How new knowledge search intensity influences firm innovation performance: the moderating roles of network assortativity and modularity

Purpose This study aims to examine how new knowledge search intensity affects firm innovation performance, focusing on the moderating roles of degree assortativity and modularity in intra-firm collaboration networks. Design/methodology/approach This research analyzes a sample of semiconductor patents from the United States Patent and Trademark Office database between 1975 and 2007, using negative binomial regression models with fixed effects for the panel data. Findings The results show that an increase in new knowledge search intensity can promote firms’ innovation performance, and this effect is influenced by the structural characteristics of collaboration networks (degree assortativity and network modularity). Firms with higher levels of degree assortativity in their intra-firm collaboration networks strengthen the positive relationship between new knowledge search intensity and innovation performance. Similarly, higher levels of network modularity also enhance the positive impact of new knowledge search intensity on innovation performance. Practical implications Firms that achieve higher innovation performance should not only enhance their strategies for new knowledge search but also consider the mechanism design of collaboration. Facilitating collaboration among inventors with similar levels of network centrality and fostering modular, community-based collaboration structures can help firms realize the innovative potential of new knowledge searches. Originality/value This research advances the understanding of how knowledge search influences innovation outcomes and highlights the moderating role of collaboration network structures. By integrating perspectives from knowledge search and collaboration network literature, this study expands current understanding of how firm-level search behavior can enhance innovation capabilities. Additionally, it enriches the literature on collaboration networks, particularly the effects of degree assortativity and modularity on innovation.