Examining the frontiers of innovation : the effect of boundary spanning vis-à-vis team performance

(1)

Examining the Frontiers of Innovation:

The effect of Boundary Spanning vis-à-vis Team Performance

Vincent Schenkeveld 10416684

Supervisor: dr. Pepijn van Neerijnen March, 2015

MSc. in Business Administration – Strategy Track Amsterdam Business School - University of Amsterdam

(2)

We love to overlook the boundaries which we do not wish to pass. - Samuel Johnson

(3)

Statement of originality

This document is written by Student: Vincent Schenkeveld 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.

(4)

Abstract

Research has shown that project leaders of new product development teams should engage in boundary spanning activities in response to the increased complexity of work tasks and changing environmental conditions. Nonetheless, empirical evidence remains thin. This study tested the impact of two boundary spanning activities on three different team performance outcomes while simultaneously exploring the effect of four moderators. A regression analysis on 31 NPD teams from a Dutch company active in the high-tech industry showed that (a) boundary spanning activities have by no means an impact on team performance, (b) ‘resource availability’ and ‘team climate’ are significant moderators, (c) two opposing facets of team climate ‘innovative’ and ‘rule-following’ have both a significant effect on team performance – providing an interesting link with contextual ambidexterity. These results represents an important contribution to understanding if NPD teams really benefit from boundary spanning activities in the high-tech industry under certain circumstances.

Keywords: Boundary Spanning, New Product Development Teams, Team Performance, Team Leader, Team Climate, Resource Availability

(7)

1. Introduction.

In today’s dynamic globalizing world one is facing ongoing change and continuous innovation, hence there is no time for organizations to rest on their laurels. Moreover while dealing with increasingly complex and innovative tasks, purely managing your own internal affairs is not enough in an ever-changing environment. Organizations need to coordinate interdependent work efforts and bridge disconnected parties by pro-actively managing their boundaries. Dorothy Leonard (1998) was a frontrunner with her statement that the most innovation happens at the boundaries between disciplines or specializations. Seventeen years later, these boundaries have become only more important, despite the fact that simultaneously they have become more difficult to manage ; since they are constantly fading, shifting and in consequence changing (Marrone, 2010). In response to this, organizations have been increasingly become team based, partially in order to anticipate quickly on change (Joshi. 2006 ; Marrone et al., 2007). Scholars rapidly found out that team performance is not merely an outcome of the internal functioning of the team, external team relationships are also immensely valuable in meeting organizational goals (Druskat & Kayes, 1999; Tsai &

Ghoshal, 1998). Those external relationships are commonly referred to as boundary

spanning.

Recent research into ‘boundary spanning’ - behaviors intended to establish relationships and interactions with external actors that can assist their team in meeting its overall objectives (Ancona & Caldwell 1992a) - has revived the notion (Faraj and Yan, 2009; Joshi et al., 2009; Marrone, 2010; Marrone et al., 2007) and has continued to build on the solid seminal work from Ancona & Caldwell (1992a). Specific team boundary behaviors may include: coordinating task activities with externals, representing the team to stakeholders, and

(8)

seeking information from outside experts (Marrone, 2007). Another important tendency that is happening simultaneously and which cannot be dissociated from aforementioned is the rising power and importance of teams (Ancona & Bresman, 2013). This is mainly due to debureaucratized, network-based and temporary structured organizations (Gibson & Gibbs, 2006). Combining the two tendencies gave birth to the concept: team boundary spanning. This is a trending and recurrent topic in the literature to find ways of dealing with, and hopefully benefiting from the ever-changing environment and meeting more effective the overall objectives of the team. The latter is not the only benefit; the individuals who carry out boundary spanning activities are often important knowledge brokers within and between organizations (Hansen, 1999) and can therefore play a crucial role in organizational learning (Carlile, 2004).

Despite the fact that researchers in recent years have generally agreed on the informational and resource benefits gained from boundary spanning which subsequently contributes positively to the team outcome (Faraj & Yan, 2009; Marrone, 2007), the group literature does not offer an extensive theoretical framework that links the different types of boundary spanning to different measurements of team outcome. Boundary spanning can be organized across three broad behavioral categories – representation, coordination and

scouting. Representation further in the text referred to as – ambassador - & coordination are

often linked as factors that increase team performance, although on a theoretical basis.

To date, empirical research has focused solely either on one type of boundary spanning, for example, probing for external knowledge (Ratcheva, 2009), or scholars consider boundary spanning as a one-dimensional construct with no distinction between the different types (Edmondson, 2003). Although the literature is crystal clear that there is discrepancy in the effects that boundary spanning has on performance; ‘ambassador’ often

(9)

referred to as beneficial while ‘scout’ might hinder performance(Ancona & Caldwell, 1992a; Keller, 2001)

This implies the need for an empirical study, in order to get a clear purification behind the division of boundary spanning and subsequently its relation to team performance. This research is designed around the notion that not all boundary-spanning activities are beneficial. Obtaining political support is an essential aspect of the performance of new product development (NPD) teams (McLoughlin et al., 2001), therefore ambassador is one of the two input variables in this model. The other variable is task coordinator, harmonizing the team with its external actors – (note that this external actors are often actors inside the organization but external to the team). Because NPD-teams are naturally linked to exploration it is used as an outcome variable next to exploitation to discover the contrast. Lastly, a general team performance variable was added since ambassador is more a generic enabler that has little to do with either exploration or exploitation. In a nutshell, this study addresses the gap – boundary spanning vis-à-vis team performance - which is caused by a theoretical snowball effect while empiricism is waiting for the bus. The gap is filled by means of a model in which two core constructs of boundary spanning are measured vis-à-vis three different team performance outcomes. Additionally, five constructs are added, because it is unclear how context dependent the hypothesized relation is and what role contingencies play. Herewith, the goal is to get a deeper understanding how this possible direct effect hold under certain circumstances.

There is consensus in previous research about the fact that boundary-spanning activities are taxing and require a great deal of effort and persistence (Ancona & Caldwell, 1990; Marrone et al, 2007). The challenge boundary spanning individuals in a team may face is the incompatible expectation from internal and external stakeholders (Marrone, 2010). This is mainly due to the fact that a team has only a limited amount of resources (e.g., time, effort,

(10)

personnel), and conducting either internal or external activities could possibly lead to a reduction of resources available for the other (Choi, 2002). Therefore boundary spanners have to perform a balancing act in order to act in moderation. Consequentially, since the balancing act is far from easy, this could lead to the erosion of the social cohesion in the team– the most common negative externality of boundary spanning. (Katz & Kahn, 1978; Marrone, 2007).

Accordingly, a pivotal role is reserved for the team leader in carrying out boundary spanning activities mainly due to the nature and complexity of those activities. There is widespread consensus in the academic literature about the essential contribution from team leaders to the overall performance of new product development (NPD) teams. (e.g.; Sarin and O’connor, 2009), part of this contribution is the ability to span boundaries and link the team to its environment (Balkundi and Kilduff, 2005). Without a championing leader who actively leads the boundary spanning practice of the team, teams may make decisions that are inconsistent with the overarching goals or constraints from the organization (Edmondson, 2003). In consequence, in this study boundary spanning is only measured by the team leaders. Despite the topicality of the subject and the recent revival, which led to an increase in academic and theoretical attention, which subsequently accelerated team boundary spanning research (Marrone, 2010; Joshi et al., 2009), recent empirical based research is surprisingly scarce. Furthermore the empirically research that is done (Marrone., et al, 2007; Faraj & Yan., 2009) is often carelessly cited in a uncritically and generalized way while industry circumstances are covered with the cloak of charity (e.g. ; Joshi et al., 2009; Golden & Veiga., 2005).

Overall, better understanding is needed about the payback of boundary spanning in different contexts and under certain conditions. This is especially the case in the high-tech industry since their NPD processes are rather complex. The high-tech industry is

(11)

characterized by uncertain, unpredictable environments with strong competitive pressure which requires continuous innovation and adaptability (Mohrman et al., 1995). Due to the highly diversified and specialized tasks, Tushman (1977) offered the insight that the optimal composition of boundary spanning in the high tech industry varied depending on the task. When facing more complex and less routine tasks , the best performing units consistently relied on boundary spanning activities. However, exactly the opposite was found for the most outstanding projects within those units. Therefore one could discourse twofold; on one hand, the complexity of the work could require above average external input and support. While on the other hand it could be too time consuming and therefore not beneficial to reach out to the external environment – as walking in a labyrinth. In order to move forward with this intriguing phenomenon: ‘boundary spanning’ it is essential to do a profound analysis on the nature of the generally assumed positive association with team performance within the high-tech industry..

Therefore the purpose of this study was to develop and test a model that would address this gap. First, I created a model that conceptualizes the two most important facets of boundary spanning as input variables on one side (measured by means of the team leader), while on the other side, the dependent variable consists of three different measurements of team performance (measured by means of the team members). Second, I included five moderating variables to investigate if the hypothesized association holds under certain team characteristics and under certain environmental factors. The theoretical model was tested with data collected from 31 new product development teams in a leading - state of the art - Dutch firm within the global high-tech industry.

This study contributes to the literature in at least three distinct ways. First this is the only study that empirically examines how boundary spanning activities from the team leader are related to the team performance outcome perceived by the team members. Marrone

(12)

(2010) suggested that boundary spanning activities are positively related to team performance. However this study did not find any significant direct relation between boundary spanning vis-à-vis team performance. Second, it has been widely suggested that there are possible moderating conditions of the boundary spanning versus team performance relationship (Choi, 2002; Faraj & Yan, 2009). This study shows that – resource availability - & - rule following team climate – are significant moderators under certain circumstances. Finally, a serendipitous finding, this research surprisingly showed that an innovative team climate as well as a rule following team climate have both a significant direct effect on team performance.

2. Theory & Hypotheses.

2.1 Boundary Spanning

Ancona & Caldwell’s (1992) seminal article is a valuable bedrock as it originated relevant concepts to refine the notion of boundary spanning, and as it served as the keystone for the few empirical articles that exist (e.g.; Faraj and Yan, 2009; Brion et al., 2012). After thoroughly gathering the primary data, they performed a factor analysis which led to the birth of four types of distinct boundary spanning activities: ‘ambassador’ mainl y protective and persuasive activities, ‘task coordinator’ the coordination with stakeholders, ‘scout’ gathering information & ‘guard’ protecting itself for outside threats.

In consonance with past empirical research this research is focusing on the two boundary spanning functions with the most distinct relationship with team performance (Brion et al., 2012; Joshi et al., 2009; Oh et al., 2004). The first is ‘ambassador’ whose core objectives are: obtaining political support – especially essential in NPD-teams (McLoughlin et al., 2001) - and ensuring the legitimacy of the project for budget, time and human

(13)

resources, literally being an ambassador of the project and keeping the overarching organization informed and satisfied. Essentially worthwhile in this research setting since it can be difficult to gain legitimacy as a NPD-team in the high-tech industry due to the unpredictable outcomes and unquantifiable timelines. Second is ‘coordinator’ who concentrates on organizing and solving issues wherefore interactions with outsiders are unavoidable. Applicable in this case because of the complexity of the problems that pop up, it is inevitable that external support is needed in order to move forward.

The division of boundary spanning is logic since the latitude of the definition, however several scholars treated boundary spanning as a single one-dimensional phenomenon (DeChurch & Marks, 2006; Edmondson, 2003). Although all different boundary spanning activities are carried out with the goal to contribute to team performance, the literature is crystal clear that there is discrepancy in the effects that boundary spanning has on performance. ‘Ambassador’ & ‘Task coordinator ’often referred to as beneficial while ‘scout’ might hinder performance; e.g. if the information received is inaccurate it could cause disruptions in team performance negatively mediated through corrosion of the social cohesion (Ancona & Caldwell, 1992a; Keller, 2001). Or teams that put too much resources in scouting activities for information, while undermining efficiency and innovation (Marrone, 2010). This implies the need for refinement and the rationale behind the division of boundary spanning because ultimately one only wants activities that are beneficial to performance.

Due to the effort that it takes to carry out boundary spanning activities, it is hard to find a consensus in the literature of the right balance of external and internal activities which can live in harmony inside a team. It is definitely not a causal relationship in terms of the more boundary spanning the better, although Ancona (1990) found out that the adverse effects of consultancy teams with the highest external outreach were in the end mitigated by the benefits gained from the team boundary spanning activities. On the occasion when

(14)

individuals are suffering, the team can still benefit. This is akin to how organizations can be benefiting while teams might suffer from boundary spanning activities. In essence, the challenge is to create boundaries that are porous enough to allow valuable resources and information to come in, but resistant enough to avoid mistrust about the exact team-composition and whether all members are equally accountable for its collective outcomes (Hackman, 2002). Marrone’s (2007) research dived deeper into this paradox and found that a team that engaged in boundary spanning performs better. Yet the individual team members carrying out boundary spanning experienced significant role overload that could possibly harm the team’s long-term viability.

Despite the fact that there is a minimum of empirical evidence, several frequently quoted papers state that there is a relationship between NPD project performance and boundary spanning. For example Brion., et al (2012 p. 709) states : ‘’ NPD project performance is strongly influenced by how teams use boundary spanning activities to access resources that are external to the team ‘’. The references used to back the statement are based on: Ancona & Caldwell (1992) though they found - in a bias sensitive study- only one significant relationship out of five different measurements of team performance outcomes; Choi (2002) who starts his implications part with : ‘’ the overall relationships proposed await further empirical testing ‘’, while his entire article is written without a single empirical observation ; Faraj & Yan (2009) did not find a direct effect between boundary spanning on team performance and was negatively linked under conditions of resource scarcity. Theoretically, there is consensus about the notion that in most of cases boundary spanning should be positively associated to team performance. However empirical results remain far behind compared to the theoretical progress that has been made since the revival of boundary spanning. This has led to a snowball effect of unjustified generalization of the little empirical boundary spanning research that exists, in spite of the fact that the most cited articles clearly

(15)

state that this relation is context dependent, thus hard to generalize (Ancona & Caldwell, 1992a; Marrone, 2010). Summarizing and emphasizing aforementioned: the relationship is not that crystal clear as several scholars have tried to indicate.

2.2 Teams

Research into teams remains strong mainly due to several studies, which clearly demonstrated the existence of the relationship between team-based organizations and new product development performance (Hoegl and Gemuenden, 2001; Gupta and Wilemon, 1996). The core benefits associated with teams are the ease of cross-functional collaborative processes plus the speed of which directions can be changed. (Brown and Eisenhardt, 1995, Ancona et al, 2012). This leaves the question: what are teams precisely?

At the beginning of the millennium Kozlowski and Bell (2003) wrote an integrative overarching article about teams in organizations. Time has taught us that this work is the foundation and standard with widely used demarcated definitions from which further research is built on. Work teams : (a) are composed of two or more individuals, (b) who exist to perform organizationally relevant tasks, (c) share one or more common goals, (d) interact socially, (e) exhibit task interdependencies, (f) maintain and manage boundaries and (g) are embedded in an organizational context that sets boundaries, constrains the team and influences exchanges with other units in the broader entity (Kozlowski & Bell, 2003). This research solely focus on project teams (e.g. New Product Development teams – NPD Teams) therefore we can specify with a more fine-grained definition. NPD- teams are characterized by (1) tasks that are externally driven, dynamic and structured by explicit workflows; (2) common goals that necessitate specific individual contributions that may shift over a work cycle; (3) roles that are specified and differentiated such that they required specific

(16)

knowledge and skill; (4) a process emphasis that focuses on task-based roles, task interaction, and performance coordination and (5) performance demands that require coordinated individual performance in real time and the capacity to continually improve over time (Kozlowski et al, 1999). All teams used in this study comply with conditions mentioned above

The current environment demands a new brand of team – one that emphasizes outreach to stakeholders and adapts easily to flatten organizational structures, changing information and increasing complexity (Ancona et al, 2002 ; Burke, Fiore & Salas, 2003). The rationale behind this strategy is to take advantage of the performance potential fundamental to teams in comparison with individuals (Schuman, 2010). Examples of advantages are efficient exchange and a flawless combination of a wide diverse set of individual resources where the outcome is often more than just the simple sum of the individuals (Schuman, 2010). According to Stough, Eom and Buckenmyer (2000) productivity, quality, and morale improves when teams are utilized. This is explained by the fact that teams produce a greater quantity of ideas and information than single individuals. Another benefit that teams have over individuals acting alone is that teamwork creates higher motivation and higher performance levels (Stough, Eom & Buckenmyer, 2000).

Organizations set up work teams with the aim of bringing together members with differentiated specializations to work on complex problems. The expectation is that there will be extensive knowledge creation and transfer across teams, organizations and even inter organizational boundaries (Argote, McEvily, & Reagans, 2003; Majchrzak et al., 2007). In order to realize the aforementioned processes – boundary spanning – is often ascribed an prominent role (Marrone, 2010).

(17)

2.3 Team leaders

The leader is the heart of the team, due to his catalyzing role; being the driver in conflict management, overseeing, governing and many other vital activities of a NPD team - one can say that leaders symbolize their teams (Balkundi and Kilduff, 2005). Several scholars made clear that the characteristics of a team leader can be an important ingredient of the contribution to the end-performance (Fisher, 2011; Sarin and McDermott, 2003). Those qualities can be expressed by the ability to span boundaries and to connect the team to its environment – a pivotal aspect of team leadership according to Balkundi and Kilduff (2005). For instance Edmondson (2003) showed that boundary spanning is crucial in hospitals operating often in multidisciplinary teams; without a clear boundary spanning vision from the leader, teams tend to act in a way that does not integrate with the bigger entity they are part of. Two core components of boundary spanning are in particular often directly linked to the tasks and responsibilities of the team leader - ‘ ambassador ‘ and ‘ coordinator ’- (Tushman & Katz, 1980; Ancona & Caldwell, 1990). For this reason, those two separate components are the independent variables in my model.

Above all, a team leader’s championing activities will positively predict team level task coordinator and ambassador activities (Joshi., et al., 2009). From the perspective of a team member boundary-spanning activities are taxing and can directly compete with internally directed activities for time and attention (Choi, 2002; Katz & Kahn, 1978). Therefore team members are unlikely to engage in those external activities unless they believe them to be a part of their responsibilities (Marrone, 2007). Resultant is that team leaders play a crucial role in either carrying out boundary spanning activities themselves or clearly delegating them to team-members. In addition, team leaders strategies regarding the

(18)

external environment provide strategic direction for the other team members and help them to better understand the external environment (Ancona, 1990).

Taken together this reasoning it suggests two corresponding notions. First, the team leader plays an extensive role in the boundary spanning activities’. Assuming there is a correlation between boundary spanning and team performance, the team leader should do it in an effective and efficient way. Second, it is not clear yet which of the four activities a project leader should focus on, and which activities he could possibly delegate to other team members.

2.4 Team Performance

The issue of how to measure the performance of a new product development team (NPD) is ongoing intense debate at the intersection of many disciplines and ideologies (Hart et al, 2003). Because of the intense debate, the methods for measuring team performance are highly fragmented. Therefore many researchers advocate the simultaneous use of several different dimensions (Hoegl and Gemendsen, 2001). This research use exploration, exploitation and general team performance as outcome variables.

Exploration and exploitation is measured apart but simultaneously since they appear to be juxtaposed, and require seemingly contradictory processes, structures and mindsets (Tempelaar, 2010). Intuitively, one would argue that NPD-teams are directly linked towards the outcome variable exploration. Therefore exploitation was added in order to explore the expected significant difference between exploration and exploitation ceteris paribus. Since the main course is boundary spanning, the explanation of exploration and exploitation will be very concise. The exploration of new possibilities including terms such as search, variation, risk taking, experimentation, play, flexibility, discovery, innovation vis-à-vis the exploitation

(19)

of old certainties includes such things as refinement, choice, production, efficiency, selection, implementation, execution (Schumpeter, 1934; March, 1991)

General team performance was used because of three reasons. First, exploration and exploitation are only measuring the extent to which the outcome of the project is innovative vis-à-vis refining an existing product (explorative vs exploitive). Therefore an NPD-team that developed a poor quality product in an inefficient way with unsatisfied team members could still score maximum points on either exploration of exploration. In order to prevent that general performance was added. Second, half of the input variable has nothing to do with either exploration or exploitation ; the function of ambassadorial activities is to either become or maintain a well-oiled machine without too much interference from outside. As a consequence of that it does not matter whether the team outcome was explorative or exploitive in nature for the relationship with ambassadorial activities. Lastly, in spite of the fragmentation of how to measure team performance, the most used outcome variable in team-level theories is team effectiveness (Choi, 2002) which has strong similarities with my general team performance variable.

2.5 Hypotheses

Strong support exists for the prominent role of team leaders in either carrying out boundary spanning activities or clearly delegating them to other team members while actively supporting them (Marrone, 2007). Notwithstanding the fact that team leaders are pivotal in boundary spanning, it is not clear from previous research which type of boundary spanning activity they should focus in order maximize the contribution to the team performance. Previous scholars found empirical evidence, but treating boundary spanning as a homogenous construct. Selected examples include the qualitative study on consultancy team by Ancona (1990), which clearly displayed that the quality of the team leaders was strongly related to a

(20)

cohesive external strategy which subsequently led to a higher performance. Edmondson (2003) demonstrated in a hospital – multidisciplinary teams in optima forma – that the team leader is paramount in regards to ensure the team following a commonly agreed line in relation to the external environment. Hence, conceptualizing boundary spanning as multi-faceted (see paragraph 2.1), it shows the gap and creates the necessity for empirical testing to identify which parts of boundary spanning contribute the most to team performance.

Since team leaders personify the project to the rest of the organization and to the outside word (Joshi et al., 2008), they are notably important in the resource and information flows circulating around the team and between the team and the organization (Brion et al, 2012) – also referred to as task coordinator. Similarly, the second boundary spanning activity outlined in the theory section – ambassador – is naturally a typical feature of team leadership. If there is someone that needs to take care of clear limits regarding external demands – potentially reducing the commitment from that person to the project – it is the team leader. The team leader is the only one capable of intervening and negotiating with top management in order to create or maintain fertile ground for the team (Brion et al., 2012).

Taken together, the core arguments outlined above suggests two complementary notions. First, if one assume that boundary spanning is correlating with team performance then the role of the team leader in boundary spanning - presumably ambassadorial & task

coordinator activities - is paramount. Second, a comprehensive investigation of this

explanatory is needed, because there is no empirical background with regard to the type of boundary spanning activity with the highest impact on team performance. These arguments led to the formulation of the primary hypothesis.

H1 - The extent to which the team leader engages in boundary spanning activities is positively associated with the team members’ ratings of team performance

(21)

In order to get a fine-grained analysis, the main hypothesis is split up into six sub-hypotheses as argued before in the theory section.

1a. The ambassador activities are positively associated with exploration. 1b. The task coordinator activities are positively associated with exploration 1c. The ambassador activities are positively associated with exploitation. 1d. The task coordinator activities are positively associated with exploitation

1e. The ambassador activities are positively associated with general team performance. 1f. The task coordinator activities are positively associated with general team performance.

2.6 Moderators

So far most research has focused on antecedents that could directly influence team boundary spanning outcomes (Joshi et al., 2009; Choi., 2002). Team level, task-based & contextual antecedents are all well represented in the academic discourse; terms such as: ‘team interdependence’, ‘resource availability’, ‘team heterogeneity’ are often referred to as possible antecedents. This study propose that several interesting constructs, generally used as antecedents can also strengthen or weaken the relationship boundary spanning vis-à-vis team performance. Considering that surprisingly, little empirical attention has been paid thus far to possible moderating conditions on the team boundary spanning – team performance relationship (Joshi et al., 2009 & Marrone, 2010). To my best knowledge, the solitary empirical study done that includes moderators, is the study by Faraj & Jan (2009) who found that when teams don’t face resource scarcity, boundary spanning increased team performance and vice versa. Supporting to this Tsai (2002) non-empirically implied under conditions of

(22)

resource scarcity, ambassador activities may become more critical to lobby for resources and the activities would therefore further increase instead of decrease.

Four moderators were selected: First group – team climate and team heterogeneity – are both related to the qualities of the team as – resource availability and team

interdependence - are both related to the contextual surroundings of the team.

2.6.1 Qualities of the team

To establish a more lucid picture of the relation between boundary spanning and team performance it is relevant to consider how this relationship is altered by factors stemming from the teams’ qualities. One particular area that does not have cause to complain about academic attention is the team climate. This construct has been general articulated as the practices, shared beliefs, and value systems that a team follows and that is influential to the behaviors of its members (Anderson & West, 1998). Team climate sets the tone for the team members on how to carry out their activities, including boundary spanning. For this reason, this study add two opposing team climate constructs as moderators in the model, namely,

rule-following climate and innovative climate.

An innovative climate promotes the creation of novel ideas and risk taking approaches (O’Reilly & Tushman, 2004), therefore this is close inclined towards the outcome variable exploration. Hence, the proposition:

H2a: An innovative team climate will moderate the relationship between boundary spanning and team performance – specifically exploration.

(23)

On the contrary, a rule-following climate embraces formal systems, established rules and prescribed procedures within the team (Anderson & West, 1998). The members of the team are directed to follow the rules and work in accordingly with the written policies (Patterson et al., 2005). In consequence, a rule-following climate where the focus is on efficiency, productivity, refinement and risk reducing processes can be linked towards exploitation (March, 1996). Therefore, the hypothesis:

H2b: A rule-following team climate will moderate the relationship between boundary spanning and team performance – specifically exploitation.

The composition of a team, which is an accumulation from the attributes of the individual team members, can be considered as a team-level construct (Klein & Kozlowski, 2000) that shapes boundary spanning (Ancona & Caldwell, 1992). Especially, the functional diversity of a team has been surrounded by academic interest. Ancona & Caldwell (1992) found that the more heterogeneous the team was, the more frequently it engaged in boundary spanning. Although on the other hand one could argue if one assume a team was perfectly heterogenic than there would be less need for spanning boundaries, especially the task-coordinator activities, because the majority of the resources are already in place within the team. Therefore I propose

H3: Team Heterogeneity will moderate the relationship between boundary spanning and team performance.

(24)

2.6.2 Contextual surroundings of the team.

Interdependence is considered a fundamental element of a team. Marks et al. (2001) describe interdependence as ‘’ interdependent acts that convert inputs to outcomes through cognitive, verbal and behavioral activities directed toward organizing task work to achieve collective goals ‘’. Kozlowski & Ilgen (2006) putting forward the notion that interdependence is pre-eminently a determinant if a group of people can qualify for the term team – some scholars go even further by suggesting that without interdependency there can be no team (Salas et al., 1992). Past research on team effectiveness found that the level of interdependence – used as an antecedent - within a team can be a relevant characteristic and predictor for team effectiveness (Campion et al., 1996). In a totally different setting regarding school psychologists; Amedore and Knoff (1993) found out that teams that are highly interdependent while dealing with uncertain and complex tasks are more likely to engage in boundary spanning activities in order to achieve team goals. However, theoretically one could also argue vice versa, that how much more interdependent a team is, the less it could be dependent on the external environment. Despite this fact I propose:

H4 Team Interdependency will moderate the relationship between boundary spanning and team performance.

Resource availability is the extent to which available resources are sufficient to support the

sustained growth or survival of the team in the overarching organizational entity (Aldrich. 2007). The literature with regard to resource availability - linked to the boundary spanning – team performance notion - is ambivalent, either as an antecedent or an moderator. In environments of low resource availability individuals display higher levels of scout activity

(25)

that thereafter is often negatively associated with team performance (Ancona & Caldwell, 1992a). This is dissimilar with the findings from Faraj and Yan (2009). They found - while perceiving enough resources in the team – a positive relationship between boundary spanning and team performance. Interestingly, ambassadorial activities involve persuading upper management to support the team, and if necessary, provide the team with extra resources (Ancona, 1990). Taking into account the high-tech industry, I expect that the team leader will more actively engage in boundary spanning activities if teams face resource scarcity, than teams in a resource abundant situation (Choi, 2002). Acute need for additional resources can create a shared sense of urgency among all team members and encourage especially the leader to look for ad-hoc opportunities, internally as well as externally (Faraj & Yan, 2009). Complementary to this one can imagine that when resources are relatively abundant, boundary spanning activities could be perceived as irrelevant by the other team members. Which could ultimately lead to the erosion of the social team cohesion. Concluding the proposition

H4 Resource availability will moderate the relationship between boundary spanning and team performance, especially the primary sub-hypotheses with ambassador as an input

variable

2.7 Conceptual Model

The aforementioned hypotheses and theory led ultimately to the conceptual model which is shown on the following page (Figure 1). The straight lines are the direct effects while the dotted lines are the interaction effects. When the interaction effect is significant the line

transformed in a straight line ; green for a positive interaction effect whilst red accounts for a

negative interaction effect. From the result section onwards it will be discussed in more depth.

(26)

(27)

3. Methodology and Data.

In this chapter the method of data collection and the level of analysis are described. Moreover, the measurement and validation of the constructs is presented, which will be used to answer the research questions.

3.1 Methodology

This research is an explaining study, which tries to observe relationships between boundary spanning activities from the team leader and team performance outcomes from the team members. Secondly, it tries to observe moderators for this relationship; team heterogeneity, team interdependence, team climate and resource availability are all adapted into the model to discover, if they have a significant moderating role on the on the proposed main effect of the study: boundary spanning vis-à-vis team performance. By means of a quantitative approach, the hypothesized constructs can be tested on a large collection of teams and is therefore considered to be a suitable and valid approach (Denzin and Lincoln 2008).

Two questionnaires have been given to the teams, one for the team members and another for the team leaders of the same team. Participants were asked to answer questions on a seven point Likert scale, ranging from ‘fully agree’ to ‘fully disagree’, furthermore there were several open questions to define general respondent characteristics. Both questionnaires can be found in Appendix 1.

This study contributes and is part of an ongoing research project at the Erasmus Research Institute of Management and the University of Amsterdam named : ‘ Innovation

Excellence ’. The main goal of the overarching study is trying to identify enablers and

(28)

teams within their respective organizations in the Dutch high tech industry. The data for this study was collected from a Dutch state of the art company active in the high-tech industry. In 2014 this company spent more than 1 billion solely on research and development. Therefore this explanatory research can give a unique look behind the scenes in the high-tech industry

3.2 Level of Analysis

A big part of the research done at exploring boundary spanning has been conducted within two separate streams of research, each concentrating at a different level of analysis and remaining largely unconnected to the other (Marrone, 2007). One stream is focusing on the team level and has identified strategies to manage its external environment – boundary spanning activities –, which contributed to team effectiveness (Ancona, 1990; Ancona & Caldwell, 1992). This is in contradiction with earlier research, mainly from a social-psychological angle, which showed that those who carry out the boundary spanning activities – often team leaders - often experiencing significant role overload by conflicting pressures (Katz & Kahn, 1978). This research is trying to overcome the seemingly contradictory sets of findings by measuring boundary spanning – the independent variable - through the team leaders while measuring team outcome – the dependent variable – through the lens of the team members.

The leader and members are all part of a New Product Development (‘NPD’) team. These teams vary in size, age in terms of matureness and composition. Each team has a leader – a vital task - due to their ascendency and the big contribution to the overall performance of the team. The data gathering for this study has been focusing on mono-disciplinary expert teams because they are clearly defined and visible within the organization and they are more stable and sustainable than project teams where the expertise of various fields is combined.

(29)

3.3 Reliability and Validity

For a scientific and accurate research process, the validity and reliability of the variables and corresponding scales should meet certain requirements. The core constructs of this study were measured using multi-item scales that were validated in previous studies (e.g. Ancona and Caldwell, 1992; Faraj and Yan, 2009). Correspondingly the following constructs are measured in this study: boundary spanning, team outcome (exploration & exploitation), team performance, team heterogeneity, team interdependence, team climate (rule following and innovative) and resource availability. Figure 2 represents a coherent and concise overview of the variables and their corresponding scales.

3.4 Data Analysis

The sample for this research consisted of 31 NPD Teams (N=31) with a response rate of 97% - one team leader decided for undisclosed reasons not to co-operate with the research. Both team leaders and team members participated in this study; in total 173 employees responded to the questionnaire, compromising of 31 team leaders and 142 team members. The size of the participating teams ranged from 3 to 18 team members ( = 6.74, σ = 2.83) and the average response rate per team was 72% (σ = 19%).

3.4.2 Independent Variables

Following Ancona & Caldwell’s (1992) seminal measurement approach for boundary

spanning; the 31 team leaders responded to seven boundary spanning activity items, which

were factor analyzed to inquire the underlying constructs. Since the imprecise and open definition of the term ‘boundary spanning’ as aforementioned in the theory section this was necessary to detect the particular types of boundary activity. Team leaders’ ratings to the

(30)

extent they engaged in the 7 boundary spanning activities were analyzed with a principal component analysis and a varimax rotation. The factor analysis table about the total variance shows two factors with eigenvalues greater than 1.0 that explained 61.7 percent of the total variance. Inspection of the scree plot supported the two-factor solution. Finally the team boundary spanning variable score was the aggregated mean of the items formed together the construct. The items 1, 3, 5, 6 and 7 with a Cronbach’s α of .72 formed together the construct: ‘task coordinator’ while items 2 and 4 combined represents the ‘task ambassador’ construct with a Cronbach’s α of .81. Most work to date in team boundary spanning literature has focused on the coordinator and ambassador task, the scales and findings are totally consistent with the leading-edge study of Ancona and Caldwell (1992), which stand the test of time so far, establishing itself as the standard and providing comprehensive coverage of the primary boundary spanning activities (Marrone, 2010). Skewness (.507 and -.033) and kurtosis (-.257 and -.486) indicated a normal distribution of the data that represents the independent variable in my model.

3.4.3 Dependent Variables

The three dependent variables: exploration, exploitation and team performance were part of the surveys of both groups. The team leaders and the team members were asked identical questions, however since the independent variable came solely from the leaders, in order to enhance the validity and reliability of this research only the answers from the members are incorporated. Hence the research did not fall prey to ‘common method bias’ since the dependent and independent variable are acquired from different sources (Podsakoff et al., 2003). Additionally, the research is preserved from illusory superiority –a cognitive bias whereby individuals overestimate their own qualities and abilities. Rating outcomes of the team one is leading would be predisposed for this type of bias (Messick et al., 2013). The

(31)

Cronbach’s α of exploration is under the ‘urban legend’ of .70 (Lance et al., 2006), but since the items are not strongly correlated and the notion of exploration with his wide definition is only measured by 4 items(which negatively affects the Alpha); the α of .65 is acceptable in this case. The skewness and kurtosis levels indicated a normal distribution of all three variables that are responsible for the dependent part of the model (see figure 2).

In line with the majority of research, perceptual measures are used mainly because of the unwillingness of managers to provide objective measures of performance (Nakata and Im, 2010).

3.4.4 Moderating Variables

For all moderating variables the data was collected from team members, for Team

heterogeneity the Cronbach α was .67 after deleting one item – regarding their own unique

knowledge and expertise that no other team member had -. All four remaining items were measuring exactly the same construct, only from a team perspective. Since the moderating

variables are team constructs the item was deleted in order to increase the Cronbach α. All

skewness and kurtosis values of the moderating variables assumed that the data is normal distributed (see figure 2)

3.4.5. Control Variables

This study also looked for possible alternative explanations by including control variables – mostly following other studies. Team size, a fixed resource that could influence individuals’ abilities to carry out certain behaviors (Marrone, 2007) was measured with the natural logarithm of the people in the project team because the initial kurtosis was too high (2.91). Organizational tenure (months worked in industry) is an individual attribute that can facilitate

(32)

broad organizational knowledge and develop emotional attachment to the industry which could promote external activities laterally (Joshi et al., 2009). Comparable was the high kurtosis for team age (3.77) & months worked in industry (2.75), therefore these variables are transformed with the natural logarithm as well. The project life cycle stage was transformed from a categorical variable into a quantitative interval variable since all stages were equally sequent. Joshi et al (2009) suggested that the stage where a team is located could influence the boundary spanning activities.

Figure 2: Summary of variables and constructs Variable

Type Construct Items Valid N Cronbach’s α Mean σ Skewness Kurtosis Dependent Exploration 4 30 .647 4.54 0.49 .290 .122 Exploitation 5 31 .756 4.73 0.58 .751 .630 Team performance 4 31 .804 5.36 0.45 -.513 .026 Independent Boundary Spanning - Coordinator 5 31 .722 4.89 0.82 .507 -.257 Boundary Spanning - Ambassador 2 30 .812 5.65 0.62 -.033 -.486 Moderating Team Heterogeneity 4 31 .668 5.44 0.48 -.285 -.962 Team Interdependence 4 31 .891 4.69 0.51 -.573 -.135 Team Climate - Rule

Following 5 31 .772 3.83 0.57 -.053 .544 Team Climate - Innovative 6 31 .831 5.04 0.51 -.904 .674 Resource availability 4 31 .712 3.43 0.49 -.596 .908 Control Months in Organization 1 31 n.a. 4.48 0.55 -.285 -.488 Team size* 1 31 n.a. 2.02 0.44 .523 .753 Team age* 1 27 n.a. 3.07 0.76 -.091 -.181 Project life Cycle* 1 31 n.a. 3.32 0.86 .105 -.113

(33)

4. Analysis and Results

4.1 Descriptive Statistics

The descriptive statistics are presented in figure 3. Due to the relatively high numbers for skewness and kurtosis in relation to team tenure, 63% of the teams are two years old or younger. The leaders of the teams are relatively experienced with over 10 years of industry experience and have a solid academic background, 75% have obtained a master degree and a quarter of them continued in academic latitudes and received a PhD. The ‘ambassador’ part of boundary spanning is carried out more while the overall general ‘team performance’ scores best of all outcome variables. In line with expectations, team members experience an ‘innovative’ team climate that consists out of ‘heterogeneous’ teams. In view of the high-tech, complex industry they are in, those are expected contextual surrounding factors. To rule out the potential that regression estimates were adversely affected by the combined magnitude of several correlations this study tested for multicollinearity. Accordingly, the variance inflation factors (VIF) and tolerance statistic were calculated from which the highest score was 3.90 (1/0.256) of tolerance level), which is below the rule of thumb of 10 (Field, 2009), illustrating no concern for multicollinearity.

The correlation coefficient of innovative team climate and general performance was relatively high (.66). Therefore a principal component analysis and varimax rotation was used with all the items from this two constructs to see if there is a lack of discriminant validity. The factor analysis table about the total variance shows two factors with eigenvalues greater than 1.0 that explained 65.19 percent of the total variance. With the Kaiser-Meyer-Olkin

(34)

Figure 3 Pearson’s correlation between variables.

Pearson's correlation coefficients

1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 Exploration - 2 Exploitation -.103 - 3 Performance -.302 -.002 - 4 Coordinator -.097 .407* -.085 - 5 Ambassador .006 .121 -.088 .263 - 6 Team Heterogeneity -.193 -.101 .296 .085 .030 - 7 Team Interdependence .322 -.186 .372 -.031 .213 -.008 - 8 Rule Following -.241 .482* .087 .011 -.060 -.140 -.334 - 9 Innovative -.107 -.178 . 656** .001 .001 .238 .487* -.244 - 10 Resource availability -.450* .302 -.043 .193 .028 .075 -.496* .553** -.147 - 11 Months in organization .375 -.026 -.371 .231 .226 -.012 .015 .008 -.410* -.153 - 12 Team size* .198 .202 -.289 .241 .021 -.115 .148 .068 -.186 .210 .243 - 13 Team age* -.310 .184 .279 -.254 .205 -.252 .268 .127 .151 .034 -.107 -.167 - 14 Project Life Cycle .128 -.099 .077 .035 .041 -.069 -.285 -.098 .168 .074 -.136 -.26 -.039 -

*Log

**. Correlation is significant at the 0.01 level (2tailed). *. Correlation is significant at the 0.05 level (2-tailed).

(35)

4.2. Testing hypotheses

Because the absence of multicollinearity, all hypotheses have been tested in a number of sequential models. Because the necessity to split boundary spanning into two namely : ‘coordinator’ & ‘ambassador’ plus the three different team outcome variables, six different regressions have been carried out – each regression consists out of 9 different models. The build-up for all 6 regressions are identical. Model 1 displays the baseline model with only the control variables. Model 2 presents the main effect; the influence of boundary spanning on team outcome. Model 3 adds all five moderators but as a direct effect. Model’s 4,5,6,7 and 8 all tested a single ‘isolated’ moderator as interaction variable, whereas model 9 includes all moderators and affiliated interaction variables.

4.2.1 Boundary spanning – Exploration

There is no significant relationship between ambassadorial boundary spanning activities and

exploration. Only model 9 where all interaction variables are included is significant (Adj.𝑟𝑟2

.444, p < .01). With team interdependence (β =.971, p < .01) as most denoting positive influence, while innovative team climate (β = - 686, p < .01) has the most significant negative impact on exploration. Despite the fact that there is no significant direct effect between ambassador and exploration, there is a significant interaction effect team interdependence (β = .754, p < .05).

There is a fragile negative non-significant direct relation between coordinator and

exploration (β = -.297, p <.10) in model 2 (𝑟𝑟2_{.366, p < .05) where the tenure of the team} leader in the organization is the most influential effect, although inserted as a control variable

(β = .400, p <,01). Model 8 is significant (Adj.𝑟𝑟2_{.444, p < .01) with team interdependence (β}

(36)

with a moderating role reserved for resource availability (β = .730, p <,01) despite the fact that there is no direct main effect.

Figure 4

Figure 4 is a visual display of model 8 and shows that task coordinator activities is positively related to exploration under conditions where resource availability is high, but negatively related to exploration under conditions where resource availability is low. This result suggests that in a resource abundant environment, exploration is enhanced through task coordinator activities. However, when resources are scarce, those activities may consume effort to the detriment of the task. The significant direct positive effect of team interdependence is totally in line with the expectations, however the finding of the significant negative direct effect of an innovative team climate is absolutely unexpected and will be further elaborated in the discussion. Also worth mentioning are the control variables in this model; team age is significantly negatively related with exploration while project life cycle is significant positively related. The last might look counter intuitive but taking into consideration that exploration was measured in the way of how explorative the new product was, while a lot of projects were in the initial stages they could not judge yet if the potential

3 3.5 4 4.5 5 5.5 6

Low Coordinator High Coordinator

E xp lor at ion Low Resource availability High Resource availability

(37)

novel product was ‘explorative’. Model 9 gives grosso modo similar results as model 8, regardless of adding all interaction variables, and resource availability is still breaking through the clutter (β = .844, p <,01).

4.2.2 Boundary spanning – Exploitation

There is unambiguously no significant relationship at all between any of the variables used in the relation boundary spanning and exploitation. Nevertheless it is worth mentioning is that although the model as such is not significant, the interaction effect of resource availability in the coordinator model closely corresponds with the exploration model. (β = .706, p <,05). Those findings are largely related to the fact that nearly all teams from the sample were in a premature stage and therefore highly explorative instead of exploitive.

4.2.3 Boundary spanning – Team performance

All models except 1 and 2 are significant (p <.001) for both types of boundary spanning activities with team climate having a pivotal role. If there is a thought-provoking finding in this thesis than it is the following. Both ‘innovative’ with an average β of .844 (p <.001) as ‘rule following’ with an average β of .419 (p <.01) have a positive significant effect on general team performance. It is interesting because it implies that a mild team climate would have a negative direct effect on general team performance.

Reiteratively, there is no direct effect of ambassadorial activities on team performance, yet two negative interaction effects have been found: ‘rule following’ (β = -.237, p <.05) & ‘resource availability’ (β = -.323, p <.05). Due to the non-existing relationship between boundary spanning and team performance in this study – the results from coordinator are in line with the ambassadorial results described aforementioned.

(38)

Figure 5

Figure 5 shows model 6, the interaction effect of a rule-following team climate. In this model rule-following itself is significant positively related to team performance (β = .463, p <.001) while at the same time ambassadorial activities are positively related to performance under conditions of a low rule following team climate, but negatively related to performance in a team climate where rule following is high.

3 3.5 4 4.5 5 5.5 6

Low Ambassador High Ambassador

P

er

for

m

an

ce Low RuleFollowing

High Rule Following 3 3.5 4 4.5 5 5.5 6

Low Coordinator High Coordinator

P

er

for

m

an

ce Low RuleFollowing

High Rule Following

(39)

Figure 6 shows exactly the opposite, if one only switch from ambassador to coordinator ceteris paribus, one can see that coordinator activities are positively related to performance in a high rule following team climate, but negatively related to performance in a low rule following team climate.

Figure 7

Figure 7 is a visual display of model 8 and shows the interaction effect of resource availability which is interestingly yet surprisingly quite the opposite from the earlier interaction effect of resource availability. Interesting because the ambiguity can also be prominently found in the academic literature. In this case ambassadorial activities from the team leader are positively associated to performance under conditions where resource availability is low and negatively related to performance under conditions where resource availability is high.

4.3. Results of regression analysis

The following six consecutive landscape pages show all the results of the 54 models. 3 3.5 4 4.5 5 5.5 6

Low Ambassador High Ambassador

P

er

for

m

an

ce Low ResourceAvailability

High Resource Availability

(40)

(41)

(42)

(43)

(44)

(45)

Performance ₁ ₂ ₃ ₄ ₅ ₆ ₇ ₈ 9 Independent variables Ambassador Coordinator -.050 -.048 -.049 .136 -.061 -.064 -.054 -.014 Moderating variables Team Heterogeneity .215† .210† .202† .174† .218† .233* .258* Team Interdependence -.013 -.014 -.077 -.099 -.015 -.064 -.268†

Team Climate - Rule Following .383** .385** .385** .395** .386** .384** .430***

Team Climate - Innovative .787*** .791*** .824*** .855*** .789*** .834*** .982***

Resource availability -.206 -.213 -.270* -.340* -.209 -.243† -.463*

Interaction effects

Coordinator* Team Heterogeneity .016 -.043

Coordinator* Team Interdependence -.240† -.317†

Coordinator*TC Rule-Following .238* .144

Coordinator* TC-Innovative .022 .274

Coordinator* Resource availability -.105 -.193

Control Variables

Months in Organization^ .023 .026 -.015 -.015 -.017 -.015 -.013 .014 .053

Team size^ -.043 -.033 .007 .006 .021 .035 .007 .010 .042

Team age^ .217 .203 .115 .112 .194† .035 .111 .160 .203

Project life Cycle .222 .222 .059 .062 .083 .095 .054 .035 -.005

R2 .094 .096 .844 .844 .871 .880 .844 .847 .911 Adjusted R2 -.046 -.085 .765 .753 .796 .811 .753 .759 .822 Δ Adjusted R2 -.039 .850 -.003 .043 .015 -.058 .006 .063 F-Value .673 .531 10.783*** 9.326*** 11.650*** 12.681*** 9.326*** 9.590*** 10.210*** Durbin-Watson 2.523 2.577 2.797 2.493 2.582 2.436 ^ = Log † p < .10; * p < .05; ** p < .01; *** p < .001

(46)

5. Discussion

Scholars in the past have shown that team leaders actions’ drive NPD performance. In this study the focus is on boundary spanning activities. Because no research has explored a fine grained analysis about which types of boundary spanning activities relates to which type of team performance. The results did not provide empirical support for the notion that NPD team performance is related to boundary spanning.

Since Ancona’s (1992) pioneering work, there has been a growing recognition that teams do not live in a vacuum and that external boundary spanning activities could be an important predictor of team performance (Joshi et al., 2009, Marrone et al., 2007). Especially taking in to consideration the present and future context of team based, knowledge focused, and complex organizations facing fast paced dynamic environments (Schreyögg & Sydow, 2010). Taking into account the present context, one would expect an association between boundary spanning vis–á-vis team performance, yet a solid empirical foundation is missing. However the main finding of this is important for two reasons. First, I provide empirical evidence that the hypothesized theoretical relationship between boundary spanning and team performance is definitely as straightforward as often assumed (Choi, 2002; Brion et al.,2012), Second, this research is a vital contribution and answer to the empirical scarcity in the academic world regarding boundary work, particularly in the high-tech industry.

Nevertheless, in the one and only auxiliary study which tries to assess the relationship between boundary spanning and performance in the high-tech industry, therefore most similar to my study found for both – ambassadorial and task coordinator – only one significant positive relation out of the five different performance outcome measurements;

ambassador with adherence to budgets & schedules and task-coordinator with innovation

(47)

spanning variables as well as four out of the five team performance variables are only measured through the eyes of the team leaders. On top of that the frequency of boundary spanning activities were highly negatively (p < .001) related to the team member rating of performance.

The other comparable study, as opposed to practically all-empirical boundary spanning research in the past, Brion et al (2012) looked also specifically at boundary spanning activities of team leaders vis-á-vis NPD outcomes in line with my research. The only positive relation they found was between ‘ambassadorial ‘ activities and new product success – built up from commercial and financial components (β = .258, p < .05). Whereas ‘coordinator’ activities was totally unrelated to NPD outcome with a N of 73 project leaders in the French manufacturing industry.

Where at first sight the findings seemed odd, after a thorough investigation of the specialized literature that corresponds most closely with my research, the results were broadly speaking consistent with existing literature. If one also takes into consideration a possible publication bias – the tendency of researchers and editors to publish only the results that are positive (showing a significant finding) – which is leading to a misleading bias in the academic literature especially in meta analyses (Peters et al., 2006) - , worthwhile since the two most cited recent published articles about boundary spanning are meta analyses (Joshi et al. 2009; Marrone, 2010). One could conclude two things. First, the difference in progress since the revival of the notion of boundary spanning is substantial between purely theoretical and empirical work. Second, the negligence of several scholars regarding the citations in respect to boundary spanning is far from optimal. There is currently an academic road being paved with ‘the benefits of boundary spanning’ as a non-significant, non-stable foundation.

Notwithstanding the fact that both existing studies using team leaders showed that ‘task coordinator ’ does not impact performance, this does not automatically imply that this

Examining the frontiers of innovation : the effect of boundary spanning vis-à-vis team performance