OOOO, Investigation of Factors Contributing to the Success of Cross- Functional Teams Edward F. McDonough II Although recent empirical research shows that most firms have implemented cross-functional teams for the majority of the new product development projects undertaken, they are still finding it hard to ensure that these teams are successful in completing the new product development task. In this article, the author first reviews the vast literature on cross-functional new product development teams 10 uncover the array of factors that have previously been demonstrated or hypoth- esized t0 relate to cross-functional feam success, when measured at the project level. He then analyzes the responses of 112 new product development profes- sionals to determine which factors are more frequently mentioned as leading to project success In looking at how to achieve successful teams, many factors have been sug- gested in the literature by a number of different researchers. The author suggests @ model of these factors that divides them into three categories that help achieve success. Setting the stage for product development by developing appropriate project goals, empowering the team with the needed decision-making power, assigning the appropriate human resources, and creating a productive climate should be related 10 fostering team success. Of these four factors, appropriate project goals is mentioned most often as being associated with success, followed by empowerment. Several specific team behaviors, including cooperation, com- ‘mitment to the project, ownership of the project, and respect and trust among team members, also have been posited to contribute to team success. Of these, this research finds that cooperation is mentioned most often as being associated with success, followed by commitment and ownership. Finally, a number of research- ers have suggested that team leaders, senior managers, and champions provide enabling support to cross-functional teams in achieving success. Team leadership is the most frequently mentioned enabler, according to these findings, followed by senior management support. The author's results also show that increased use of cross-functional teams in new product development is related to higher project success. However, achieving cross-functional team success appears to be more complicated than previously thought. For example, across the set of factors identified in this research, the most frequently mentioned is obtaining the team behavior of cooperation. Setting Adress correspondence to Edward F. MeDonoagh Ill, College of Business Administration, Northeastern University, 04 Hayden al, Bow: ton, MA 02215-5700, J PROD INNOV MANAG 2000:17-221-255 © 2000 Elsevier Science In Al rights reserved (0737-6782/00-se fron mater (655 Avenue ofthe Americas, New York, NY 10010 Pus1 F MCDONOUGH I Appropriate project goals, a stage-setting step that is completed early in the Project, follows closely in relative importance. Finally, providing good team leadership as an enabler is the third most frequently mentioned factor in achiev- ing success. This suggests that companies must work in all dimensions to maxi- mize the probability of achieving team success. © 2000 Elsevier Science Inc Introduction ‘The rapid change and diffusion of technology as well as burgeoning global competition have intensified the need for complex and highly novel product innova- tions, with the result that the use of cross-functional product development teams! has become increasingly important [27]. Donnellon [14], p. 377] notes, “Whether teams represent a thoughtful strategic ad- justment to environmental contingencies or a reactive emulation of successful competitors, they are now widely recognized for the potential that theorists have asserted for many years.” This may explain why, as Adler [1] points out, more organizations are now seek- ing to replace the traditional sequential model of new product design with a reciprocal process, one element of which is a cross-functional team. Indeed, recent research indicates that companies now rely on cross- functional teams between 70% and 75% of the time [12,21], Part of the attraction of cross-functional teams is the perception that their use will lead to greater new product success. But, as Hitt et al, [27], p. 3] noted, “While much research has been conducted on the function and process of groups, less is known about the specific characteristics and success of eross-func- mal teams [author's emphasis] in the design and development of new products.” Thus, despite this recognition of their potential, for many firms cross-funetional teams are still “terra in- cognita.”? This may stem from the different orienta- tions, goals, departmental cultures, and language that each functional representative brings with them to the team [2,34,36,54,55]. As well, what makes one effec- tive and another a failure may not be well understood 27,32], as suggested by the conflicting results of research that has examined the relationship between performance and the use of cross-functional teams [1,7,9,11,12,15,17,20,23-25,46,48,52,55,56,59,60) Some studies found that the involvement of multiple "Ror ease of discussion, fom heron T refer to cross fetinal prot evelopment teams ss simply eros-funetional teas xiely [32}- nis rile, "Innovation Consregations,” refers to a stay by Susan Cohen, who looked at 17D teams in 11 companies, One of her key Findings was that ours are stil “era incognita.” functions in new product development (NPD) has a positive effect on performance [15,43,55], whereas, others found a less clear-cut relationship between the use of cross-functional teams and performance [2,9]. These conflicting results may stem from the orga- nizational context within which cross-functional teams, ‘operate, including the size of the firm and the industry, ‘or they may stem from the internal infrastructure of the firm using cross-functional teams. Cross-functional teams may simply work in particular circumstances and contexts but not in others. Alternatively, they may work well in a variety of contexts, but the firm's internal infrastructure may cause them to fail. Even those companies that acknowledged the importance of cross-functional teams found that implementing effec- tive organization systems, structures, practices, and procedures is an extremely difficult task and one that they have not always performed successfully [25] Hence, although cross-functional teams have been ballyhooed as a key organizational tool for developing new products, it is not clear why and under what circumstances they should be used. These contradic tory research findings and gaps in our research suggest that our understanding of them is still very much incomplete. The purpose of this article is twofold. On the one hand, it seeks to enhance our understanding of cross-functional team use and suecess by investigating the following questions: What is the impetus for companies to begin to use cross-functional teams? 2. What factors are perceived as being associated with cross-functional team success? On the other hand, it seeks to provide direction for future research through the development of a model that proposes relationships among factors associated with cross-functional team success and the impact of contextual variables on these factors, Literature Review One stream of research on cross-functional teams has examined the relationship between the use of cross-SUCCESS OF CROSS-FUNCTIONAL TEAMS functional teams and various measures of performance [2,21,43,61]. This research has yielded inconsistent results. Ancona and Caldwell [2], p. 338], for exam- ple, found that the functional diversity of cross-func~ tional teams was negatively related to performance. They concluded that, “simply changing the structure of teams (i.e., combining representatives of diverse function and tenure) will not improve performance.” Other researchers found that the functional diversity of teams may help to speed up product development [61], particularly in the early stages of product development {22}, but may not contribute to innovativeness and team performance. Still other research suggests that innovation strategy may moderate the relationship between cross-func- tional team use and performance at the firm level (43). MeDonough and Griffin [43] examined the moderat- ing effects of three innovation strategies—prospect ing, analyzing, and defending [45}, each of which varies along two dimensions: technological innova- tiveness and speed of development. For prospectors, speed of developing a product is less critical than is devetoping a highly innovative product. Analyzers, on the other hand, emphasize speed of response to retain competitiveness. Defenders attempt to locate and maintain a secure niche in a relatively stable product or service area. They are rarely highly innovative, nor is their response especially rapid. It was found that in firms that employed either a prospector or an analyzer innovation strategy, the use of cross-functional teams was associated with a firm meeting its NPD program objectives [43]. Cross-funetional team use also. was associated with industry success for firms employing a prospector innovation strategy. Industry success was defined as an organization’s overall new product suc- ess compared to their primary competitors [43]. The use of cross-functional teams was not significantly related to performance for firms employing a defender innovation strategy. These results suggest the need to take into account moderating factors when examining the relationship between NPD performance and the use of cross-functional teams A second stream of research has investigated the relationship between success and the characteristics of ‘cross-functional teams, as opposed to simply their use [6.26,27,40,42,50,57]. Although several relatively dis- tinct areas of focus can be identified within this liter- ature, there have been few attempts to organize it in a way that helps us to understand the interrelationships among those factors found to affect success {13]. Be- low, the research that has been conducted within three major areas of focus, stage-setting elements, enablers, and team behaviors, is reviewed, Stage-Setting Elements One group of studies has focused on variables that play an antecedent role in the product develop- ment process, including setting project goals [6,33,49,50,57], empowering project team members [6,14,18,26,33,39,41,50,57,58}, establishing a project, climate [42,57], and the human (as opposed to physi cal or financial) resources of the team [6,10,38,60] ‘These antecedent elements reflect management actions that initially direct the development effort and set the stage for the product development that follows. They are put in place at the outset of a project and create the foundation on which the work of the project takes place. Goals. Establishing goals for a project provides several benefits. On the one hand, they provide project members with a common frame of reference, which, in turn, promotes a higher lever of cross-functional ct operation [50]. Superordinate goals also help structure tasks and, in doing so, facilitate cooperation by k ing team members oriented toward a common task outcome [50]. Establishing goals can impact project performance, but only indirectly by affecting cross- functional cooperation [50] Providing teams with clear, consistent goals is also a way to create boundaries for the project team so that, it is not continually redefining its direction [6]. Goals have the effect of not only telling teams what to do, but also what not to do. In this way, they focus the team’s efforts. Thus, providing clear goals for cross- functional teams accomplishes two objectives. On the one hand, they provide a common focal point for the team’s efforts, while on the other hand, they constrain, the team’s efforts within boundaries. Empowerment, Goals also are important to set boundaries within which teams can be empowered to make project related decisions [40]. Studies that looked at the relationship between empowerment in cross-functional teams and performance yielded sur- prisingly consistent results (6,14,26,41,57]. Giving i dividuals greater decision-making responsibility leads to their being more committed to the project and to meeting its goals, At the same time, it can lead to greater satisfaction on the part of team members. Me- Donough and Barezak [41], for example, found that the speed with which new products are developed was significantly related to the amount of freedom andna [PROD INNOY MANAG aunt 98s responsibility given to project team members by the project leader. In an earlier study, it also was found that driving decision-making down to the lowest pos- sible levels enabled firms to respond more quickly {39}. By locating decision-making authority at the project team level, firms were able to reduce the time it takes to make decisions, solve problems, and take actions, and thus reduce product development cycle time, But, giving the team decision-making authority may not directly impact on project performance. Instead, by allowing teams to establish their own rules and pro- cedures (two aspects of empowerment (33,58}), cross- functional cooperation is fostered, which, in tum, ef- fects project outcomes [50]. Thus, empowerment may affect performance indirectly by generating coopera- tion among team members, which, in turn, may lead to faster decision-making and higher quality products. Empowering a team has multiple impacts. It not only can result in faster product development and greater cooperation, but it is also one means of gen- erating a performance enhancing climate. When sub- ordinates are empowered, they perceive their work ‘groups, their managers, and themselves as more influ- ential and, as a result, a more innovative climate is created [18]. Climate. The climate that surrounds a. product development effort can play an important role in set- ting the stage for the NPD effort. At the outset of a project or even before it officially begins, management consciously or unconsciously creates a climate. AS noted earlier, one means of creating a more innovative climate is by empowering team members. Another ‘means is to create a sense of urgency about the project [42]. In order to make clear the importance of a project, senior management needs to develop what ‘Thamhain [57] refers to as a “priority image” for the project. Creating a sense of urgency about the project and conveying to tea members its importance stim- ulates a feeling of excitement about working on the project and generates a feeling of commitment to it. Feelings of commitment, as well as trust, greater com- munication, and better cooperation, can be enhanced by the selection of team members, i.c., a project’s human resources. Human resources. The capabilities that are em- bodied in the members of the team represent a signif icant project resource. The extent to which there is a good fit among these resources and the project can impact on its effectiveness [5,25]. Brown and Eisen- hardt [5], p. 367] suggest that the functional diversity 5. MCpONOUGIE IE of cross-functional teams increases the amount and variety of information available to design products, which, in tur, “helps project team members to under stand the design process more quickly and fully from 4 variety of perspectives, and thus it improves design process performance. Moreover, the increased infor- mation helps the team to catch downstream problems such as manufacturing difficulties or market mis- matches before they happen, when these problems are generally smaller and easier to fix.” Thus, the variety of perspectives that can be brought t0 bear on prob- Jems and the development process is increased by including individuals from different functional disci- plines on project teams. ‘Management can set the stage for product develop- ment by putting together a team of individuals who possess the “right” skills and by ensuring that the needed technical expertise and interpersonal skills are present on the team [6,10,25,60]. But the “right skills” can depend on the type of development project under- taken. For example, McDonough [38] found that using ‘more highly educated team members on routine NPD projects resulted in faster development, whereas using, individuals who were newer to the company on more, radical projects resulted in faster product develop- ment, These four variables—goals, empowerment, cli- mate, and human resources—are elements that, in some sense, precede the development effort itself. They set the stage for subsequent development by creating clarity of direction for the project, providing team members with the decision-making authority they will be able to exercise during the project, creat- ing a climate enveloping the team’s efforts, and se- lecting individuals who will enhance the likelihood of project success. Stage-setting elements also work in conjunction to create a basis for undertaking the development work of, the team. As Donnellon [14], p. 388) notes, to “realize the benefits afforded teams as an integrative device capable of balancing and reconciling multiple sub- goals, organizations need to communicate the priority of strategic goals to teams and [Donnellon’s empha- sis] give teams the authority and autonomy to resolve the conflicts and manage their progress towards these goals over time.” For example, management has the potential to influence how quickly the team will progress toward the attainment of goals by establ ing a climate of urgency and by empowering team members to make decisions, Similarly, by assembling a team of individuals who possess the “tight” capabilities, management can enhance the possibility of suc~ cessful project performance, but only if they have established clear goals and boundaries [40] But simply setting the stage for product develop- ‘ment is not sufficient to ensure that development will progress effectively. It is during the process of devel- oping a new product that managers and leaders can play a key role by assisting the team in overcoming problems and providing support. These individuals act as enablers of the NPD process. Enablers A second area of focus for research has been on individuals—team leaders, managers, and champi- ons—who have the potential to enable cross-fun tional team success. These individuals can facilitate the development team’s efforts and impact on the stage-setting elements that have been put into place ally. Individuals who act as enablers can be lo- cated within different hierarchical levels of the orga- vation. Regardless of their location within the hier archy, though, each is in a position to potentially facilitate the team’s efforts toward successfully devel- oping and bringing to market a new product by playing an enabling role. Researchers found that these individ- uals play different roles, including leading the cross- functional team [4,16,40], providing support joning the project [35]. ‘Team leadership. There has been surprisingly little empirical research that has examined the rela- tionship between performance and the leadership of cross-functional teams [5,19]. The research that has been done indicates quite strongly that team leaders do not take direct action to precipitate project success. Rather, they operate indirectly as enablers of the NPD process. This enabling function can take several forms For example, McDonough and Leifer [40] found that effective project leaders delineated task boundaries for the team and then allowed team members to perform within those boundaries without specifying how the work itself was to take place. In this way, they enabled the development process, i.e., the work of the team, rather than engaging directly in carrying out develop- ‘ment tasks themselves. Another means of enabling is for team leaders to engage in a participatory style of leadership, where team members are given the freedom to explore, dis- cuss, and challenge ideas and make their own deci- sions about what technologies to pursue, problems to solve, and tasks to undertake. When this style is em- (14,27,57,60), and chamy ployed, the team leader gives considerable control to the team to conduct product development as they see fit (41). In this way, the team leader is an “enabler” by virtue of ceding responsibility to team members for making decisions, i.e., empowering them, rather than taking that responsibility for himself or herself. In a study of British firms, McDonough and Barczak [41] found that just such a participatory style of leadership was associated with higher project performance, Using an open and apolitical style of leadership also may facilitate successful NPD [44]. Apolitical leaders share information and knowledge broadly with the team and other groups within the organization so that, realistic decisions can be made based on facts and 80 that realistic expectations for the projects are known, by and accepted at all levels of the organization Leaders also enable NPD by performing a variety of roles, including keeping team members challenged and instilling a positive attitude toward the project and being a communicator [4]. In this role, the leader acts, to maintain the climate of the project and so facilitate project performance, In the role of communicator, the leader communicates with team members about the focus of the project, project changes and develop- ments, and individual member responsibilities. The leader thus ensures a continued awareness of project goals and any changes to them that may have o curred. These leadership behaviors are “behind the scenes” roles that leaders engage in to create a context, within which team members function in carrying out the work of the project. Finally, it has been suggested that the project leader may be a linking pin or bridge between the project team and senior management [3,5,8]. As Ancona and Caldwell [3] observed, pow- erful project leaders are particularly effective politi- cians in lobbying for resources, protecting the group from outside interference, and managing the impres- sions of outsiders. Powerful leaders may command greater respect and, thus, may be able to attract better, project team members to the group (i.e., human re- sources, a stage-setting element) and to Keep groups focused and motivated [8]. By undertaking activities and actions of which team members may not even be aware, leaders may moderate the relationship between, stage-setting elements and project performance and so facilitate more effective performance. Senior management support. Senior manage- ‘ment support of a project team can have a direct affect ‘on performance {27,57,60}. Their support can take a variety of forms, including demonstrating commi ment, helping the team to surmount obstacles, making26 4. PROD INNOV MANAG aowaeBi as things happen, and providing encouragement to the team [6,26,27,57,60]. For example, at 3M’s Occupa- tional Health and Environmental Safety Division, se~ nior management sponsorship was crucial to the suc~ cess of their Action Teams [26]. Their job was to make things happen and to overcome obstacles in the path of the project team, ie., facilitating and enabling success- ful product development, Just as senior management commitment can in- crease the likelihood of project success [60], a lack of support can increase the chances of failure [27]. Hitt et al. [27], who studied one NPD team over time, found that the lack of top management support, as evidenced by a failure to provide encouragement to team members to remain active in the team after they had been transferred to distant locations, played a major role in the failure of the project they studied Champions. Champions, who have been defined a8 someone who “takes an inordinate interest in seeing that a particular process or product is fully developed and marketed” [51], can play an enabling function. ‘Their role can vary significantly, from doing little more than stimulating awareness of an opportunity to playing a major role in overcoming strongly en- trenched resistance by management. Although cham- pions have generally been viewed as contributing to project success, there has, in fact, been little empirical research that supports this contention [35]. The result of recent research has found that the effect of cham- pions on project outcomes is “indirect rather than direct [35]. They operate in concert with processes and strategies the firm uses to improve NPD program performance, which, in tum, leads to improved objec- tive firm-level_ performance.” In this sense, then, champions can be viewed as enablers of the process of developing new products and are, therefore, perhaps most accurately seen as having a moderating effect on the relationship between stage-setting elements and project performance. From the review of the literature on enablers, it is clear that the roles that all three types of Ieaders— senior managers, champions, and team leaders—play are complex. Although a few studies suggest that leaders can have a direct impact on project perfor- ‘mance, the preponderance of research suggests even more strongly that they play an indirect role, moder- ating the effects of stage-setting elements and in so doing impacting indirectly on cross-functional team success. Team Behaviors A third area of research has investigated the relation- ship between stage setters and enablers and team be- haviors, including cooperation, commitment, owner- ship, and mutual respect (6,50,57]. Although st setting elements do not impact directly on performance, both stage setters and enablers can in- fluence team behaviors in important ways [50]. Re- search that has investigated the relationship between team behaviors and both stage-setting elements and enablers is discussed in the following, Cooperation. Cooperation, i.e., working together to accomplish the work of the team, has been variously defined as collaboration, teamwork, interaction, com- ‘munication, and integration (6,28,29,54]. Several stud- ies linked goals—a stage-setting element—with coop- eration [29,50]. Pinto et al. [50], for example, stiggested that goals stimulate cooperation by focusing dividual efforts on specific, common goals, Greater cooperation, in turn, has a positive impact on project performance. Establishing specific goals that the team. ‘can use to measure project progress is another means of promoting integration across functions [6]. Enablers also impact on integration. Project leaders with supe- rior interpersonal skills can play an instrumental role in facilitating relationships and promoting integration, across functions [36]. Commitment. Commitment refers to a sense of duty that the team feels to achieve the project's goals, and to the willingness to do what's needed to make the project successful [6]. Projects that have the “right” mix of human resources, i, the expertise and skills, that are needed to successfully execute project tasks, ‘may enhance each member's confidence in their fel- Jow team members. This confidence, emanating from the expertise and skills that team members possess, may increase the willingness of team members to commit themselves to making a project successful [6] Project managers who gain the commitment of their project team members and actively manage that com- ‘mitment throughout the planning and execution of their projects have a higher probability of achieving, their project objectives [53]. This research reinforces the findings of Pinto et al. [50], who found that team behaviors are intermediate variables. ‘Thus, by properly setting the stage, ie., by staffing the project with the right mix of people and through effective project leadership, the organization can gen- erate greater commitment on the part of the team forSUCCESS OF CROSS-FUNCTIONAL TEAMS, achieving the goals of the project, which can lead to high performance outcomes [31]. Ownership. Ownership refers to the feeling of wanting to make a difference. It goes beyond duty, ic., commitment, in that members of the team begin to tie their identity to a project's outcome, thus putting forth extra effort to ensure its success, Several stage-setting elements have been found to affect ownership of the project, including empowering the team, establishing a climate, and setting goals. Of these three elements, setting clear, focused project goals early on in a project is most likely to foster ownership [6]. To do so, however, it is necessary to involve the project team in translating those goals into specific statements that they can use as guidelines for their work [6], p. 180]. Other researchers found that the actions of a team leader, such as using a participative leadership style, can facilitate or hinder the development of a sense of ‘ownership on the part of the team by fostering greater involvement in translating goals into specific state- ments (26,40). Respect. The respect that team members have for each other can lead to open communication among the team and to feelings of trust [58]. When team mem- bers trust each other's judgments and interact honestly With each other, they are, in effect, exhibiting a form of respect for others. The success of 3M’s Action ‘Teams was dependent, in part, on team members being ‘open and honest with each other [26]. Through their ‘own modeling behavior and interpersonal skills, team leaders can promote respect among team members, thus ensuring the successful use of the human re- sources on the team, In these ways, then, team leaders can serve to moderate the relationship between stage- setting variables and team behaviors. Summary Although the studies reviewed have been useful in highlighting issues of importance to team. perfor- mance, most include only a few of the many factors identified as important in the literature review. Thus, whereas one study found that teamwork is critical, others found that management support is key, and still others emphasize the importance of team leadership. Further, these studies pay little attention to the con- textual effects of such factors as the size of the com- pany or its industry on the performance of its teams. AS a consequence, we have little insight into the relative importance of these factors. The focus on one or a handful of factors cannot help but leave managers | PROD INNOY MANAG: om in a quandary concerning where to concentrate their efforts so as fo implement their cross-functional teams successfully. Studying only a few variables at a time also poses problems for researchers as they attempt to develop models of effective cross-functional team be- havior. Methodology Seven hundred seventy-six questionnaires were mailed to a random sample of the membership of the Product Development & Management Association (PDMA). Because PDMA members are from many different organizational positions, the questionnaire asked re- cipients to forward the questionnaire to the appropriate and knowledgeable individual in their organization, in the event the recipient was not the appropriate person, to respond to the questionnaire, One hundred seventy- two questionnaires were returned, of which 112 were useable in the analysis of success factors, thus yielding a useable response rate of 14%. T-tests were per- formed to address nonresponse bias. A comparison between this sample and two other random samples of PDMA members on annual sales revealed no statisti- cally significant differences, suggesting minimal non- response bias (F = 9.63, p = .002 and F = 12.43, p = 001, respectively).+ Sample The average size of the companies in the sample in terms of number of employees was 2,000 and in terms, of sales revenues was $300 million. Companies in this, sample were quite “healthy,” on average, as indicated by their response to the question concerning change in, revenues. On average, firm revenues increased for the past 5 years by 20%. Respondent firms came from a variety of industries, including consumer goods, ser- vices, and business to business. Variables Following earlier research [37,50], measures of cross- functional team success, ie., project performance, are The studies used in this analysis were Edward F. MeDonough IH and Francis C. Spi, "Managing the New Product Project Potfolio;” the International Product Development & Management Association Confer fence, Atlanta, GA, 1998; and Edward F. MeDenough Ill, Kenseth B. Kab, and Gloria Barczak, “Efectively Managing Global, Co-located and Distributed New Praduct Development Teun,” the Intimatonal Product Development de Management Assocation Conference, Atlant, GA, 1998 respectelym8 1 PROD INNOV MANAG: Monn 23 185 distinguished from firm-level NPD performance, e.g., new product sales and profits. Project performance was measured by asking respondents the extent to which (1) new products were brought to market quickly, (2) budgets for the development of new prod- uucts were met, (3) the quality of products was high, and (4) members of the team were satisfied. The measures were drawn from earlier research that exam- ined project-level performance [37]. Respondents were asked to respond on a five-point scale, where | = not at all and 5 = to a very great extent. As can be seen in Table 1, these items were highly correlated. To explore whether these items should be viewed as a single factor, a factor analysis was conducted. This analysis yielded a single factor comprised of all four items, Factor loadings ranged from .69 to .80. To assess the reliability of combining these four items into a single measure, a reliability analysis was con- ducted, which produced a coefficient of reliability of 76. This level of reliability is considered acceptable for research of the type being conducted here [47]. Firm-level performance was measured with a sin- gle-item scale that asked, “On average for the past 5 years have sales revenues remained unchanged, in- creased (by approximately what percent), or decreased (by approximately what percent).” This variable was labeled “Change in Revenues.” Data were gathered on the extent to which eros functional teams were used to develop new products. Respondents were asked, “What percent of the time is a cross-functional team approach used, where different functions (e.g, R&D, marketing, manufacturing, sales) are members of the same project team and work in collaboration with each other?” Size was measured by the number of full-time em- ployees and by the total sales revenues for the last completed fiscal year. Because these measures wer highly correlated (r = .82, p < 01), only the number of full-time employees was used in the subsequent analysis. Respondents wete asked to indicate into which of five categories their organization fell, <100 ‘Table 1. Correlation Matrix for Performance Measures (n = 106) Budget Quality Quality 40° Satisfection 38 50° Speed 40° 42 53° “7 BF MeDONOUGH It employees, 101-1,000, 1,001-3.000, 3,001-5,000, or >5,000 employees, To obtain data about each company’s industry af- filiation, respondents were asked to identify the prin- cipal products or services offered. Based on these responses, each company was placed in one of three industry groupings: business to business, consumer goods, and services, ‘An open-ended question was used to gather data on why firms adopted a cross-functional team approach, Respondents were asked, “What was the primary rea- son for moving to the cross-functional team ap- proach?” A second open-ended question was used to gather data on the relative importance of factors asso- ciated with cross-functional team success. Respon- dents were asked, “Thinking of your most successful cross-functional project team, why was it so success: ful?” Other research, which investigated factors that, differentiate between successful and unsuccessful cross-functional teams, obtained data on teams that are successful and those that are not. Here, the goal is to obtain data about factors that prior research had found to be associated with cross-functional team success, and then to analyze those data to determine the relative importance of the factors Analysis ‘The following approach was used to analyze responses, to both open-ended questions. Based on a preliminary reading of responses to each question, a coding sheet was created with definitions of (1) reasons for adopt- ing cross-functional teams and (2) success factors (Ex- hibits 1 and 2). All responses by each respondent were coded by two coders using these definitions, In those cases where the same item, ie., reason or factor, was mentioned by a single respondent more than once, the item was listed only once, Where respondents men- tioned multiple and different items, each response was, listed in the appropriate category. Thus, although it, was possible for a single respondent to mention mul- tiple items, the coding scheme did not permit multiple listing of the same item by the same respondent. In instances where a response did not fit into any of the definitions, it was placed in an “Other” category, which consisted of miscellaneous and idiosyncratic items. A reliability analysis was performed to assess the reliability of the coding by the two coders for re- sponses to the question, “What was the primary reason for moving to the cross-functional team approach?”| xn Coding Sheet for Responses to Question about Reasons for Adopting Cross- | Functional Teams Performance Outcome Reasons ‘Speed: To speed up product development, reduce eyele time | j ' | cut To inrve te uty of sew produc ing developed by the organization | neds respond to cusomer expectations, batter consumer | value. caslomer eisfction Success rate: Improve success rate | Larer costs: Lower wait prtuction ests, consol con] improved development cost | Need for new products: Need for new products | I Moinaining schedule: On-target schedules | | Process Improvement Reasons Cross-functional interaction Leverage cross-functional | _scengths. improve coordination | Ownership: Give functions more say and ownership ot products that are developed, increase accountability Motivation and satisfaction: Motivation of staff, employee | eo Resource use: Better utilization of limited statf [_ tno comm ater cort of development proces Convergence was 89%, which is in excess of 85%, the minimum level of acceptability [30]. Because some respondents listed multiple reasons for adopting cross functional teams, a total of 201 responses were ob- tained from 136 respondents. A reliability analysis also was performed to assess the reliability of the coding by the two coders for the question, “Thinking of your most successful cross functional project team, why was it so successfull?” Convergence was 93%, which is in excess of 85%, the minimum level of acceptability [30]. Responses were {PROD INNOW MANAG ns aon. 231 235 sorted into one of the 11 elements discussed in the literature review. In instances where a response did not, fit into any of the II elements, it was placed in an “Other” category. Because some respondents listed multiple success factors, a total of 210 responses were. obtained from 118 respondents. Six respondents pro- vided answers that fell into the “Other” category, and, 23 responses from the other 112 respondents also were. placed in the “Other” category. Three responses in this category pertained to team size, 2 to meeting customer needs, 3 to physical proximity, 2 to having defined responsibilities, and 2 to a shorter time frame. Each of the remaining 12 responses were single mentions of different factors, including, for example, using a struc- tured methodology and being fully staffed, Findings/Discussion Cross-functional teams have not always been an inte- gral part of the NPD process. Indeed, even today not all companies use cross-functional teams and, of those that do, cross-functional teams are not always used in the development of new products. Thus, the interest ‘was in determining the frequency with which firms in the sample used cross-functional teams for NPD. As Table 2 shows, companies in the sample are using cross-functional teams even more extensively than ‘was found by other researchers [12,21]. Ninety-seven percent have used cross-functional teams and 33% use them 100% of the time. In order to ascertain whether the use of the cross functional team approach is the domain of certain companies and not others, the correlations between the use of the cross-functional team approach and the number of full-time employees, sales revenues for the last completed fiscal year, age of the company, and percent change in revenues for the past 5 years were examined, As Table 3 shows, the use of a cross- functional team approach is not correlated with age or size, as measured by number of employees or reve- nues. These data suggest that the use of the cross- functional team approach is scattered across widely differing organizations and is not simply the province of certain size companies or companies of a particular age. It also would appear from these data that the use of cross-functional teams is not associated with firm- level revenue performance ‘The research also sought to determine why compa- nies started using cross-functional teams. Understand- ing the impetus behind the adoption of cross-func- tional teams and understanding what firms hope to2 J,PROD INNOW MANAG Dawn at235 Speed Customer Satstacton Figure 1. Outcome reasons for adopting cross-functional fon bar chart equal percent of repondents jem (n= 136). gain by using cross-functional teams is important be- cause it can provide researchers with a rationale for selecting which performance measures to include in their studies. Hence, respondents were asked, “What was the primary reason for moving to the cross-funi tional team approach?” Responses to this question fell into two major categories, One category of responses pertained to performance outcomes that firms hoped would result from the use of cross-functional teams (Figure 1). The other category pertained to improve- ‘ment in the process of developing new products (Fig- ure 2) As can be seen from Figure 1, the most frequently 60 > — . gop 40 20 | 20 | ee 10 | o/ a. — a8 4 Figure 2. Process reasons for adopting cross-functional teams. ‘Numbers on bar chart equal percent of respondents mention- om (n = 136). ‘mentioned performance outcome reason for moving to, ‘cross-functional team approach was to improve speed to market. Fifty-two percent of respondents in- dicated that reducing cycle time or speeding up devel- ‘opment was the primary outcome reason for adopting cross-functional teams. This is three times the number of respondents who cited developing new products that were of higher quality (18%) or doing a better job of satisfying customers (16%) as the most important reason for adopting cross-functional teams. As can be seen in Figure 2, the principal process reason for adopting cross-functional teams (mentioned by 24% of respondents) was to enhance cross-functional interac- tion, i., the involvement, interaction, and communi- cation across functions involved in NPD, as well as to leverage cross-functional strengths. Because promot- ing closer and more frequent communication and in volvement across functions is an important means of speeding up NPD, it is not surprising that these are seen as the primary reasons for adopting cross-func- tional teams. However, it is surprising that so few respondents viewed them as a vehicle for improving, the NPD process (4%). Finally, it is interesting to note that a few respon- dents (3%) seemed to be unsure as to why their com- panies implemented cross-functional teams, as indi- cated by responses such as it was the philosophy of the president, it was a new approach by other companies, not sure—sounded like a good idea, and new presi- dent. These responses may reflect a tendency on the part of some companies to jump on the bandwagon of new approaches and fads without giving full consid- eration to the appropriateness or desirability of adopt- ing them. The Cross-Functional Team-Performance Relationship A multiple regression was conducted to examine the relationship between cross-functional teams and project-level performance. Project performance was regressed on the use of cross-functional teams, number of employees, revenues, change in revenues, and com- pany age. The overall equation is significant (R? = 21, F = 24.25, p = .000). The results (as shown in Table 4) indicate that only the use of cross-functional teams is significantly associated with performance, indicating that employing cross-functional teams does have a positive impact on project performance. These findings contradict the results of Ancona and Caldwell {2}, who found that the use of cross-functional teams‘SUCCESS OF CROSS-FUNCTIONAL TEAMS {PROD INNOY MANAG avast 221235, Exhibit 2. Coding Sheet for Responses to Question about Success Factors Goats: The objectives for a project eam, Terms used to describe goals include vision, common focus, direction. Empowerment: The decision-making authority or responsibility that has been given to the team. ‘Terms used to describe ‘empowerment include freedom, decentralization, allowing teams to establish their own rules and procedures. Climate: The feelings that the team has about the project or being on i, Terms used to describe climate include a sense of urgency, excitement, importance ‘Human resources: The capsbilities that arc embodied in the members ofthe team. Terms used to describe a team’s hnman resources include expertise, experience, skills, and competence, The “chemistry” of the team, Le, Working well ogether, is also a resource. Senior manager support: The support that senior management provides the team. Terms used to describe senior management ‘suppor include senior management commitment, enabling, overcoming obstacles, making things happen, facilitating, encouraging ‘Champions: Someone who “takes an inordinate interest in seeing that particular process or product is fully developed and ‘marketed. The distinetion between senior management support and champions is that a “champion” is not a formal role in aa organization Team leadership: ‘The support, facilitation, and encouragement provided by the team leader. Cooperation: Working together to accomplish the work of the team. Terms used to describe cooperation include collaboration, teamwork, interaction, communication, and intergration, Commitment: The sense of duty that the team feels to achieve the project’s goals. Refers to a willingness to do what's needed to make the project successful or toa sense of “buy-in” tothe project. Ownership: ‘The feeling of being able to make a difference and wanting to do so. It goes beyond duty, i.c., commitment, in that ‘members of the team begins to tie their identity toa project’s outcome, thus putting forth extra effort to ensure its success. Respecvfrust: The faith that team members have in each other, their abilities, or their judgments, Terms used include the trust and ‘mutual respect team members have for each other. was associated with lower team-rated performance. ‘These results also indicate that not only are cross- functional teams associated with bringing new prod- uucts to market quickly [22,61], but that their use is associated with meeting development budgets, devel- oping high-quality products, and team member satis- faction. Thus, although cross-functional team use was not found to be associated with company-level reve~ nue performance, the findings do suggest that teams are a powerful tool for affecting performance at the project level. The Relative Importance of Success Factors ‘The findings regarding the relative frequency with which success factors were mentioned are shown in Figure 3. These results have been organized into the three categories used to organize the literature review (stage setters, enablers, and team behaviors) and have been rank ordered in terms of the frequency with which each element was mentioned. As shown in Figure 3, establishing clear, unchanging goals, team. leadership, and cooperation were the three most fre- quently cited reasons for cross-functional team suc- cess. These results did not differ significantly by size or product type. It is interesting to note that each of these three factors (goals, team leadership, and cooperation) is, from a different category. This suggests that the route to project success is a complex one and may involve the intricate interplay among elements from different categories and in different combinations to bring about{PROD INNOY MANAG Montane Table 2. Frequency of Use of a Cross-Functional ‘Team Approach ‘Number of Percent of Frequency of Use Firms Firms 0% 5 328 110% 8 5.1% 11-49% 16 103% 50-78% a 153% 75-09% 31 32.7% 3338 00% 52 successful performance. For example, establishing clear, stable goals may not only provide the team with direction and focus for their decision making, but it also may help to foster cooperation. In addition, the team leader also may play an enabling role in fostering cooperation among team members and help to keep the team focused on its goals. Within each category, it was found that one variable ‘was mentioned much more frequently than the others as being associated with project success. For exampl of the enabler variables, team leaders, who are typi cally closest to the team and interact with them mo often, were associated with cross-functional team suc cess more often than were senior managers or cham- pions. It is possible that the team leader may have a greater influence on the team’s behavior due to their ability to intercede more immediately when conflicts arise and to more continually ensure that the team’s activities remain focused on the project's goals. ‘The fact that senior managers are less often associated with success may be due to their being more removed from the day-to-day workings of the team than the team leader. It also may suggest that strong and effective team leaders may be able to succeed despite the lack of senior management support. Given the extensive writ- ing about the apparent importance of champions, it was surprising to find they were rarely seen as con- tributing to success. It may simply be that compared to ‘Table 3. Correlation Coefficients Among Cross- Functional Teams and Demographic Variables Size Revenues Age cris Size 00 Revenues 6 sit Age 08 234 324 Change Cee) —2* a7 * Concation is significant at the 05 level + Comtelation is significant atthe 1 level 5. MCDONOUGH AT ble 4. Multiple Regression analysis of Cross- Functional Teams, Number of Employees, Company Age, Revenues, and Change in Revenues on Project, Performance Project Peformance Constant 256" Cross-functional teams a6" No. of employees 03 Revenues ol ‘Change in revenues 06 Company age —07 Adjusted 20 F 24.25 P 000 peo formal role players—team leaders and senior manag- cers—they are needed only infrequently, when the team, leader alone is unable to achieve results or when senior management support is totally lacking Cooperation was found to be the most frequently mentioned element in the team behavior category. Only goal setting was mentioned with equal frequency as contributing to success. Although commitment and, ownership also were mentioned frequently, these ele- ments refe in a sense, to a deeper level of coopera- tion, Although commitment and ownership may be needed in some projects in order for them to be suc- cessful, more commonly, achieving cooperation among team members may be all that is. required. ‘When cooperation is viewed as an initial step toward sgt eters erates Temas Figure 3. Success elements by category for entire sample. "Numbers on bar chart equal percent of respondents mention- ing an item (n= 112),commitment and ownership, it is perhaps not surpris- ing that these two elements would be mentioned less often. Although this research found that all of the variables included in the literature review were perceived to be associated with project success, they were not men- tioned with equal frequency. Clearly, more research is needed that investigates the relative importance of these variables to project success and, of equal impor- tance, that investigates the relative importance of vari- ables within each category. It is important to keep in mind that the findings of this stady can only suggest possible links among these variables and, given the nature of the research presented here, itis only possi- ble to speculate the about the underlying re: relationships among them. A Proposed Model It is clear that more methodologically and statistically rigorous investigation of the factors associated with cross-functional team success is needed in order to more confidently suggest the relative importance of success elements and the effects of contextual factors on the relationships among these elements, To help organize and focus future research, a model of the relationships among stage-setting elements, enablers, team behaviors, and performance that is based on the review of the literature, as well as the results of this study, is proposed (Figure 4). ‘The model proposes that stage-setting elements op- crate on project outcomes indirectly by affecting the behavior of the team, which in turn influences project performance, thus implying that, in the absence of ___ Sy Team tears) | } | Leconte Cremer) 7 Figure 4. Proposed model of interactions among stage setters, ‘enablers, team behaviors, and cross-functional team success. {PROD INNOY MANAG: = aoweatas cooperation, ownership, and commitment, stage-set- ting elements are unlikely to have a sufficient impact to generate successful project performance. The model also proposes that enablers influence project perfor- mance principally by influencing team behaviors rather than by having a direct impact on project out- comes, However, the model proposes that enablers play another important role, Senior managers, team leaders, and champions are able to influence team behaviors by moderating the effects of the different, stage-setting elements. It is the behavior of these in- dividuals, in addition to the foundation that they have laid for NPD projects, that can have an important influence on the cooperation, commitment, ownership. and mutual respect among team members. Thus, the model recognizes that managers are responsible for both establishing stage-setting elements and, while the project is underway, engaging in behaviors that can significantly influence the behaviors exhibited by the team, Finally, the model recognizes the impact of contex- tual effects on stage setters, enablers, and team behav- iors. Although the findings of this study suggest that size and industry may not impact on the relative im- portance of the elements in the model, clearly addi- tional research is needed in order to more fully under- stand the effects of these, and other, contextual variables. Conclusion ‘One of the goals of this research is to provide insight into which factors affect the success of cross-fune- tional teams. Within each of the three categories, it was found that one element is mentioned especially frequently as being important to success. At the same time, a set of secondary, but still important, elements was identified within each category. Recognizing the existence of a primary and secondary set of elements can be quite important to managers who are often forced to make project-related decisions within time and resource constraints. A second goal of this research is to propose a model to be used in future research on cross-functional teams, effectiveness. Inherent in the model is the recognition, of the complex and dynamic character of the NPD. process and of cross-functional teams. This complex- ity argues for an equally complex and more encom- passing approach to future research. Such studies would need to explicitly examine the interrelationships among elements within categories, as well as betweendentate AS them, for example, investigating to what extent team leaders and senior managers actually affect team be- haviors and how they moderate the relationship be~ tween stage setters and team behaviors. Research em- ploying quantitative analysis of the model would require the use of multi-item measures of each of the elements in the model and on more objective measures of cross-functional team performance. These, in turn, would enable the use of path analysis, e.g., to deter- mine relationships among elements and causality. We also need a deeper understanding of how each of the elements in the model interact and in what sequence. Although both senior managers and team leaders are important in enabling project success, we need to determine when intervention by each is most effective and what form of intervention is most useful. Thus, while further quantitative analyses are useful, it is pethaps even more important to understand what actions are taken, by whom, when, and for what pur- pose. To obtain insight into these sorts of questions requires a case-oriented approach and the application of a grounded theory methodology. In addition, although a few contingencies that im- pact on managing cross-functional NPD teams suc- cessfully have been identified here, clearly there are many other variables that require additional investiga- tion, including the nature of the development being undertaken, i.e., the technological complexity and newness involved in the development process, the newness of the market, the effects of geographical dispersion of team members, and the importance of speed to market. Again it would be appropriate to use ‘a qualitative research approach, such as grounded the- ory, (0 generate a more detailed and in-depth under- standing of the moderating effects of contextual fac~ tors, including, for example, industry and size, on other elements in the model. Such research might include in-depth interviews with new product teams, as well as with individuals at multiple levels within the organization, | would like to acknowledge support from PDMA for this research, | would also like to thank two reviewers, xs well as the editor of PIM, for their very helpful comments and extensive feedback, References I. Adler, PS, Itendepartmenta intenependence and coontination: The ‘ease of the desighfmanufactting interface. Organization Science (6147-167 (1995), 2. 2 2, 2, 2 EF. MCDONOUGH IT ‘Ancona, D.G. and Caldwell,D.F. Demography and design: Predictors of new product team performance. Organization Science 3:321-341 (4993), ‘Ancona, DG. and Caldwell, D. F, Bridging the boundary: External process and peformance in organizational tems, Administrative Sei- fence Quarterly 37:634-665 (1992), Barczak, G, and Wilemon, DL. Leadership differences in new pr! ‘et development teams Journal of Preduct Innovation Management (6289-267 (1989) Brown, S. L. and Eisenhardt, K. M. Product development: Past re search, present findings, nd ature detions, Academy of Manage: ment Review 20:343-78 (1995). Bowen, H.K. Clark, K.B., Holloway, C. A, and Wheehwriht S.C (eds), The Perpecual Enterprise Machine. New York: Onford Univer: sity Press, 1994, ‘Carmel, E. Cycle time in packaged software firms. Journal of Product, Tanovation Management 121-14 (1995). Clark, K.B. and Fojimovo, T. Prodet Development Performance Boston, MA Harvard Business School Pes, 1991 Clark, KB. and Wheelwright, 8. C. Organizing and leading Nesey ‘weight development feams, California Munsyenent Review 34-28 (982), Cooper, R. G. and Kleinschmidt, PJ. New products: What separates winners from losers? Joural of Product Innovation Management 4490-111 (1987) Cooper, R. G. and Kleinschmidt, B. J. Major new products: What disnguises the winners in the chenica industry, Jounal of Prodet Innovation Management 10:90-111 (1993) Cooper, R.G, and Klcinschmidt, E Determinants of tieliness in rredvet development. Joan of Product Innovation Management 11:381-396 (1954), Denison, D. R., Hat, S. L. and Kabo, J. A, From chimneys 10