there have to be some further mechanisms in place that enable the team members to work together as a team in a coordinated fashion with smooth interactions (O’Reilly and Tushman, 2008), rather than working as a group of individuals put together. In my thesis I address this gap with a model in which transactive memory system (hereinafter: TMS) partially mediates the relationship between proactivity and ambidexterity and provides basis for concentrated action.
The concept of transactive memory was first introduced by Daniel M. Wegner, according to whom “[t]ransactive memory refers to the idea that people in continuing interpersonal relationships often develop a specialized division of labor with respect to the encoding, storage, and retrieval of information from different substantive domains (Wegner, 1987, in Hollingshead, 1998, p. 425).”, whereas a TMS “describes the active use of transactive memory by two or more people to cooperatively store, retrieve, and communicate information.” (Lewis, 2003, p. 588). TMS emerges through the sharing of members’ “contributions across knowledge domains” (Heavey and Simsek, 2014, p. 4). Investigating internal relations within a team is important to understand how a team can achieve its ambidextrous goals. A well-developed TMS facilitates organizational members to find and make use of all the explorative and exploitative knowledge within the organization (Argote and Ren, 2012). It can connect members throughout the organization, even from lower levels given its autonomous nature (Peltokorpi, 2012).
In my thesis I argue that there is a positive relationship between individual team members’ proactiveness and the extent of how a TMT is capable of addressing two
conflicting objectives at the same time. The seemingly opposite goals of exploitation and exploration can be resolved through the various ways of learning, at the same time as helping to transcend the paradoxical tensions challenging TMTs. I contribute to the proactivity literature by extending the possible research fields of the consequences of proactivity and discover further connections between previously not researched constructs.
However, while proactiveness may have a direct effect on team ambidexterity, I argue that proactiveness does not fully stimulate TMTs to address exploration and exploitation simultaneously. Rather this relationship is partially mediated through a well-developed TMS. Previous research has investigated the relationship between TMS and personality traits (Ren and Argote, 2011; Pearsall and Ellis, 2006). Haas (2010) also suggests the relevance the future research between TMS and team ambidexterity. However there has been no research conducted before investigating the relationship between proactiveness and team ambidexterity, through the partial mediating effect of TMS. I argue that TMS contributes to team members’ ability to effectively and efficiently locate and retrieve information to meet their ambidextrous goals (Argote and Ren, 2012). In my research I expand previous research on TMS. On the one hand, I further develop previous reseach by investigating the connection between a further personality trait as a team composition antecedent of TMS (Ren and Argote, 2011). On the other hand, I also broaden the previous research on TMS’ possible consequences in connection to achieving ambidextrous goals.
In my thesis I will focus on organizational ambidexterity, as well as investigating the dynamics within TMTs. Consequently, through gaining insights on teams in connection with TMTs, I will immerse myself in organizational ambidexterity at that same time. Thus, the ultimate contribution of my thesis is expanding the academic research of ambidexterity, with special emphasis on ambidexterity within TMTs. Because of the changing environmental pressures, organizations change their approach towards the markets at which they are present.
This slight shift requires different attitude from organizations in order to be successful. Thus academic research has to react to this shift as well and discover the new avenues of research. In my thesis I address this shift and further investigate the necessary antecedents for a team, and especially for a TMT, successfully achieving high level of ambidexterity.
In the following, I will first review previous research on ambidexterity in general and on the team level, as well as on TMS and proactivity, and consequently establish the links between them. As next, after establishing my hypotheses and introducing the sample, I will test my hypotheses, followed by data analysis. Subsequently, I attend to discuss the results and insights gained based on the analysis, and draw up possible avenues for future research. Figure 1 shows the conceptual model used in the thesis.
Figure 1: Conceptual model
H4 H1 Organizational Ambidexterity Proactivity Transactive Memory System H2 H3
2. Theory and hypotheses
“To survive, organizations must execute in the present and adapt to the future. Few of them manage to do both well.” – Eric D. Beinhocker
2.1. Ambidexterity
2.1.1. Organizational Ambidexterity
The word ambidexterity derives from Latin, meaning being able to use both hands equally well. Organizational researchers used ambidexterity in regards of organizations as well. It is an abstract construct, only used by academic researchers (Birkinshaw and Gupta, 2013). The first mentioning of the construct of ambidexterity by Duncan (1976) referred to the fact that an organization is capable of pursuing opposite goals, through putting structures in place which focus on various or even opposite goals. The importance of pursuing seemingly opposite goals is especially important in fast changing environments, where organizations need to stay fit with their environment in order to keep their competitive advantages and have to adapt appropriate solutions in these circumstances (Levitt and March, 1988; Teece, 2007). Being ambidextrous ensures an organization to maintain a dynamic fit with its environment.
March (1991) introduced the twin concepts of exploration and exploitation, and argued that these two concepts shall be regarded as two extremes (Cao et al, 2009). According to his argumentation, these two concepts create paradoxical challenges and it is extremely difficult for an organization to pursue both of them at the same time, since these concepts are self-reinforcing (Birkinshaw and Gupta, 2013). The exploration-exploitation paradox lies in the fact that members of an organization have to embrace these paradoxical
tensions and simultaneously differentiate between exploitative and explorative tasks, but at the same time integrate knowledge resources of these opposite goals (Smith and Lewis, 2011, Smith and Tushman, 2005). They need to be able to transcend the seemingly controversial goals of exploitation and exploration and integrate the resources stored in various places throughout the organization to work towards the integrative goals of the organization. Transcending these tensions is problematic for organizations as their members face the challenge of reframing the coexistence of seemingly opposites (Lewis, 2000). Especially difficult in these situations is that organization members tend to favor exploitation over exploration when faced with cognitive biases (March, 1991; Levinthal and March, 1993). To illustrate the possible tensions and the arising potential confusions among members of an organization:
“I can't stand this," said a senior manager of a Standard& Poor's 500 company recently. "One minute the management team is telling us to innovate, and the next
minute they are giving us our marching orders in deploying Six Sigma. It's crazy to
tell people they should be focused on becoming more efficient while at the same time
you want them to explore untapped growth potential. This is making me nuts.” (Rae, 2007)
The management of an organization has to make certain trade-offs between exploration and exploitation, in order to find a balance between these two concepts. This approach of organizational ambidexterity is the balance dimension of ambidexterity (BD) (Cao et al, 2009). However, other researchers in the past argued that exploration and exploitation should not be regarded as two ends of ambidexterity. Rather, organizations can simultaneously achieve high levels of exploration and exploitation and they are orthogonal to
each other (Gupta et al., 2006; Jansen et al., 2006). Cao et al. (2009) refer to this as the combined dimension of ambidexterity (CD).
According to the CD approach of ambidexterity, the ability of organizations continuously focusing on both concepts enables them to stay fit with their environments (Helfat et al., 2007; Simsek, 2009). Burgelman (1991) suggests that firms should simultaneously keep induced processes, as well as autonomous processes in place in order to keep up the competition with their environment, even though this might lead to efficiency losses in both extremes. This continuous pursuit of the seemingly paradoxical concepts leads to higher firm performance (Lubatkin et al., 2006; He and Wong, 2004) and eventually to sustained competitive advantage (Crossan et al., 1999).
Many researchers have tried through various perspectives to define ambidexterity, but because of the numerous research streams a universally used definition is still not in place (O’Reilly and Tushman, 2013). Depending on the level of analysis, big part of the definitions can be classified as structural, behavioral and realized definitions (Simsek, 2009). Structural definitions mostly refer to the design set up by an organization in order to achieve ambidexterity and look at an organization’s subunits responsible for exploration and exploitation (Tushman and O’Reilly, 1996). Behavioral view describes an “organization’s or business unit’s behavioral capacity to simultaneously demonstrate alignment and adaptability across the business unit as the organizational context encourages and supports individuals in their efforts to heed both of these concerns.” (Simsek 2009, p. 602). Where the realized view refers to a state where organizations have attained both high levels of exploration and exploitation. Further classification of organizational ambidexterity research is how researchers try to categorize the way an organization achieves ambidexterity. The two major groups of categorization are structural ambidexterity (Benner and Tushman, 2003) and contextual ambidexterity (Gibson and Birkinshaw, 2004). Despite of the diverse approaches
defining ambidexterity, there is an agreement that achieving exploration as well as exploitation is beneficial from a financial perspective (He and Wong, 2004; Lubatkin et al., 2006) and in terms of organizational durability (O’Reilly and Tushman, 2011).
2.1.2. Ambidextrous TMT
Despite of the many research streams, most of the researchers agree that “ambidexterity is an organizational capability that makes it possible to resolve different tensions that arise within an organization” (Nosella et al., 2012, p. 450). As presented above, ambidexterity can be discussed on the level of the organization. However, that is not the only possible level of analysis to get further insights on ambidexterity. The ambidextrous behavior of a firm has to be present not just on the organizational level, but also at the team level. Gaining insight on ambidexterity of teams can help to better understand and analyze ambidexterity at the organizational level (Haas, 2010) and help to gain insights on the appropriate management of these tensions throughout the entire organization. That is because in more and more organizations and business units of exploration and exploitation are not only integrated on a macro level, but their teams need to focus on having to meet ambidextrous goals simultaneously.
These teams and the top management of a firm have to blend the opposite agendas of exploration and exploitation, otherwise “the firm’s orientation may remain fragmented”. (Heavey and Simsek, 2014, p. 2). TMTs play a key role in organizations that aim to become ambidextrous on organization level. Their role requires them to differentiate and integrate the seemingly opposing agendas of exploration and exploitation (Heavey and Simsek, 2014). Differentiation calls for managers’ ability of enriched awareness: being able to recognize new ways of operation, extend current solutions and look for new ones (Langer, 1989). The ability
of integration requires managers to “integrate the various agendas into a coherent strategic form” (Heavey and Simsek, 2014, p.4).
Ambidexterity is a dynamic capability where a firm has to simultaneously differentiate between exploitation and exploration and allocate resources to areas which are apparently contradictory to each other (O’Reilly and Tushman, 2008; Smith and Lewis, 2011). Referring to ambidexterity as a dynamic capability means an organization’s ability to stay fit with its environment and is manifested in the management team’s decisions, including activities such as new product development and organizational learning (O’Reilly and Tushman, 2008; Helfat and Peteraf, 2003). Since these tensions coincide to the largest extent at the level of management, management teams need to be able to reconcile these paradoxical tensions and integrate seemingly controversial factors (Lewis, 2000). “Senior teams make those decisions regarding organizational forms and resource allocation processes such that their firms might balance exploration as well as exploitation” (Smith and Tushman, 2005, p. 523). The leadership style of management teams can play an important role as well. As Nemanich and Vera (2009) suggest, transformational behaviors of team leaders may lead to a culture where learning is promoted and eventually lead to an ambidextrous team.
Teams represent a solution to firms which have the sufficient expertise to efficiently exploit previously acquired knowledge to meet business goals. At the same time teams are also exposed to continuous new experiences coming from the market and their customers, and thus have hand-on information for future development opportunities. Continuously addressing these tensions, an ambidextrous TMT is capable of simultaneously addressing the seemingly controversial constructs of exploration and exploitation to improve its current competencies and create opportunities for the future.
Nevertheless, the amount of research on how TMT eventually become ambidextrous and which capabilities managers need to demonstrate to achieve that is significantly lagging
behind compared to the research on organizational ambidexterity (Birkinshaw and Gupta, 2013). Perretti and Negro (2006) focus on the specific design of a team and argue that the level of ambidexterity of a team depends on its composition of team members. Their article is based on BD arguments. They make a distinction between newcomers and oldtimers, whose mixing and matching in a team determine the level a team can exploit its knowledge from prior experiences and explore creative solutions. Similarly, Taylor and Greve (2006) examined the potential effects of a team’s composition on the ambidexterity of a team. They argue that the recombination of various knowledge domains potentially leads to new innovations. Their research point out that ambidexterity within a team is increased by learning and exchange of knowledge among the team members (Gupta et al., 2006). Complementing the arguments of Taylor and Greve (2006), Beckman (2006) focuses in her research on the past experiences of members of a founding team. Based on her arguments, founding teams whose members previously worked at the same company share a set of common values that facilitate to efficiently implement new strategies and improve processes in place. However, members with different company affiliations bring diverse beliefs to a team which might lead to new ways of knowledge combination. In her research she found evidence and points out that there is a direct connection between the appropriate combination of founding team members and organizational ambidexterity, and ultimately to greater firm performance.
As the above brief summary of the previous research on ambidexterity suggests, it is a diversified field of academic research. The majority of academic articles have been published in connection with ambidexterity at organizational level. However, far less is known about ambidexterity at the team level. The importance of learning and the sharing of knowledge within a team has been pointed out in previous research (Taylor and Greve, 2006; Gupta et
al., 2006), yet little is known about the personality of team members and team dynamics necessary for a highly ambidextrous team.
2.2. Proactivity
The personality trait of proactivity has been broadly researched in the past, not just in organizational context. According to Bateman and Crant (1993), proactive behavior is a type of behavior in which individuals influence their environments and vice versa. There are numerous ways how individuals can behave proactively; among others seeking feedback, selecting in what situations to participate, acting in advance to form and manipulate individuals’ and groups’ opinion or building social networks. Bateman and Crant (1993) described proactive behavior as rather foreactive than counteractive, transcendent, seeking for primary control and rather as agency than passivity. They also argue that proactivity as a personality trait is connected to openness, extraversion and conscientiousness of the ‘Big Five’ personality traits.
Proactive individuals within an organization take a proactive approach towards forming their working environment, and seek information and opportunities instead of waiting it gets to them (Crant, 2000). Becherer and Maurer (1999) suggest proactive behavior is particularly appealing as an entrepreneurial habit; seeking for new opportunities, proactively changing and adapting to the environment, focusing on the future.
Despite the fact that proactive behavior has been broadly discussed and researched, there is no one universal framework that is accepted to measure proactivity. The framework introduced by Crant (2000) summarizes the most important antecedents and outcomes of proactive behavior which is based on the types of variables that have been studied previously in relation to proactive behavior. Crant (2000) points out that proactive behavior is a complex
phenomenon with numerous antecedents and possible outcomes. As he suggests, because of the multiplicity of factors influencing proactive behavior and its possible outcomes, there is still room for further research in this field. One of these potential issues is to further determine the relationship between proactive individuals in teams and team performance.
2.2.1. Proactivity and learning
The concept of organizational learning has been approached from various aspects in the past and has been defined in numerous ways, yet an ultimate definition is not in place (Fiol and Lyles, 1985). Levitt and March (1988) defined organizational learning as a concept which is history-dependent; the behavior of an organization is based on routines and is target oriented. Organizations that intentionally implement structures that facilitate the ways their employees can learn are called learning organizations and can be defined as “an organisation which facilitates the learning of all of its members and continuously transforms itself” (Pedler et el., 1989, p. 2).
Some argue that individuals within organizations are the primary entities of learning and not organizations themselves (Simon, 1991; Grant, 1996). In his article, Dodgson (1993, p. 378) points out that learning is “the highest form of adaptation, raising probability of survival in changing environments” but can also be “as a purposive quest to retain and improve competitiveness, productivity, innovativeness in uncertain technological and market circumstances”. Thus, learning can serve seemingly opposite goals, to retain current knowledge base and improve effectiveness, as well as adapt to changing environments through acquiring new ways and practices.
Learning can be achieved through two distinct ways, through so-called single-loop and double-loop learning (Argyris and Schön, 1978; Argyris, 1991). Single-loop learning
refers to the situation where individuals incorporate their previous experiences in their existing objectives and practices in order to continuously improve their operations. Through single-loop learning proactive members will seek every opportunity to improve their ways and constantly advance their solutions. Single-loop learning activities support the ability of organizational members to differentiate between exploitation and exploration activities. Propensity to engage in single-loop learning encourages organizational members to incorporate experiences from previous actions into existing norms and objectives (Vashdi et al., 2007). Consequently, organizational members are guided to focus their attention on either exploitation or exploration activities through single-loop learning, and thus are able to differentiate these activities.
On the other hand, double-loop learning enables individuals to question their previous and current routines and freely explore new ways of meeting their goals. The impact of double-loop learning can be enhanced by proactive individuals who tend to change and adapt to their environments, which requires successfully exploring new opportunities. Double-loop learning behaviors free organizational members from perspective-limiting assumption (Gray, 2007) as they are able to actively question their existing mental models. This way they can move beyond the deeply rooted assumptions and routines relating to exploitative and exploratory resources and objectives (Argyris and Schön, 1978; Lewis, 2000). Proactiveness triggers double-loop learning through stimulating organizational members to approach each other and seek for new information from other members and critically apply the newly acquired knowledge to their current routines. As proactive members engage in discussions with each other, they automatically allow themselves to escape the paralysis that often accompanies paradoxical tensions (Lewis, 2000).
Lewis (2000) argues that members of organizations have to reframe cognitive ways of perceiving these paradoxical tensions in order to transcend contradictions and come up with
ambidextrous solutions through single- and double-loop learning. In case of exploitation, where the underlying principles and the expected outcome of an activity is already proven right, single loop learning helps to incorporate proactively all the relevant new information and come up with more effective solutions to achieve the same goal. At the same time, individuals continuously have to be ready to question their current routines and their understanding of their surroundings. Since an organization’s capacity to become ambidextrous depends on the individuals’ ability to be able to transcend the ambidexterity paradox, proactive behavior among organizations members leads to increase in organizational ambidexterity. Being ready to change and search for new solutions to implement in their everyday operations helps organizations to overcome the paradoxical tensions between exploitation and explorations. Thus, the above points lead to the argumentation that:
Hypothesis 1: TMT member proactivity positively relates to organizational ambidexterity
At the same time, I argue in my thesis that having proactive managers is not enough for a TMT to achieve ambidextrous outcomes. All the experiences, information and knowledge of the individual team members have to be communicated and transformed among the team in order to effectively pursue seemingly opposite goals. Proactive team members are eager to share and seek for new information in their own environments. In order to achieve there needs to be a certain mechanism that facilitates the flow and transformation of information. I suggest that teams need a well-developed TMS to overcome these challenges.
2.3. Transactive Memory System
The concept of transactive memory was first introduced by Daniel M. Wegner, according to whom “[t]ransactive memory refers to the idea that people in continuing interpersonal relationships often develop a specialized division of labor with respect to the encoding, storage, and retrieval of information from different substantive domains (Wegner, 1987, in Hollingshead, 1998a, p. 425).”, whereas a TMS “describes the active use of transactive memory by two or more people to cooperatively store, retrieve, and communicate information.” (Lewis, 2003, p. 588). The transactive memory theory introduced by Wegner in the 1980’s has received a lot of attention ever since in many disciplines such as social psychology, management and communication. Three main characteristics of TMSs are specialization, credibility and coordination among the members of a TMS. Based on the above definition, specialized knowledge emerges among individual members (specialization), when they can rely on other members’ knowledge (credibility) and they will combine this expertise in a task-specific way to achieve common goals (coordination) (Lewis, 2003).
Based on Peltokorpi (2011, p. 17) I define organizational TMS as “a network of interdependent work groups that use each other as external cognitive aids to accomplish shared tasks”. A well-developed TMS helps TMTs to leverage the full extent of organizational members’ knowledge and can be characterized as the “division of labor for learning, remembering, and communicating team knowledge” (Lewis 2004, p. 1519). At the same time, teams might fail to fulfill their full potential in case team members fail to share their unique expertise (Nonaka and Takeuchi, 1995), which can have effect on the entire organization, especially on heavily team-based organizations. TMSs emerge following numerous iterative interchanges between members, where the simple ‘who knows what’ concept becomes a more refined and detailed construct (Lewis et al., 2005).
Since the construct of TMS has received a lot of attention since its introduction (Wegner, 1987), numerous antecedents and consequences of the construct has been identified and researched (Ren and Argote, 2011). Ren and Argote (2011) suggest that the antecedents of TMS can be categorized into three levels: team composition inputs, team level inputs and organizational level inputs. Despite the many streams and categories of research regarding TMS, the effect of personality traits of team- or organization members has not been researched in too many occasions. Pearsall and Ellis (2006) studied the connection of dispositional assertiveness between team performance through the mediating effect of TMS. They found evidence that the assertiveness of team members plays an important role since they are capable of communicating effectively throughout a team and enhance the flow of information in the team.
I suggest that managers’ proactive characteristics trigger a TMT’s TMS. Proactive individuals excel in building out social networks in their environments and actively form their environments (Thompson, 2005). This allows them to have a better understanding of team dynamics, as well as to learn fast and easily about the location of specific knowledge among team members. Social networks of proactive individuals contribute to awareness of knowledge repository within the team and can speed up the retrieval of necessary knowledge upon need. Proactive team members are ready to take control and act in case certain tasks require that. They facilitate the communication of the necessary knowledge which eventually leads to a better developed and functioning TMS in the team. They are ready to do this because proactive behavior is by definition foreactive and seeking for control and to act in advance, anticipating future actions (Bateman and Crant, 1993; Den Hartog and Belschak, 2007). This characteristic can also encourage proactive individuals to share information with the other members of their team and thus contribute to a well-functioning TMS (Thomas et al., 2010). I expect team member proactiveness to positively affect team level ambidexterity
through the mediating effect of a team’s TMS. Thus, the above points lead to the argumentation that:
Hypothesis 2: Team member proactivity positively relates to a well-developed TMS within a team.
A TMT that is aiming to become as ambidextrous as possible can potentially benefit from a well-developed TMS. A TMS enables team members to access the required knowledge of the other team members in an efficient and effective way (Argote and Ren, 2012). This does not simply mean that members can benefit of the awareness of the ‘who knows what’ concept, but since they are also aware of other members’ expertise, thus can guide knowledge seekers to experts unknown to them. The knowledge exchange occurs based on the task-expertise-people associations (Brandon and Hollingshead, 2004). Thus, TMS can serve as the basis for reconfiguration of resources in new ways (Argote and Ren, 2012). It helps to exchange knowledge in TMTs through trust and coordination (Argote and Ren, 2012, Heavey and Simsek, 2014).
Efficient knowledge retrieval is enabled through a well-developed TMS since members of a team are closer to each other in terms of path lengths (Gulati et al., 2012). Thanks to a TMS, experts can be easier identified to react to certain developments in the market, or to new product developments, which allows providing the team with a more appropriate interpretation to the occurrences. At the same time, TMS can also serve to spread newly acquired knowledge through the team and get this new information to those members who mostly need it. These two distinct directions of information flow within the team lead to effectively seeking exploratory and exploitative solutions.
TMSs also facilitate the development of new, non-overlapping knowledge rather than redundant knowledge (Hollingshead, 1998b). The new cognitive structure is greater and more distinct than the sum of individuals’ knowledge (Heavey and Simsek, 2014), and leads to better team outcomes (DeChurch and Mesmer-Magnus, 2010). Thus, the above points lead to the argumentation that:
Hypothesis 3: A well-developed TMS within a TMT positively relates to the ambidexterity of that team.
Despite the fact that a TMS of a team can facilitate to locate and access the appropriate knowledge within a team effectively and efficiently, by itself TMS is not enough for integrating exploratory and exploitative solutions. It is however the intention of the proactive team members who aim to make use of a well-developed TMS in order to establish and keep up an ambidextrous team. Although TMS has the combinative potential, by itself it does not intend to come up with new knowledge recombination. Rather helps to raise awareness of the knowledge resources and facilitates proactive team members to come up with ambidextrous resource recombination (Van Neerijnen, 2012). In this relationship TMS can be seen as a mechanism that facilitates the members of the top-management to efficiently reach out to each other and help them to strive together in a coordinated and smooth fashion. TMS enables members of a TMT to be eased from cognitive responsibilities, and allows them to access the specialized knowledge residing in the team (Brandon and Hollingshead, 2004). Members can focus on fields of their expertise, and widen their information base of their environment (Zajac et al., 2014). By providing a wider knowledge base to every member of a TMT, TMS helps to coordinate knowledge exchange which results to in facilitating the TMT to better differentiate and integrate “strategic agendas for exploration and exploitation”
(Heavey and Simsek, 2014, p. 3) TMS provides the means for coordinated action. Thus, the above points lead to the argumentation that:
Hypothesis 4: TMS partially mediates the relationship between team member proactivity and team ambidexterity within the TMT.
3. Methodology
This thesis makes use of the innovation benchmark database in order to test the set-up hypotheses, which is owned by the innovation benchmark team of the Rotterdam School of Management. The data of the benchmark project were gathered through an online survey measuring a wide range of various concepts and certain parts of it will be used for the constructs in this thesis. The online survey was sent to companies in six different innovation intensive industries. Innovation intensity was measured based on the total costs of innovation over sales. These innovation intensive industries were selected since it is argued that the hypothesized relationships are best observed in these industries (He and Wong, 2004).
The data is gathered as follows: an online survey is conducted among Dutch and German firms between 25 and 250 in employee size. The headcount of 250 employees is the widely used upper-level point for SME categorization. Firms with less than 25 employees are removed from the sample. These companies mostly lack a degree of organizational and coordinative challenges and are not very likely to have the capability to be ambidextrous, especially on the team level. The online survey was completed by the top management of the companies, with assured confidentially.
After screening the data on non-random missing values for proactivity, TMS and ambidexterity, data from 292 organizations were retained. On average, these organizations employed 91 full-time employees (s.d. = 56.9), and were 46.3 years old (s.d. = 8.7). 64.4 per cent of the organizations were located in Germany against 35.6 per cent in the Netherlands. The final sample frame covered a wide range of industries including manufacturing (3%), professional services (17%), wholesale (15%), administrative support services (7%), construction (5%), and other industries (53%).
3.1. Measurement and Validation of Constructs
All of the study’s constructs were measured on a seven-point Likert scale ranging from (1) “strongly disagree” to (7) “strongly agree”. Ambidexterity and TMS were measured by using existing scales from the literature (Appendix 1).
Ambidexterity (𝛼=0.86). As prior scholars (Gibson and Birkinshaw 2004, He and
Wong 2004) suggest, the measure for ambidexterity was developed in two steps. First, information on exploration and exploitation was captured by means of two separate scales, adopted from Jansen et al. (2006). The exploration scale measures the extent to which organizations depart from existing knowledge bases in order to pursue radical innovations for emerging customers. Example items are “We regularly experiment with new products and services in the market” and “We frequently utilize new opportunities in new markets”. The exploitation scale, contrarily, measures the extent to which organizations rely on their previously acquired expertise in order to create value as efficiently as possible for existing customers. Example items are “Our organization expands services for existing clients” and “We regularly implement small adaptations to existing products and services”.
To provide evidence of convergent and discriminant validity for exploratory and exploitative innovation, an exploratory factor analyses is performed. Exploratory factor analyses indicated that because of too high crossloadings three items from the exploitation scale had to be dropped. This resulted in a seven-item scale for exploration (α = .87) and a four-item scale for exploitation (α = .74). A joint exploratory factor analysis showed that both factors had eigenvalues greater than one. Each item furthermore clearly loaded on its intended factor. In order to capture the extent how ambidextrous a team is, previous research is followed where exploration and exploitation is considered orthogonal and the product of
exploratory and exploitative innovation is computed (Gibson & Birkinshaw, 2004, Cao et al., 2009).
Transactive Memory System (𝛼=0.78). Team level TMS measures the extent to which team members are able to make use effectively of other team members’ knowledge (Ren and Argote 2011, Wegner 1987). TMS “describes the active use of transactive memory by two or more people cooperatively store, retrieve, and communicate information.” (Lewis, 2003, p. 588). To be able to measure team level TMS, all three reflective behaviors (localization, coordination and credibility) have to be present (Van Neerijnen, 2012). Localization shows the position of expertise within the organization. Example item is “Employees have a keen awareness of the talents and skills of others”. Coordination refers to how team members combine their expertise in a task-specific way to achieve common goals. Example item is “Cooperation’s between different departments are well coordinated.” Credibility refers to team members’ reliance on other members’ knowledge. Example is “Employees in our organization have a lot of faith in other employees’ expertise.” Exploratory factor analysis showed all three expected behaviors with factor loadings above .6. The reliability scores of the three distinct behaviors of team level TMS were respectively 𝛼=0.81 for localization, 𝛼=0.91 for coordination and 𝛼=0.85 for credibility.
Proactivity (𝛼=0.86). Proactivity is a type of behavior where individuals influence
their environments and vice versa, described as rather foreactive as counteractive (Bateman and Crant, 1993). Foreactive behavior suggests that proactive team members would continuously search for new opportunities (explore) while making use of their past experiences as efficiently as possible (exploit). Proactivity of a team in a business setting is measured in terms of proactive market orientation within a management team of a firm. Example is “Our management team tries to discover additional needs of our customers of which they are unaware.”
Control Variables. In order to test the robustness of the suggested model and further
explain the relationships between the tested constructs, five control variables are included as well. First, the number of employees of the organizations was included to control for the firm size. With the growing size organizations may have a wider pool of resources, however might lose their flexibility in terms of exploration and exploitation (Ahuja and Lampert, 2001). Second, Gilbert (2005) argues that higher level of inertia might occur with the maturing of companies. Thus the natural log of companies’ age was included. Third, as part of the survey the top management of the companies were asked questions regarding the paradoxical cognition, measuring the extent of distinction between existing and new products and execute different strategies. Fourth, frame agility has been controlled for as well, measuring the extent to which companies’ management teams continuously challenge among others their companies’ long-term strategy, competitive positioning and procedures. Lastly, reflexivity has been included as a control variable, which is considered as an antecedent of ambidexterity (Lewis, 2000).
3.2. Analysis
Since level of analysis of the constructs is at the team level, measurements are created by aggregating individual answers, and thus inter-rater agreement scores are calculated first (James et al., 1984). The average inter-rater agreements for the constructs are 0.88 for exploration, 0.73 for exploitation, 0.86 for proactivity and 0.77 for team level TMS. According to James et al. (1993), values above 0.7 demonstrate a good fit.
In order to assure evidence for convergent and discriminant validity, factor analysis has been performed. During the exploratory factor analysis principal component analysis and varimax rotation was used. Based on the exploratory factor analysis a few items have been
eliminated. In case of the underlying constructs of TMS, two items from both localization and credibility and one from coordination have been removed because either of too high cross loading (higher than 0.3) or because of too low reliability scores. Because of the same criteria three items from the underlying construct of exploitation has been removed as well. As a result, Proactivity has been measured on a four-scale item. The underlying behaviors of TMS have been measured respectively on a three-item scale for localization and credibility and on a four-item scale for coordination. After removing a few items because of too high cross loadings, the two separate scales measuring Ambidexterity consisted of a seven-item scale for exploration and a four-item scale for exploitation.
3.3. Analytical procedure
In order to test the mediator hypotheses, regression analysis and analysis process for the analysis of mediator hypotheses has been used. According to Baron and Kenny (1986, p. 1176) “moderator variables specify when certain effects will hold, mediators speak to how or why such effects occur.”
Figure 1 below describes the steps that need to be followed in the testing process of mediated relationships. As the figure suggests, first step is to test whether the independent variable x has a significant direct effect on the dependent variable y. If this step is true, as second step, the independent variable x should also have a significant on the mediating variable m, and at the same time the mediating variable m should have a significant direct effect on the dependent variable y. As the final step, the dependent variable y has to be regressed both on the mediating variable m and on the independent variable x. If the relationship described in the second step is stronger than the relationship in step one (while the relationship of the first step even becomes nonsignificant), then the effect of the
independent variable x on the dependent variable y is mediated through the mediating variable m (Baron and Kenny, 1986).
Figure 1: Baron and Kenny procedure for testing for mediation (Liu and Leitner, 2012, p. 105)
In order to confirm the findings of the Baron and Kenny (1986) procedure and address its limitations, the PROCESS macro developed by Hayes (2013) has been used for testing the mediation effect of team level TMS on the relationship between proactivity and team level ambidexterity. The need for testing the hypotheses with not just one method is that, despite the wide usage of the Baron and Kenny (1986) procedure, this method has received many criticisms. Some argue that the first assumption of the model, stating that there has to be a direct link between the independent variable x and the dependent variable y, can be disregarded (Pardo and Román, 2013). Further criticism states that the model set up Baron and Kenny (1986) does quantify the one relationship that the model is testing – the
intervening effect (Hayes, 2009). Based on the model the mediating effect of the mediating variable m is only inferred in case a set of assumptions are met.
3.4. Results
The inter-correlations and the descriptive statistics are presented in Table 1 below. In order to check for multicollinearity issues variance inflation factors (VIFs) have been calculated. VIF provides an acceptable indication whether multicollinearity has an effect on the variance of regression coefficients (O’Brien, 2007). The highest VIF was 2.124 in the sample, which indicates no multicollinearity threat, since it is considerably lower than 10, which is considered to be a clear sign of multicollinearity (Hair et al., 2006).
Mean SD 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 1. Proactivity 4.73 1.22 - 2. TMS 4.98 .88 .48** - 3. Ambidexterity 22.67 8.46 .50** .48** - 4. Firm size*** 4.32 .61 .03 -.04 -.04 - 5. Firm age*** 3.82 .20 .13* .12* .02 -.08 - 6. Paradoxical cognition 4.81 1.02 .49** .44** .47 .01 .07 - 7. Frame agility 4.53 1.13 .41** .42** .39 -.02 .22** .46** - 8. Reflexivity 5.05 1.16 .51** .58** .41 .03 .19** .58** .52** - 9. Construction .05 .221 -.02 .01 -.05 -.04 .10 -.04 -.03 -.08 - 10. Manufacturing .03 .164 -.02 -.11 -.06 .05 .09 -.12* -.08 -.08 -.04 - 11. Wholesale .15 .358 .09 -.02 .03 -.11 -.02 .08 .04 .04 -.10 -.07 - 12. Professional services .17 .374 -.02 .01 -.01 -.03 -.14* -.06 -.01 -.01 -.10 -.08 -.19** - 13. Administrative support services .07 .259 -.04 .05 .07 .03 .01 .04 .05 .10 -.07 -.05 -.18* -.13* -
Table 1: Inter-correlations and descriptive statistics (N=292), *** natural log, ** p < 0.01 (2 tailed), * p < 0.05 (2 tailed)
Adjusted R squared
Dependent
variable Independent variable Beta Significance VIF
Step 1 0.25 Ambidexterity Proactivity 0.50 <0.001 1.00
Step 2 0.23 Proactivity TMS 0.49 <0.001 1.00
Step 3 0.22 Ambidexterity TMS 0.48 <0.001 1.00
Step 4 0.32 Ambidexterity Proactivity 0.35 <0.001 1.31
TMS 0.31 <0.001 1.31
Model 1 Model 2 Model 3 Model 4 Model 5
Predictor Transactive Memory
System
Transactive Memory
System Ambidexterity Ambidexterity Ambidexterity
B SE p B SE p B SE p B SE p B SE p Control variables Firm size -.046 .069 .332 -.052 .067 .268 -.043 .710 .398 -.051 .678 .297 -.039 .665 .414 Firm age -.005 .218 .922 -.012 .213 .801 -.073 2.258 .167 -.084 2.160 .098 -.081 2.112 .102 Paradoxical cognition .114 .052 .060 .057 .052 .346 .307 .534 .000 .225 .527 .000 .212 .516 .001 Frame agility .134 .044 .021 .105 .044 .065 .192 .458 .002 .150 .442 .012 .126 .435 .032 Reflexivity .452 .048 .000 .391 .048 .000 .145 .493 .034 .057 .487 .393 -.033 .507 .641 Construction .046 .194 .346 .042 .189 .376 -.018 2.014 .729 -.024 1.920 .635 -.034 1.880 .495 Manufacturing -.049 .261 .317 -.059 .255 .214 .012 2.715 819 -.003 2.591 .955 .011 2.540 .826 Wholesale -.055 .122 .269 -.069 .120 .157 -.013 1.274 .816 -.032 1.217 .529 -.016 1.194 .744 Professional services .013 .118 .798 .007 .115 .880 -.003 1.226 .954 -.011 1.169 .832 -.013 1.143 .802 Administrative support services -.010 .166 .840 .007 .162 .877 .035 1.726 .514 .059 1.652 .241 .058 1.615 .244 Independent variable Proactivity 0.22 0.41 0.00 0.31 0.41 0.00 0.26 0.41 0.00 Mediator Variable
Transactive Memory System 0.23 0.59 0.00
R2 0.37 0.40 0.28 .35 0.36
∆Adjusted R2 0.03 0.06 0.10
Table 2 presents the results of the aforementioned four-step Baron and Kenny (1986) procedure. As the first step, measures the direct effect of Proactivity on Ambidexterity. This effect turns out to be positive and significant (β=0.50, p<0.001). The next step involves whether Proactivity has a positive effect on TMS. As the results suggest, there is a positive and significant effect between these two variables (β=0.49, p<0.001). As a following step TMS’ effect on Ambidexterity is tested, and again it turns out to be positive and significant (β=.048, p<0.001). At last, Ambidexterity is regressed both on Proactivity and TMS. As the results suggest, the relationship between Proactivity and Ambidexterity turns out to be weaker (β=0.35, p<0.001), while the relationship between TMS and Ambidexterity remains positive and significant (β=0.31, p<0.001). These results suggest that there is a mediation effect of TMS on the relationship between Proactivity and Ambidexterity.
Table 3 presents the results of the hierarchical regression analyses for ambidexterity (Models 3 - 5), and TMS (Models 1 and 2). Models 1, and 3 are baseline models that contain the control variables. Model 2 introduces Proactivity as an antecedent to TMS. Model 4 investigates the direct effect between Proactivity and Ambidexterity. Model 5 examines the mediating effect of TMS in the relationship between Proactivity and Ambidexterity. According to hypothesis 1 there is a positive relationship between Proactivity and Ambidexterity. Model 4 shows that the relationship between Proactivity and Ambidexterity is positive and significant (B=0.31, SE=0.41, p<0.00), thus hypothesis 1 is supported. Hypothesis 2 proposed that there is a positive relationship between Proactivity and TMS. Model 2 shows that the relationship between Proactivity and TMS is positive and significant (B=0.22, SE=0.41, p<0.00), hypothesis 2 is supported. Moreover, the results show that the relationship between Proactivity and Ambidexterity does not become insignificant but it decreases (from B=0.31, SE=0.41, p<0.00 to
B=0.26, SE=0.41, p<0.00). This indicates that TMS partially mediates the relationship between Proactivity and Ambidexterity, therefore hypothesis 4 is supported.
To address some of the critical issues in connection with the Baron and Kenny (1986) procedure, other way of testing the mediating effect of TMS on the relationship between Proactivity and Ambidexterity is to make use of Hayes’ PROCESS tool (2013) in SPSS. The results of this analysis are presented in Table 4. The bootstrapped Sobel test was performed using confidence intervals using 1000 samples to confirm the mediating effect. As the results suggest, the findings of the PROCESS analysis confirm the findings of the Baron and Kenny four-step procedure. There is a positive and significant mediating effect of TMS (β=0.35, BCa CI 0.12-0.75). The bootstrapped Sobel-test likewise confirmed the mediating effect of TMS (Sobel z=2.23, p<0.026).
Indirect effect of Proactivity on Ambidexterity
Mediator Effect Boot SE Boot LLCI Boot ULCI
TMS 0.35 0.15 0.12 0.75
Sobel-test
Effect SE Z Significance
0.35 0.16 2.23 0.026
4. Discussion
The ambidextrous challenge faced by companies - to exploit their current expertise, but also constantly explore possible future solutions (Lewis, 2000) - has been a popular topic for press, and organizational and management research literature resulting in the publication of numerous research papers (Birkinshaw and Gupta, 2013). More and more organizations and business units of exploration and exploitation are not only integrated on a macro level through ambidextrous TMTs, but their teams need to focus on having to meet ambidextrous goals simultaneously. Gaining insights on ambidexterity of teams and incorporate them with previous research on TMTs can help to better understand and analyze ambidexterity at the organizational level (Haas, 2010). TMT plays an important role in any organization, since it is the decision-making body that sets and influences the future direction of an organization (Hambrick and Mason, 1984). The dual importance of TMTs makes further research essential. TMT has to be considered as a team that is striving for ambidexterity, but since they are the ones making decisions on macro level, organizational ambidexterity depends heavily on them.
Antecedents of ambidexterity have been researched before (Raisch and Birkinshaw, 2008), yet little is known about the personality traits of team of a TMT and team dynamics necessary for a highly ambidextrous TMT. The partial mediation model presented in this study postulates that TMS partially mediates the relationship between the proactivity of TMT members and team ambidexterity, such that teams with more proactive members are able to develop a well-functioning TMS on the team level, which ultimately helps the team to be more ambidextrous. Overall, this study not only extends our understanding about ambidexterity and its
antecedents, but also provides some practical handholds that might enable organizations to actually be able of setting up top-management teams that can become more ambidextrous.
4.1. Theoretical and Practical Implications
In my thesis I hypothesized that proactivity has a positive effect on achieving ambidexterity. I investigated how proactive members can overcome the paradoxical challenges of exploration and exploitation (Lewis, 2000) through the various ways of learning, especially through single and double-loop learning (Argyris and Schön, 1978; Argyris, 1991). In every organization individuals play an important role, especially when it comes to learning, since they are the primary entities of learning and not organizations themselves (Simon, 1991; Grant, 1996). Proactive members are able and ready to not just influence their own environments but also make use of their social networks. As Thompson (2005) suggested, proactive members excel in building out extensive networks. Proactiveness triggers members to make use of their extensive social networks and share, access new information and process the newly acquired information. This enhances both single and double-loop learning. Through single-loop learning top management is able better understand their current routines and develop a more consistent operation. At the same time double-loop learning allows them to look beyond current ways of running their business and continuously look for better resource allocations. Further investigation of the linkages between proactive personal characteristics, social networks and organizational learning can potentially expand the research field of ambidexterity.
This study furthermore examined the mediating role of TMS in overcoming the tension between exploration and exploitation and proactivity. Prior studies suggested that top managers should overcome this tension in order to contribute to the organization’s success (Smith et al.
2010). The social networks of individuals play an important role for setting up and developing a well-functioning TMS in an organization, since these networks promote trust and coordination (Argote and Ren, 2012; Heavey and Simsek, 2014), which helps to exchange knowledge among the members of a TMS. There are promising avenues for future research of TMS using an approach of social networks. This might allow to examine TMS characteristics on multiple levels within an organization (Lewis and Herndon, 2011). Looking at TMS is also important from a paradoxical perspective. In today’s world, access to information is easily made possible through the newest technologies. However, making sense of all the information at hand is challenging for any top-management of any company. Processing all the information, singling out the relevant ones and making use of it remains a task hard to overcome. A well-developed TMS, where all the members trust each other and rely on each other’s expertise, can help top-management to overcome this task and make the right decisions in terms of exploration and exploitation. Being able to process all the information and use those for the right goals can lead to successfully achieving ambidextrous goals. These ambidextrous goals lead a TMT and the entire organization to be able to stay fit with their environment, keep up with their environment and end up with competitive advantage in the long run.
4.2. Limitations and Future Research
The several limitations of this study open up avenues for future research. First, literature suggested that small and medium organizations do not have the resources to focus entirely on either exploration or exploitation (Nooteboom 1994, Voss and Voss 2013), thus this study exclusively focused on small and medium organizations. Small and medium organizations rely
on their ability to balance exploration and exploitation even more than large organization do, which makes them ideal candidates to study ambidexterity. However, focusing exclusively on small and medium organizations has implications for the external validity of this study’s findings. Future research could therefore expand this study by focusing on large organizations as well.
Second, the data in this study have been obtained in a cross-sectional manner. For that reason, the results of this study should be interpreted with caution. In that sense, it remains unclear how proactive characteristics of team members and TMS contribute to a team to become and stay ambidextrous over time. It may be that over time, the social networks set up by individual members may become so strongly established that it even prevents sharing relevant information to members outside of the network. As argued before, it takes time until a TMS is set up within a team and works efficiently. However, based on cross-sectional data it is unclear how much time it would take until a TMS can show its positive effects. Future research could therefore replicate the study after the collection of longitudinal data.
Finally, this study examined the impact of proactivity on the ambidexterity of a TMT. A measure for ambidexterity was developed in two steps. First, information about both exploration and exploitation was captured by means of two separate scales that were adopted from Jansen et al. (2006). Second, the multiplicative interaction between exploration and exploitation was computed, resulting in a measure for team level ambidexterity as a whole (Cao et al. 2009, Gupta et al. 2006). Future research could focus on how proactive members within teams influence exploration and exploitation separately. Such research would lead to a deeper understanding about the role of proactivity in achieving ambidexterity. In addition, although the scales of exploitation and exploration were proven to be valid and reliable, they remain proxies.
Nevertheless its shortcomings, this study provides several new insights about ambidexterity and its antecedents, both at organizational and team level. Most importantly, it is shown that proactive members within TMTs play an important role in achieving ambidexterity, and that this can be achieved through the support provided by a well-developed TMS. This study furthermore shows the importance of different ways of learning and how it helps to transcend the paradoxical tensions of exploration and exploitation.
