How the timing of cooperation affects innovation outcomes: An agent-based model of postponing the coupled search process

Organizations may face the problem of whether and when they should cooperate with others to accomplish a novel task. Prior literature answers this problem based on the transaction costs of cooperation, but ignores the effect of cooperation timing on innovation outcomes. Will postponing cooperation influence innovation outcomes? We develop a conceptual agent-based search and learning model to explore this question. We find that the timing of cooperation, the initial mental representations, and the search strategy jointly determine innovation search outcomes. Agents with initial mental representations that are inclined towards suboptimal outcomes can benefit more from postponing cooperation when the search strategy is exploitative. This study inserts the practical innovation problem into the framework of adaptive learning models and provides a novel perspective from which to understand the timing of cooperation.