Global customer team design: dimensions, determinants and performance outcomes
Abstract
Designing and implementing global customer teams (GCTs) represents a key task for suppliers that are expanding the scope of their customer relationships. However, research has not provided an explanation of how these teams function and what determines their performance. Using an interdisciplinary combination of concepts from customer management and organizational behaviour research streams, we develop an integrative framework of GCT design and performance. The framework is conceptualized with qualitative interview data and validated with survey data from 273 members of 113 GCTs in six multi-national companies. Our results indicate that team performance is influenced directly by three team processes: communication and collaboration, conflict management and proactiveness. Team design in terms of goal and role definition, customer coverage, empowerment, heterogeneity, skills adequacy and leadership indirectly influences performance, mediated by team processes. In addition, three factors of the organizational environment – top-management support, rewards and incentives, and training – have similar indirect effects.
Introduction
Driven by the search for both new business opportunities and competitive advantages, companies in business-to-business markets increasingly have moved away from transactional forms of exchange (Dyer 1997) to look for closer, more collaborative relationships with their customers (Cannon and Perreault 1999; Heide and John 1990; Narayandas and Rangan 2004). A common approach to exploiting the potential of long-term supplier–customer relationships (Anderson and Weitz 1992) has been to adopt various customer management techniques, such as relationship management (Subramani and Venkatraman 2003), relationship marketing (Grönroos 1994; Morgan and Hunt 1994; Webster 1992), or national and key account management (McDonald, Millman and Rogers 1997; Shapiro and Moriarty 1984; Weilbaker and Weeks 1997).
However, these approaches often become more challenging as customers become more global and powerful. With global expansion, customers establish a direct presence in more and more countries and simultaneously expect the supplier to provide consistent, coordinated service worldwide, which entails moving away from traditional relationships with local subsidiaries and towards uniform prices, terms and service in markets in which the supplier may lack operations (Montgomery and Yip 2000). Furthermore, these customers recognize the strategic value-adding potential of global procurement (Cohen and Huchzermeier 1999; Ellram and Carr 1994) and therefore adopt integrated centralized purchasing practices (Olsen and Ellram 1997; Sheth and Sharma 1997) to reduce their supplier base (Capon 2001). The resulting shift of power to the customer increases through industry consolidation (Birkinshaw et al. 2001) and advances in information and communication technologies, which enable customers to track suppliers' quality and prices globally (Narayandas et al. 2000). These factors have heightened the challenges facing suppliers and made international approaches, such as global customer management (GCM) (Birkinshaw et al. 2001; Harvey et al. 2003b; Shi et al. 2004, 2005), the new frontier in customer management (Yip and Madsen 1996).
Defined as ‘an organizational form and process in multinational companies by which the worldwide activities serving one or more multinational customers are coordinated centrally by one person or team within the supplier company’ (Montgomery and Yip 2000, p. 24), GCM represents a key organizational design issue for suppliers (Homburg et al. 2002). Meeting global customer demands requires a coordinated cross-functional effort, including establishing a customer dimension that crosses existing product, country or functional units and mobilizes organization-wide resources to deliver on customer expectations (Galbraith 2001). In its simplest form, this dimension might be a specialized global customer manager who plays a pivotal boundary-spanning role to manage external and internal relationships and coordinate dispersed value-adding activities (Millman 1996). As the relationship with the customer develops, the number of contacts between companies and the need for dedicated resources and relationship-specific adaptations increases (Dwyer, Schurr and Oh 1987), as does the number of people involved in the relationship. A dedicated team, possibly composed of sales representatives from various product lines and countries or with a cross-functional composition including manufacturing, distribution, finance, R&D and other functional units, works to present an integrated presence to the customer (Galbraith 2001). Ultimately, in an overall GCM organization, all customer managers and teams are integrated to reconcile external alignment requirements by the customer with the organization's existing configuration of activities.
Consequently, three main levels of analysis emerge in the context of GCM organizational solutions: the individual global customer manager, the global customer team (GCT) and the overall GCM programme. Whereas prior work addresses the role of global customer managers (Harvey et al. 2003b; Millman 1996; Wilson and Millman 2003) and structural aspects on the programme level (Birkinshaw et al. 2001; Homburg et al. 2002; Kempeners and van der Hart 1999; Shapiro and Moriarty 1984), no GCM research considers the design and functioning of the team in detail. As a result, we suffer a lack of understanding about when teams should be formed, how they should be structured and managed, and, most importantly, what determines their performance. Following the call of Workman et al. (2003) for more research that examines how team type influences effectiveness, we develop and test a framework of GCT design and performance that draws on relevant literature from the fields of GCM, team selling (Moon and Armstrong 1994; Moon and Gupta 1997; Smith and Barclay 1993) and small groups within the organization (e.g. Ancona and Caldwell 1992a; Cohen and Bailey 1997; Gladstein 1984).
Specifically, this paper aims to fill the research gap identified above by clearly delineating key GCT performance determinants. It identifies domains of GCM team functioning with their specific dimensions and tests their impact on team performance with a relatively large and truly global data sample. It contributes to the international marketing literature by narrowing the traditional focus on the overall GCM approach to concentrate on one of its less researched building blocks – the team. Moreover, it complements prior suggestions regarding the key decisions involved in designing a customer team (Kempeners and van der Hart 1999) by proposing additional design dimensions and linking them to outcomes. Finally, the novel interdisciplinary approach of combining concepts from customer management and organizational behaviour research allowed us to identify and employ team-related variables that, although used in the context of smaller teams, have never been empirically tested in the more complex, dynamic and boundary-transcending context of GCM.
Literature review
Teams in the organization
The extensive body of small group research provides valuable insights into how teams work. The broad use of teams in organizations has provoked classical team effectiveness models rooted in social psychology (McGrath 1964; Steiner 1972), socio-technical theory (Cummings 1978; Pasmore et al. 1982) and organizational psychology (Gladstein 1984; Hackman 1987), as well as myriad empirical studies in diverse organizational settings. Reviews of team research in general (Cohen and Bailey 1997; Gist et al. 1987) and team diversity in particular (Milliken and Martins 1996; Williams and O'Reilly 1998) demonstrate the bewildering range of identified constructs and relationships and provide some organizing frameworks. Although these findings can help explain some aspects of GCT interaction and performance, the special characteristics of GCTs prevent a direct transfer of the frameworks and results from small group contexts.
For example, the responsibilities of GCTs include complex interactions with groups external to both the team and the organization, whereas most small group research uses bounded or isolated groups as the unit of analysis. One exception is the stream of research that focuses on cross-functional teams, particularly in new product development contexts (Ancona and Caldwell 1992a, 1992b; Denison et al. 1996; Pelled et al. 1999), which recognizes the importance of external activities. However, its concentration remains on task-focused, limited-duration groups, which run counter to the long-term commitment expected from GCTs. Furthermore, general models of team effectiveness assume that teams exist to perform a well-defined task and have clear goals, whereas in a dynamic GCM context, tasks are likely to be unstructured, customer specific and evolving, which means they pose different challenges. This blurring of tasks and responsibilities becomes exacerbated further in GCTs by the involvement of people from different units and the complicated work matrix that may require GCM members to report to different managers (Birkinshaw et al. 2001). Therefore, small group literature cannot provide a complete answer to the question about what determines the performance of teams in more dynamic and high-velocity environments, such as GCM.
Team selling
Another body of literature tries to enhance understanding of teamwork in a more specific selling context. In response to the industrial buying centre approach (Johnston and Bonoma 1981), it follows the shift of focus away from the individual salesperson toward the collective selling effort (Weitz and Bradford 1999). Recognizing the increasing role of functions other than sales in servicing customers (Spekman and Johnston 1986), suppliers began to form selling centres to facilitate cross-functional interactions in the communication, coordination and exchange of resources and assistance (Ruekert and Walker 1987). These centres usually consist of a permanent core selling team and members of other functional groups that participate on an ad hoc basis (Moon and Gupta 1997).
Several studies examine the roles of selling centre members (Hutt et al. 1985; Moon and Armstrong 1994) and develop frameworks to incorporate the linkages among team composition, processes and effectiveness (Moon and Gupta 1997; Perry et al. 1999; Smith and Barclay 1993). These frameworks enhance understanding of how the sales division cooperates with other functions, as well as how various group and environmental factors influence team interactions and effectiveness. However, these studies contain two major limitations. First, existing models remain at a conceptual level, with no empirical support for their hypothesized relations. Second, they focus on a broader customer base that does not explicitly reflect the increased organizational complexity, cultural diversity and complexity of the solutions associated with global customers (Millman 1999). Therefore, team selling stops short of providing an explanation of the performance determinants in a GCT context.
The global customer team
Although teams facilitate the complex task of coordinating efforts by individuals across functional, product and geographic units to serve customer needs (Shapiro and Moriarty 1984) and form an integral part of many GCM programmes (Homburg et al. 2002), existing literature does not offer a precise definition of what constitutes a GCT. We might deduce that such a team comprises all persons involved in developing and maintaining relationships with one or several related key customers on a global basis and that its responsibilities include developing a customer strategy and account plan, creating innovative solutions and coordinating various networks (Galbraith 2001). Therefore, a GCT must operate in multi-actor structures with different tasks and levels of authority, and its success depends on its ability to secure continuous support and manage the various technical, legal, economic and political dimensions of the relationship (Harvey et al. 2003b). In these circumstances, the composition of the team becomes a critical managerial issue, because it determines the operating format and effectiveness of the GCM effort (Harvey et al. 2003a).
Despite the importance of this topic, research in the area remains sparse. Some studies touch on team design issues but stop short of identifying the determinants and performance effects of various team dimensions (Harvey et al. 2003a; Harvey et al. 2003b; Kempeners and van der Hart 1999). Shapiro and Moriarty (1984) note the importance of integrating two groups of support functions – sales support and other functional specialists – into the team and suggest several organizational solutions that range from using independent sales and functional forces to shared teams to fully dedicated teams. These ideas have been extended to formulate specific design decision topics (Kempeners and van der Hart 1999) and analyse the skills and knowledge diversity needed in GCT composition, depending on the challenges the team faces (Harvey et al. 2003b). Thus, several distinguishing characteristics of such groups have emerged, including geographic dispersion and multi-national composition (Harvey et al. 2003a; Montgomery and Yip 2000), cross-functional involvement (Galbraith 2001), the involvement of several hierarchical levels (Harvey et al. 2003a), full- and part-time membership (Kempeners and van der Hart 1999), parallels with competing organizational arrangements (Arnold et al. 2001) and continuous long-term tasks. However, the lack of an overarching theory or framework that consolidates these dimensions and explains how they lead to different outcomes remains a serious concern.
Furthermore, as a distinct type of group, GCTs prevent any direct extensions of frameworks and concepts from other research streams. Instead, they require a more profound investigation in the specific GCM context that includes both the integration and adaptation of existing team concepts and the identification of context-specific factors. Therefore, we develop a framework that conceptualizes some key design dimensions and performance outcomes.
Conceptual framework of GCT design and performance
Our framework proposes that the design of GCTs should incorporate six key dimensions, each of which influences team performance in terms of relational and financial outcomes through the mediating effect of three key processes: communication and collaboration, conflict management and proactiveness. In addition, three aspects of the organizational context of the GCTs – top-management support, rewards and incentives, and training – have similar effects on performance (see Figure 1).
Figure 1 Conceptual framework of GCT design and performance
GCT design
The GCT design construct builds on the Hackman (1987) group design concept and is similar to organizational behaviour constructs such as team structure (Gladstein 1984; Smith and Barclay 1993), team composition (Cohen and Bailey 1997; Stewart 2006) and team characteristics (e.g. Perry et al. 1999). However, it reflects the specificities of the GCT setting and the dimensions with highest relevance in this context. We define GCT design as the structural and competency characteristics of the team in terms of the distribution of skills, roles, tasks and power among the team members. Because of its practical complexity, this construct inevitably incorporates several related, but distinct, aspects. Methodologically, we tackle this issue in line with other relevant studies (Denison et al. 1996; Zou and Cavusgil 2002) modelling GCT design as a second-order variable that reflects six underlying first-order constructs: goal and role definition, customer coverage, empowerment, heterogeneity, adequate skills and leadership.
Goal and role definition
A key problem in forming groups is that team members might have conflicting group and individual goals that create mixed motives in terms of cooperating and sharing information (Bettenhausen 1991). Because GCT members often have multiple, organizationally defined roles and identities, conflicts arise, and operating managers' support for the GCT may be minimal if its role and responsibilities are not clearly delineated (Arnold et al. 2001). Aligning GCT goals with individual and overall corporate goals, as well as with customer goals, can therefore enhance collaboration, lower conflict levels and improve performance (Weldon and Weingart 1993). We therefore define this variable as the extent to which team members' goals and roles are clearly defined and aligned with team and corporate objectives.
Customer coverage
The GCT plays a boundary-spanning role, in which its ability to deal with the complexity marking the interface between the supplier and the customer depends on the extent to which it provides appropriate customer coverage across various dimensions, such as countries, divisions and functions. Structural fit between the supplier and the customer along these dimensions can improve supplier performance, because it enables the supplier to increase its information-processing capability and bargaining power in relation to its global customer (Birkinshaw et al. 2001). In addition, such fit can enhance the joint supplier–customer outcomes in terms of dyadic competitive advantages and joint profit performance (Shi et al. 2004). Therefore, customer coverage refers to the extent to which the team structure adequately corresponds to the customer's organization and its requirements.
Empowerment
Empowerment in a GCM context consists of two distinct aspects: decision-making authority and resource availability. Providing the team with sufficient authority tells the customer that customer managers have the power and resources to ‘get things done’ in a timely and effective manner (Homburg et al. 2002; McDonald et al. 1997). In a more general aspect, empowering teams by providing them with sufficient authority to make decisions for which they are responsible, can improve team processes (Mathieu et al. 2006), lead to greater proactiveness (Kirkman and Rosen 1999) and enhance positive customer-related outcomes such as customer service, customer satisfaction and process improvement (Kirkman et al. 2004). This empowerment requires the support derived from providing the team with sufficient resources to enable it to develop distinctive value propositions for its customers (Denison et al. 1996; Mohrman et al. 1995).
Heterogeneity
Heterogeneity refers to the diversity in skills and knowledge represented on the team. Following the resource-based view (Barney 1991; Teece et al. 1997), Harvey et al. (2003a) argue that individual and collective organizational knowledge as a resource enables the customer team to develop unique competencies valuable to global customers. Empirical evidence about the performance benefits of this type of diversity remains equivocal (Milliken and Martins 1996; Webber and Donahue 2001), but skill diversity clearly has positive performance effects on non-routine diverse tasks that require a wide range of competencies (Hambrick et al. 1996; Murray 1989) – a good description of GCT responsibilities. Therefore, incorporating a variety of backgrounds and areas of expertise within the GCT probably improves the team's performance.
Adequate skills
As opposed to skill heterogeneity, this construct focuses on individual characteristics that combine, such that more is always better for a team. The role of the global customer manager might be compared instructively to that of a political entrepreneur (Wilson and Millman 2003) who must overcome cultural barriers and navigate the sensitive political aspects of multiple interfaces while generating continuous business opportunities. These demands require a very broad range of skills that goes beyond that of a traditional salesperson, and teams whose members possess such a skill set are probably characterized by more collaborative and efficient team processes and thus superior performance.
Leadership
Because GCT leaders often have no formal authority, rely heavily on other members and are themselves contributing members to team activities, they must have strong influence both within and outside the team. Teams with leaders who can create effective working environments and motivate team members also feature more collaborative processes, reduced conflicts and higher effectiveness. Similarly, teams with leaders who exert a strong influence within the organization are more effective and collaborative because they communicate better with top management and other areas of the organization, more easily convince others that the team's activities are important, and secure more resources.
Hypothesis 1. GCT design has a positive influence on GCT processes.
Organizational context
Teams can be understood best in relation to their external surroundings and internal processes; therefore, most models of team effectiveness incorporate aspects from the organizational context of those teams (e.g. Campion et al. 1993; Denison et al. 1996; Shea and Guzzo 1987). Although there may be a large number of factors from outside the organization that play a role, it is usually the immediate surroundings within the organization that has the strongest impact (Stewart 2006). We identified three such factors that are of highest relevance in a GCT context: top-management support, rewards and incentives, and training. In line with Denison et al. (1996), we model organizational context as a second-order variable and these three factors as underlying first-order variables.
Top-management support
Top-management support often is identified as crucial for all aspects of GCM activities (Arnold et al. 2001; Homburg et al. 2002; Millman and Wilson 1999b), and some support emerges for the positive relationship between top-management support and GCM effectiveness (Workman et al. 2003), as well as marketing performance (Townsend et al. 2004) and market orientation (Jaworski and Kohli 1993). In particular, support in the form of resource allocations and public recognition of the GCT's efforts conveys the importance that top management attributes to the GCM initiative and is pivotal for overcoming potential resistance or power struggles within the firm, and legitimizing the efforts of the team as a strategic undertaking (Harvey et al. 2003b). This support enhances internal and external collaboration, reduces conflict and enforces a more proactive approach to global customers.
Rewards and incentives
A reward system that recognizes and reinforces contributions to the GCT can significantly amplify the motivational incentives provided by top-management commitment or well-designed goals and responsibilities. Because GCTs often work alongside pre-existing national sales organizations, and both units have vital roles in managing customers, it is important that the incentive structure helps reconcile any potential tensions between the units and further enhance collaboration (Arnold et al. 2001). A compensation system that enhances team processes and stimulates proactive contributions to the team usually requires two aspects: a measurement system that tracks the sales of global customers and the implied results of the GCT efforts worldwide, and an incentive system that rewards all team members appropriately.
Training
Due to the high demands on the skill sets of everyone involved with global customers, GCT members require professional development and training throughout their careers (Weeks and Stevens 1997). Even if these skills are available, team members, who come from different backgrounds and parts of the organization, need training to help them understand their tasks, processes, objectives and roles (Parker 1994). Team members who have successfully undergone rigorous training are more likely to contribute cooperatively and effectively to the team. Furthermore, external parties such as customers and other units in the supplier company probably perceive them as more competent and respectable and therefore become more willing to collaborate with them.
Hypothesis 2. The organizational context has a positive influence on GCT processes.
GCT processes
Although the team's structure and composition may have direct impacts on team outcomes, in most cases, this relationship also is mediated by the processes that characterize the interactions of team members (Lawrence 1997). We follow the definitions of Cohen and Bailey (1997, p. 244) who describe the team process as ‘interactions such as communication and conflict that occur among group members and external others' and of Marks et al. (2001, p. 357) who expand the definition to ‘members’ interdependent acts that convert inputs to outcomes through cognitive, verbal, and behavioural activities directed toward organizing taskwork to achieve collective goals'. Consequently, our framework suggests three key mediating processes that have an impact in a GCT context: communication and collaboration, conflict management and proactiveness.
Communication and collaboration
Communication and collaboration refer to interactions both within the team and with external parties. The link between effective internal collaboration and team performance receives broad support from organizational behaviour studies (Bunderson and Sutcliffe 2002; Pinto et al. 1993), supplier–buyer relationship literature (Buvik and John 2000; Heide and John 1990; Mohr and Spekman 1994) and GCM studies (Birkinshaw et al. 2001; Shi et al. 2005). Similarly, the active management of external linkages and engagement of outsiders is crucial from a resource-dependence perspective (Pfeffer 1986) and is a key determinant of team performance (Ancona and Caldwell 1992a,b) and overall GCM performance (Homburg et al. 2002).
Conflict management
In GCTs, conflicts can arise as a result of diverse aspects of the home organizations (functions, units and regions) represented on the team, which lead to contrasting views and divergent task interpretations. Furthermore, conflicts may occur because of resource disparities or disagreements about resource allocations (Denison et al. 1996; Donnellon 1993). Therefore, we argue that the extent and frequency of conflicts, as well as the existence of working mechanisms that help resolve disputes in a timely and effective manner, influence GCT performance.
Proactiveness
Although many GCM activities result from customer demands, the potential for a cooperative and synergistic relationship is greatest when the supplier adopts more proactive behaviours (Harvey et al. 2003a; Homburg et al. 2002). Consequently, proactive team behaviours that actively seek areas for continuous improvement, identify opportunities and innovative solutions to problems and address issues before they become major problems represent a critical success factor (Bateman and Crant 1993; Hyatt and Ruddy 1997).
Hypothesis 3. GCT processes have a positive influence on GCT relational performance.
Hypothesis 4. GCT processes have a positive influence on GCT financial performance.
GCT performance
Global customer team performance often gets represented as two distinct components: financial, or quantitative, and relational, or qualitative (Birkinshaw et al. 2001).
Financial performance relates to key quantitative performance measures, such as sales growth and profitability (Birkinshaw et al. 2001; Shi et al. 2005), as well as measures such as reduced cost of sales to customers, more efficient use of salespeople's time in serving customers, cross-selling to divisions in which the supplier traditionally has been weak (Arnold et al. 2001), securing desired market share, attracting new customers (Homburg et al. 2002), or growing the share of the customer's wallet (i.e. share of the customer's total purchasing volume devoted to the supplier). Measuring GCT performance using purely quantitative criteria, however, runs the risk of missing vital aspects of customer partnerships, as well as the broader goals and objectives of developing relationships with key customers. Therefore, some qualitative performance indicators should be taken into account as well.
Relational performance refers to achieving goals such as long-term customer relationships, learning and innovation through various joint initiatives, and access to the customer's new product ideas (Birkinshaw et al. 2001). Other qualitative measures may include customer assessments of the relationship and customer satisfaction, which can increase customer value and mutual trust (Gladstein 1984; Homburg et al. 2002). In addition, the team's ability to meet its goals and objectives (Moon and Gupta 1997), gain broader market access and achieve a better competitive position in the market can indicate its performance (Smith and Barclay 1993). We expect that improvements in these indicators ultimately also enhance the team's financial performance.
Hypothesis 5. GCT relational performance has a positive influence on GCT financial performance.
Research methodology
We conduct this study in two separate stages, using both qualitative and quantitative methods. In the first stage, we conduct 15 interviews with managers in various industries, all of whom have direct responsibility for managing GCM programmes or teams. Each interview focused on the critical success factors for the customer teams in which the respondent had been involved, with the goal of better understanding the performance drivers of GCM teamwork. Through a content analysis of these interviews, we develop the questionnaire for the study. To ensure the reliability and validity of the measurements, our questionnaire development consists of several stages, following established guidelines (Churchill 1979; DeVellis 1991; Venkatraman and Grant 1986). In the first step, we review all relevant literature, with a specific focus on scales used in previous empirical research. Our analysis of the expert interviews and the thorough literature review helped us identify existing scales, adapt them where appropriate and develop new measures if no available scales existed. We pre-tested the questionnaire with several academic experts and many senior GCM executives, including managers from each of the six research companies and some of the interviewees. In accordance with their comments, we added, deleted or reworded several questions without causing a significant loss of validity in the constructs. The final versions of the scales in each of the four domains appear in Tables 2,3,4, and respectively. The coefficients alpha for all measurements is satisfactory and demonstrate their reliability.
In the second stage, we administered the questionnaire to 113 teams in six multi-national companies. We followed a rigorous selection process, whereby we focused on the top companies from the Forbes Global 2000 list and profiled them on the basis of our understanding of their GCM practices. Following several rounds of thorough assessment and contacts with potential participant companies, we narrowed the list to the six focal companies for two main reasons. First, we aimed for an in-depth study, with the team (rather than the company) as the main unit of analysis, which required the involvement of a large number of teams within relatively similar environments. Selecting six companies helped us achieve this objective and reduced the variance in the external factors. Second, these six companies are highly representative of the population of international companies with established GCTs; all have conducted structured GCM programmes for at least six years and use GCTs extensively to manage their global customer relationships. They earn average revenues of US$18.7 billion (in 2008), employ 105,000 people on average and represent different industries, including construction equipment, engineering, hospitality services, electronics and office equipment.
Within these six companies, we targeted 425 team members in total and received 273 complete questionnaires, for an effective response rate of 64.2%, which compares favourably with similar studies (e.g. Denison et al. 1996). All respondents are qualified to complete the survey: they directly support or are involved in the GCT, in positions ranging from key or global account managers to regional or country general managers to members of the top-management team. More detailed sample characteristics appear in Table 1. The respondents' average tenure with the company is 12.5 years (SD = 8.6) and only 19% had been with their company for three years or less.
Table 1: Sample description
| Positions of respondents | Total (n = 273) |
|---|---|
| Managing director, general manager, head of region/business unit | 24% |
| Sales director, sales manager, head of sales | 17% |
| Global account manager, international key account manager, head of GCM | 24% |
| Key account manager, national account manager, account manager | 20% |
| Other | 15% |
| Represented countries | Total (n = 273) |
| UK | 28% |
| USA | 23% |
| Germany | 8% |
| France | 7% |
| Rest of Europe | 29% |
| Latin America | 2% |
| Asia | 3% |
To test for potential non-response bias, we use an extrapolation procedure (Armstrong and Overton 1977) in which we divide the data set into thirds according to the date we received the completed questionnaire, then test for differences between the first and last third. The t-tests of the mean responses of early and late respondents indicate no significant differences (p < .05), suggesting non-response bias is not a concern.
Results
We analyse the data using a three-step process. In the first step, we assess the validity of the variables suggested by the literature review and our qualitative analysis in each of the four dimensions (design, organizational context, processes and performance) on the basis of an exploratory factor analysis. In the second step, we perform separate confirmatory factor analyses (CFA) for each dimension to validate the relationships between the dimensions and the individual constructs. In the third step, we test the full structural model.
In Table 2, we provide the results of the exploratory factor analysis of the item pool for the GCT design dimension. To extract the factors, we use a principal component method, based on correlation matrices and Varimax rotation. The Kaiser criterion (i.e. factor loadings greater than or equal to .50) dictates which items we take into consideration; we highlight factor loadings above this cut-off point in bold. The six factors with eigenvalues greater than 1.0, according to our analysis, correspond directly to the proposed design constructs, namely, goal and role definition, customer coverage, empowerment, heterogeneity, adequate skills and leadership. Table 2 reports also the Cronbach (1951) coefficient alpha for the six variables. All of them are above the recommended level of .7 (Nunnally and Bernstein 1994), indicating that the measurement scales have a sufficient degree of reliability.
We present the results of the exploratory factor analysis for the organizational context dimension in Table 3. They reveal three factors, as posited by our literature review and qualitative research: top-management support, rewards and incentives, and training. All of them can be considered reliable as indicated by the high values of the coefficient alpha.
Table 3: Results of exploratory factor analysis of organizational context domain
Next, in Table 4, we present the results of our exploratory factor analysis of the team processes dimension. These results support the three suggested factors – communication and collaboration, conflict management and proactiveness – and show a good level of reliability (with the exception of conflict management whose coefficient alpha is slightly below .7 but still satisfactory).
Table 4: Results of exploratory factor analysis of team processes domain
Finally, Table 5 displays the results of our exploratory factor analysis of the team performance dimension, which supports our proposed distinction between relational and financial performance.
Table 5: Results of exploratory factor analysis of team performance domain
| Constructs and items | 1 | 2 |
|---|---|---|
| Relational performance (α = .859) | ||
| 1. Our team is characterized by strong and harmonious long-term relationships with global customers. | .745 | .305 |
| 2. Our customers are satisfied with the overall performance of our team. | .767 | .291 |
| 3. Our team provides real value to our customers. | .800 | .277 |
| 4. Our team has successfully learned some critical skills or capabilities from our customer relationships. | .728 | .211 |
| 5. Our company's competitive position has been enhanced due to our team's GCM achievements. | .736 | .214 |
| Financial performance (α = .826) | ||
| How has your team, over the past three years, performed with respect to | ||
| 1. … growth in sales? | .251 | .881 |
| 2. … profitability? | .296 | .779 |
| 3. … growth in the share of your global customers' wallets? | .282 | .789 |
| Eigenvalues | 40.416 | 1.020 |
| Variance explained by factor after Varimax rotation | 38.6% | 29.4% |
| Total variance explained | 67.9% |
In Table 6, we include the correlation matrix of the indexes from the exploratory analysis, as well as their means and standard deviations.
Table 6: Means, standard deviations, and correlations among constructs
Following the initial exploratory factor analysis (EFA) support for the validity of the proposed constructs, we perform CFA to specify the relations of the underlying constructs to their second-order factors. This approach is in line with the recommendations for a two-step approach in which the measurement model is estimated separately prior to the simultaneous estimation of the measurement and structural submodels (Anderson and Gerbing 1988; Fornell and Yi 1992) and is consistent with other studies in the marketing area (e.g. Townsend et al. 2004; Zou and Cavusgil 2002)
Table 7 reports the results for the second-order measurement model. The chi-square/degrees of freedom ratio (χ2/df), root mean square error of approximation (RMSEA), root mean square residual (RMR), comparative fit index (CFI) and goodness-of-fit index (GFI) values are satisfactory for all four dimensions and thus suggest the CFA models fit the data adequately (Hu and Bentler 1999).
Table 7: Results of the second-order CFA
Next, we assess the internal structure of the models and the convergent validity of the constructs. The loadings of all items on their respective factors are positive, high in magnitude and statistically significant, which indicates satisfactory convergent validity (Anderson, 1987). To assess the discriminant validity of the factors, we use the procedure recommended by Bagozzi et al. (1991) and therefore constrain the correlations between each pair of first-order factors to equal 1 and then release this restriction in a series of consecutive CFAs. In all cases, the chi-square statistics of the nested model are significantly higher than those of the unconstrained models, which indicates all constructs are distinct and that the model achieves discriminant validity.
Following the purification of the measurement model through CFA, we test the complete structural path model using the maximum likelihood procedure. The full model includes the structural model as well as the multiple indicators of all first- and second-order constructs obtained in the previous stages. We provide the results in Figure 2, including the structural relationships, standardized estimates of the path coefficients and fit indexes. The t-values for all estimates appear in brackets.
Figure 2 Structural path model
Furthermore, we thoroughly examine the solution by inspecting the parameter estimates, standard errors and modification indexes, as well as any possible irregularities. We detect no problems. The χ2 for this model is 1,621.29, which with 1,104 degrees of freedom results in a χ2/df ratio of 1.469, indicating a good fit. The other fit indexes are as follows: RMSEA = .042, RMR = .041, CFI = .909 and NFI = .765. Given the relatively complex nature of the model, with its many indicators and three second-order factors, these indexes suggest a satisfactory fit (Bollen 1989).
The estimates of the path coefficients in model 1 provide support for H1, H3 and H5. Although we do not find support for H2, we identify an alternative path of influence through which organizational context may affect the performance of GCTs. Organizational context has a positive and significant effect on team design and, from there, indirectly influences team processes, which supports the importance of the context construct. In addition, H4 is not supported, so team processes may influence financial performance only indirectly through their impact on relational performance.
Discussion
Global customer teams represent an important part of many GCM programmes, but existing research provides few directions for clarifying their formation, structure, composition, interaction processes or performance. Furthermore, GCTs differ from other forms of groups because of their complexity in terms of their cross-regional, cross-functional and cross-divisional span, blurred responsibilities and lines of authority, heightened external activity requirements and long-term tasks. This character of GCTs further complicates the managerial tasks of building and managing high-performance customer teams. Consequently, we examine what determines GCT performance by developing and empirically testing an integrative model of team composition, interaction and performance. On the basis of a comprehensive literature review, exploratory interview research and an analysis of survey data from six multi-national companies, we identify 12 determinants of team performance in three broad domains. We evaluate these determinants against both qualitative and quantitative performance criteria.
GCT design
The empirical results indicate that GCT design is a complex, multi-faceted construct that encompasses six key dimensions: role and goal clarity, customer coverage, empowerment, heterogeneity, adequate skills and leadership. These structural and compositional dimensions influence team performance through the mediating role of three key processes: communication and collaboration, conflict management and proactiveness. The resulting implication is that building high-performance GCTs starts with a clear specification of objectives in relation to the customer and the overall corporate goals, as well as a delineation of the roles and responsibilities that team members should fulfil. Furthermore, the team composition should receive sufficient attention to identify people with adequate skills who can develop relationships and navigate the interface between global organizations. Team performance could be enhanced further if these arrangements were accompanied by a leader with a strong presence and influence in the organization, as well as by the provision of sufficient resources and authority to the team. These factors would enable the team to make fast and efficient decisions and take courses of action that are in the best interests of the customer, even if they are unpopular or face opposition within the supplier organization. As a result, the team will experience less conflict and higher levels of collaboration and proactive behaviour. These results extend previous research in GCM by identifying a novel dimension, namely, customer coverage, as well as testing the skills-performance relationship empirically.
Organizational context
Three key elements of the organizational context of GCTs – top-management support, rewards and incentives, and training – emerge from our analysis, with the implication that developing high-performance GCTs probably will not succeed without genuine support from senior managers and proper incentives that encourage team contributions and stimulate global thinking throughout the organization. Furthermore, building strong teams requires that team members receive well-rounded training opportunities to develop their relevant selling, interpersonal and technical skills. Although a direct link to team processes is not supported (H2), we find an indirect link through team design. These constructs may contain several aspects that remain implicit in the study. For example, we predicted top-management support would influence team processes by helping the team in its work with the customer and creating a more cooperative and customer-oriented organizational environment. However, other aspects of top management support, such as involvement in the goal- and role-setting process, appointment of a team leader and members, or team empowerment, may explain these results. Therefore, this finding should prompt further investigation into the specific elements of the context dimension.
Team processes and performance
The results indicate that all design, process and context dimensions have either direct or indirect positive influences on performance. The finding that the link between design and performance is mediated by team processes falls in line with studies that call for a means to open the ‘black box’ of team interactions (Lawrence 1997) and implies that decisions about a team's structure and composition should be based on a careful assessment of their potential consequences for the way team members interact within the team and with their broader environment.
In addition, we find that GCT performance has two associated components, relational and financial performance, and that relational performance is only an intermediate outcome that ultimately leads to improved financial results in terms of sales growth, profitability and greater share of wallet. The lack of support for H4 suggests some interesting implications. We formulate it on the basis of prior studies that indicate a direct influence of GCM processes, such as communication and coordination, on financial measures of profitability and sales growth (Birkinshaw et al. 2001; Shi et al. 2005). Therefore, our findings suggest that prior research might have overlooked an important mediator of relational performance; we contribute to the literature by providing this missing link.
This finding also has implications for organizational behaviour research. Most studies in this research stream distinguish between performance outcomes, such as goal achievement, productivity or other relevant criteria, and attitudinal outcomes in terms of team satisfaction, commitment and trust. However, these two performance groups usually represent ultimate outcomes of effective teamwork, with no explicit interrelations. Our findings imply that the relationship between the two may require closer consideration.
Conclusions and implications
The major contribution offered by this paper is the development and empirical validation of an integrative model that sheds light on the performance determinants of a relatively new and under-researched phenomenon, namely, GCT. Its novel approach lies in the interdisciplinary combination of concepts from customer management and organizational behaviour research.
Our focus on the team level has not been investigated sufficiently in previous customer management literature. Therefore, our detailed delineation of the domains of team functions, along with their specific elements, provides a more complete set of performance determinants and extends prior research on GCT, which addresses only fundamental organizational decisions (Kempeners and van der Hart 1999). The introduction of concepts from research on organizational groups and cross-functional teams also broadens understanding of GCM teamwork by moving a step beyond purely structural issues. Moreover, we suggest and inspect variables that have never been empirically tested in a GCM context, including empowerment, leadership, conflict management and training.
Our contribution to the organizational behaviour field involves identifying factors that gain greater importance in teams with more complex and dynamic structures, tasks and environments. For example, constructs such as proactiveness, top-management support, skills and rewards have not been broadly emphasized or have produced inconclusive results in previous studies of teams with more structured and stable settings. However, they receive strong confirmation in this study, which implies a distinctive set of individual, team and organizational capabilities that may generalize to teams with similarly dynamic natures.
From a managerial viewpoint, this study highlights the importance of adopting a holistic approach to the composition and management of GCTs. As the results show, all team characteristics and the context in which the team works have important performance implications. Consequently, neglecting some or all of these issues will represent a key barrier to team effectiveness and the overall GCM operations of the supplier. Moreover, we develop a comprehensive model that can guide companies attempting to implement customer teams and help them improve their existing global customer relationships. Many of the identified elements fall directly or indirectly under the control of senior managers. Therefore, the complete set of 12 team characteristics and their corresponding questionnaire items, along with the 9 performance indicators, may serve as a practical tool for designing new GCTs or enhancing the effectiveness of existing ones, because they enable the firm to assess its current performance, map out areas for improvement, develop courses of action and track progress along each dimension.
Finally, the tested linkages among the variables illustrate the mechanisms through which different aspects of GCM teamwork are interrelated. This can help managers better understand and foresee potential consequences of their actions in each domain.
Limitations and suggestions for further research
Due to the exploratory nature of our study, we acknowledge that further research is needed to validate our proposed framework and could envisage several avenues for future investigation. First, based on our research design, most of the findings are a posteriori. Further research, however, may aim to establish additional research questions and hypotheses based on previous research and the current findings. Second, future contributions might include an analysis of the individual performance impact of each factor. The result could be a ranking or weighting of the factors depending on the magnitude of their effects, which would have important managerial implications. Third, researchers might build upon our findings to deepen the understanding of team processes which play such an important role in our framework. The conceptualization and empirical examination of complex concepts such as communication and collaboration could benefit greatly from further endeavours in this area.
In addition, some of the methodological limitations of our study could be addressed in the future. First, we use subjective performance ratings. Further research might test the effect of the identified independent variables on objective measures, such as actual sales growth, profitability, share of wallet or customer satisfaction ratings obtained in customer surveys. For measures that are rather subjective in nature (e.g. relationship building, customer value, learning, competitive position), a more objective rating might be obtained through managerial or customer assessments of these indicators. Second, our study is based entirely on data from supplier companies and lacks a customer perspective. Although some dimensions may be less visible or more difficult to evaluate from the customer side, a worthwhile extension of the findings would be to test the model with a sample of members from the customers' buying teams which work with the GCTs.
Also, due to the exploratory nature of our study, we consider the sample of six companies sufficient for the sake of depth and clarity. However, additional studies should seek to test the framework in other settings or with a larger sample of companies and thereby study each team characteristic more deeply by identifying their individual antecedents and consequences or by building on the model to identify other dimensions and measure their performance outcomes. One such research direction could be a study comparing global companies across countries to understand to what extent team design is determined by the practices in different countries. Finally, another interesting avenue for further research involves conducting a multi-level analysis to clarify the individual-, group- and organizational-level variables that may influence team performance.
Note
© 2010 Elsevier Inc. All rights reserved. Reproduced with the kind permission of Elsevier, originally Atanasova, Y. & Senn, C. (2011) Global customer team design: Dimensions, determinants, and performance outcomes. Industrial Marketing Management. 40(2). p. 278–289.