Turning creative ideas into successful innovations: differential effects of network structure for radical and incremental innovation

This paper examines how the collaboration network structure of an innovation site influences the adoption and future use of its innovations. We explore the effects of tie strength and network cohesion, with a particular focus on the moderating role of innovation radicalness. While prior research emphasizes the benefits of strong ties and network cohesion for idea transfer—due to increased trust, information exchange, and reciprocity—we argue that these effects are contingent on the innovation’s radicalness. Specifically, we suggest that these effects hold for incremental innovations but may become negative for radical innovations, as the impact of radical innovations may not align with reciprocity norms and could be sanctioned by the network. Additionally, the lack of information diversity may hinder the identification of new applications for radical innovations. Our empirical analysis is based on a dataset of 93 of the most innovative U.S. pharmaceutical and biotechnology companies, with 16,011 unique sites observed from 2001 to 2013. This results in a panel dataset with 19,343 site-time observations, using 3-year rolling windows. Our findings support our hypotheses, contributing to the literature on social networks, creativity, and innovation. We show that different types of innovations require different network conditions for diffusion, and that reciprocity norms can be burdensome, particularly for radical innovations. We also demonstrate that non-redundant information is crucial not only for generating novel ideas but also for identifying new applications for radical innovations. The findings have implications for innovation management, particularly at geographically dispersed sites.