| Abstract: | Integration of research and development (R&D) with marketing remains a frequent topic in the new product development (NPD) literature, largely because it represents a critical antecedent of new product performance (NPP). Two divergent opinions about this integration exist, such that those who contend that firms should pursue high levels of integration in every case provoke criticisms from those who propose that various NPD processes require different levels of integration. This paper proposes that the two perspectives can be reconciled by taking into account the fact that R&D and marketing are integrated mainly to combine critical knowledge (technological and market) that otherwise would be separate to achieve market success. Following Danneels's approach, we investigate how the effect of R&D–marketing integration on performance change across four types of NPD processes: pure exploitation, pure exploration, technological competence exploitation, and market competence exploitation. Data derived from a deep study of 11 NPD projects by five firms, analyzed through qualitative methods, highlight the necessity to vary the level of integration according to the type of competence to be developed during the NPD process. Our analysis suggests two main conclusions. First, the effect of integration depends strictly on the type of competence that the firm uses to develop and launch a new product. Second, integration does not have a unique effect on performance, but it is necessary to distinguish between market performance (e.g., sales and market share) and process performance (e.g., meeting the planned budget and time to market). In some projects, the effect of integration on the two types of performance is diametrically opposite. In particular, we propose that (1) higher performance will be associated with lower integration in pure exploitative projects; (2) in projects that exploit existing market knowledge, higher market performance will be associated with a higher integration, although these projects tend to offer poor process performance regardless of integration level; (3) in projects that exploit technical knowledge, higher performance will be associated with higher integration; and (4) higher integration will be associated with higher market performance but poorer process performance in pure explorative projects. |
