Economics of organizing product development in the extended enterprise

Firms are increasingly facing the challenge of organizing product development effort in extended settings, i.e., across the boundaries of several firms. This dissertation looks at the implications of this trend to the development cost structure, in particular to the coordination cost.; The theoretical framework is based on transaction cost economics, design theory and organizational learning. Building on this literature, the dissertation addresses the following questions: what are the indicators of coordination costs for different types of development tasks? Do these help explain the task sourcing decision? And, do firms learn how to coordinate development tasks?; Three empirical models were developed to address these questions. The first model proposes that a set of task attributes derived from transaction cost economics, including a proxy for asset specificity, affect both the task sourcing decision and its coordination cost. Then, I test for differences in the coordination cost of 'generation' and 'test' types of tasks, a classic distinction in design theory. Finally, the dissertation tests a 'learning model' for development task coordination effort. The data was collected from 11 software development projects conducted at a global firm in the medical device industry. The total sample size consists of 71 system development tasks.; The main findings were as follows. The proxy for asset specificity, internal problem solving, is a significant predictor of both the likelihood of outsourcing a development task and the associated coordination cost. Moreover, the impact of internal problem solving on the coordination cost is significantly larger for outsourced tasks. The results also reveal asymmetries in the coordination cost for generation and test tasks. Finally, I find evidence that projects were able to reduce coordination effort over time.; The findings of the dissertation should be useful in several ways. Recognizing indicators of coordination cost and the asymmetric nature of coordination costs for different types of tasks may provide a more principled approach for organizing extended product development. By systematically exploring the ability to reduce coordination effort through competence development and management of the task dependence structure, we may be able to discover more efficient ways for engaging external partners in development efforts.