آشکار ساختن پویایی ها و پیامدهای تناسب و غیرمتناسب با محیط بین شبکه های رسمی و غیر رسمی در همکاری های توسعه محصول چند نهادی

The study presents a longitudinal examination about dynamics and consequences of fit and misfit between formally ascribed design interfaces and informal communication networks in two large multi-institutional product development collaborations in space industry. Findings: (1) formally ascribed design interfaces and informal communication networks correlate only marginally. The main reason is that informal communication is much more dense than ascribed; (2) although the formally ascribed design interfaces change, the structure of informal communication remains largely stable throughout time; (3) the most intriguing finding is that this communicational misfit is associated with higher effectiveness, but, it negatively impacts the institutional unit's efficiency.

In today's fast-paced, knowledge-intensive environments, multi-institutional product development collaborations have become a popular vehicle for executing innovative efforts and serve as an inevitable answer to environmental and market developments (Oxley and Sampson, 2004). The exponential growth in knowledge is likely the most prominent driving factor for the increasing number of multi-institutional product development collaborations. In addition as product development and research become increasingly multidisciplinary, research managers and policy-makers are relying more on multi-institutional collaborations to develop strong, intellectually diverse projects that can answer complex research questions (Carley et al., 2006). Multi-institutional product development collaborations are large-scale projects consisting as building block of units dispersed over a number of different institutions and most often also over a number of different countries. Engineering technically advanced and complex products requires tapping into and capitalizing on knowledge concentrated in pockets of excellence around the globe (Leenders et al., 2003). As recent research exemplifies the number of multi-institutional product development collaborations and their size exponentially grows in many business sectors (e.g., Carley et al., 2006, Carlsson, 2006 and Roijakkers and Hagedorn, 2006). Researchers and practitioners unanimously agree that effective management plays a critical role in ensuring the success of multi-institutional product development collaborations (Pinto and Prescott, 1988). However, most managerial concepts and knowledge in the field of new product development refer to smaller-scale projects consisting of a single (or few) project unit(s) within a single institution. This obvious deficiency may be responsible for a ‘performance paradox’ as Flyvbjerg et al. (2003) stress: “At the same time as many more and much larger infrastructure projects are being proposed and built around the world, it is becoming clear that many such projects have strikingly poor performance records ….” (p. 3) (see also Gruen, 2004 and Morris and Hough, 1987). Today, it is common practice to manage the embedded knowledge in product development by using a stage-gate new product development process (e.g., MacCormack et al., 2001). In employing a stage-gate approach, new product development processes are usually modeled by dividing design activities into a number of sequential phases or stages and separating the stages by milestones called gates (e.g., Cooper, 1990 and Ulrich and Eppinger, 2004). Although the number of phases or stages differ, most stage-gate processes contain the same activities—conceptualization, design, and integration (e.g., MacCormack et al., 2001). In the early conceptual phases of the design process, the design space is likely to be very large and complex (Dym, 1994). In order to reduce this complexity, a hierarchical decomposition procedure is used to break down the design task into semi-independent sub-tasks corresponding to the functional components of the planned product (Simon, 1996). In executing the design tasks in multi-institutional product development collaborations, most commonly a number of units distributed over the involved institutions are responsible for designing the components of the final product (Sosa et al., 2004). In the later integration stages, after completion of the unit-specific tasks, the project break-down structure reverses, thereby forming a hierarchical ‘recomposition’ of the project. The managerial process of ‘breaking down’ the project into components, and its subsequent reversal, can be visualized and managed using the design structure matrix framework of Steward (1981). The ‘Design Structure Matrix’ (DSM) framework is a managerial tool used to address and define the unit-specific interfaces within the collaboration and their dynamics throughout the stages of new product development (Hellström, 2005 and Yassine et al., 1999). The unit-specific interfaces represent product component related dependencies, which inherently create the requirement of interrelated design activities. For example, an electric shaver is composed of six main functional parts or components, the protection cap, the shaving unit, the driving unit, the power supply, the housing assy, and the cover assy. There are different degrees of dependencies between these components (Oosterman, 2001). The driving unit and the shaving unit have certainly strong dependencies and exemplify a strong interface, the dependency between power supply and shaving unit is weaker and can be characterized as weak interface, whereas the power supply and the protection cap have no direct dependencies, thus no interface. Subsequently interfaces can be rather weak or strong reflecting the required extent of the unit's mutual design adjustments or absent indicating that there is no requirement for mutual design adjustments. The dynamical changes throughout product development phases highlight the importance of addressing and defining the interfaces between product components at the design and integration phases; it is also important to determine if the units actually interact according to their formally ascribed interfaces, an inevitable requirement for the collaborations to function (e.g., Dahl and Pederson, 2004, Thompson, 1967 and Galbraith, 1973). Unfortunately, informal communication networks often compete with aspects of organizations as formal structure (Cross et al., 2002). One of the most consistent findings in the social science literature is that who you know often has a great deal to do with what you come to know (e.g., Granovetter, 1973, Burt, 1992, Rogers, 1995 and Szulanski, 1996). In multi-institutional product development collaborations, therefore, it would be naïve to expect a perfect alignment between design interfaces – the ‘Design Structure Matrix’ (DSM) – and the informal communication network – the ‘Informal Communication Matrix’ (ICM) – as Sosa et al. (2004) have shown. These considerations raise the following research questions: (1) How strongly do the ascribed design interfaces (DSMs) and informal communication networks (ICMs) overlap? (2) How do the ascribed design interfaces and informal communication networks, and their overlaps, change when the process moves from the design phase to the integration phase? (3) What are the reasons for possible alignments and misalignments of DSMs and ICMs? (4) What consequences do alignments and misalignments have for goal achievement? (5) How can the management alignments and misalignments of DSMs and ICMs and goal achievement be brought into balance? Obtaining answers to these questions is of major importance to improving the management of multi-institutional product development collaborations because it allows concurrent balancing of the formal organizational structure, as mirrored in the DSMs, and the informal organizational structure, as reflected in the ICMs, throughout phases of new product development. In addition, answering these questions broadens the base of systematic research and empirical data on large multi-institutional collaborations, as suggested by Gruen (2004), and it supplements the work of Sosa et al. (2004). Our paper contributes to the field of innovation management by offering new insights and explanations for the very nature and dynamics of multi-institutional product development collaborations. Because there are few empirical and theoretical insights into the functioning of such projects, our study is organized in an exploratory and a testing part. In the first part of the study we explore the dynamics of formal and informal networks in multi-institutional product development organizations. In this part we are unable to formulate explicit hypotheses based on the present knowledge. In the second part examining the consequences of the revealed network dynamics we are able to formulate hypotheses based on our exploration and preceding research in social networks. In order to answer the questions raised above, we investigated two typically multi-institutional product development collaborations in space industry; these projects feature large sizes, complex structures, long tenures, and highly complex design tasks. An overview of the organization of our paper and a preview of our main findings is as follows. In Section 2 we discuss our research sample and methods. In Section 3 we explore the fit and misfit between the DSMs and ICMs at the design and integration phases of our two multi-institutional projects,1 finding that: (1) there are large misalignments between the DSMs and ICMs, (2) the informal communication networks are much more vivid than are the formally ascribed interaction networks, (3) the DSMs undergo structural changes in the transition from the design to the integration phase, and (4) the structure of informal communication is stable during transition from the design to the integration phase of the collaborations. In Section 4 we introduce, define, and empirically explore factors that may explain the structural characteristics of the DSMs and ICMs in the transition from the design phase to the integration phase. This exploration leads to the result that: (1) work-related changes of interfaces based on hierarchical decomposition and ‘recomposition’ account for the changing structure of DSMs, and (2) the structure of the ICMs is determined by a core-community of institutions, by larger-sized units, by a certain location, by a higher percentage of astronomers, and by the structure of DSMs. In Section 5 we empirically test the consequences of the misalignments between the DSMs and ICMs for units’ goal achievement based on a number of hypotheses. We find that vivid and stable informal communication contributes to unit effectiveness at the design phase and is a drawback to unit efficiency at the design and integration phases. Finally, in Section 6 we summarize and discuss the implications of our findings, and in Section 7 we draw scientific and managerial conclusions from our study.

The global business world is denoted by an increasing number of multi-institutional product development collaborations. However, there is little empirical evidence to inform managerial decisions and practice on how to improve the success of such collaborations. Our study attempts to lessen this knowledge gap and is innovative in several respects. Whereas previous studies have typically investigated the formally ascribed design interfaces and the informal communication networks from separate viewpoints, we integrated these two perspectives and examined the structure and alignment of both. Moreover, ours was a longitudinal study through two phases of the new product development process, not a cross-sectional study as typifies the empirical literature on multi-institutional product development collaborations. In this way we added to the literature about stability and change of social networks. Moreover, our study directly tackled the controversial debate that has arisen in recent years about the contributions and drawbacks of utilizing informal networks to achieve organizational goals. In attempting to resolve this debate, we showed and discussed the positive and negative effects of misalignments between formally ascribed design interfaces and informal communication networks at design and integration phases of two multi-institutional product development collaborations. As result, we believe that our paper offers entirely new insights for managers regarding how to balance the structural set-up of multi-institutional product development collaborations throughout development stages and offers advice about how to manage the dynamics of informal networks to obtain the best trade-off between two project goals—unit efficiency and unit effectiveness. Hence, we offer insights into the new managerial challenges of current and future multi-institutional product development collaborations. Generally, our work contributes to the product innovation literature (1) by uncovering factors that explain, in part, the stability, structure, and alterations occurring in both formally ascribed design interfaces and informal communication networks as product development move from design to integration phases, and (2) by investigating the alignments between formally ascribed design interfaces and informal communication networks, as well as their consequences for achievement of goals at these two phases. Our results have four important scientific implications. First, we clearly demonstrated that informal communication networks have their own characteristics and dynamics. In contrast to a prevalent assumption in prior literature, we found highly dense and stable informal networks. Our analyses also imply that these dense and stable informal communication networks are likely to be the same in other collaborations because density and stability are largely caused by such general structures as composition, size, and location of units, and by the formally ascribed channels of communication among them. In fact, we strongly assume that our main results can be generalized for similar collaborations in other branches, for example, the automobile or airplane industry because most characteristics in these industry sectors are the same or at least comparable settings to space industry, such as the degree of specialization, the product development phases, and the international composition. These findings, however, raise the question: How can informal networks be managed in multi-institutional product development collaborations? Our study cannot fully answer this question and we urge that it be central in pursuing this kind of research. A second scientific implication of our study is that the systematic design methods of hierarchical ‘decomposition’ and ‘recomposition’ are powerful factors in explaining the structural patterns of informal communication, and also in explaining alternating formal structures. Therefore, the relationships between and among systematic design methods, the resulting informal structures, and the changing formal structures, should receive more scientific attention. A third important implication of our results is that frequency of communication may be a two-sided coin; greater frequency of communication does not necessarily result in a higher performance and vice-versa. Moreover, it could be that efficiency and effectiveness are not one-dimensional factors, and may actually display contradictory aspects of goal achievement. Ensuring goal achievement may be a case of trying to find an appropriate managerial trade-off in support of effectiveness and efficiency. Our study has shown that vivid and misaligned informal communication fostered achievement of effectiveness goals, but, at the same time, prevented high efficiency. And finally, we found that social network methods proved to be extremely fruitful and valuable when investigating organization issues that are somehow based on structures of interaction. Translated into managerial terms, our results indicated that it may be advisable to define a phase-specific hierarchy of goals. As we found, the degree of overlap between the formally ascribed design interfaces and the informal communication networks appear to serve different goals. Also our findings suggest that unit location, size, and composition, and their formally ascribed channels of communication, may help to increase the frequency of communication, and in turn may enhance the effectiveness of such collaborations. In contrast, the same kinds of units may prevent project efficiency. So it is up to general managers of the project to balance these central variables in order to achieve the best trade-off between efficiency and effectiveness at different product development phases. Likewise, as De Meyer et al. (2002) argue, such collaborations are exposed to altering types of uncertainty so that situation- or phase-specific managerial approaches are inevitable. Moreover, the special dynamics of the informal network should be more strongly considered at the outset of such collaborations. When it is known that the informal structures are very dense and stable, certain communication tools may be applied to help managers accommodate and if possible, to help them balance these structures; one such tool is communication media with different ‘bandwidths’. Our empirical study, like others that involved collection of data in a single or a very small number of multi-institutional product development collaborations, limit the range of our findings; a limitation that would be lessened by completing a number of similar studies in other types of products and in different industries. We would like to encourage future researchers to do just this and to enlarge the body of such empirical investigations. That said, our study is partly explorative, as such, we avoid drawing explicit predictive or prescriptive conclusions about the dynamics of formal and informal networks in multi-institutional product development collaborations. We hope that future researchers will build upon our work, and working toward drawing more explicit predictive and prescriptive conclusions. We only could partly fulfill this research demand. In addition, we express a hope that future researchers may have greater concentration on options for managerial interventions to control and balance informal communication.