Abstract
The major premise of this study is that managers purposefully shape the business context for radical innovation. Particularly, the strategic option of developing radical innovation in collaboration with direct competitors offers opportunities otherwise unattainable. We tap into its cognitive underpinnings by running an experimental study of coopetition design for radical innovation. We have collected 5760 binary decisions from a sample of 160 managers. Their indications are used to run a choice-based conjoint analysis in order to identify utilities attributed to coopetition-shaping decisions in a radical innovation project (using a scenario of self-driving/electric cars produced by VW, Daimler or Tesla). We use Hierarchical Bayes Multinomial Logit Regression to test a set of four hypotheses, each addressing a different coopetition factor to unveil manager's preferences in coopetition design for radical innovation. Our findings pinpoint a clear preference for network coopetition, using formal governance, and being based on intensive knowledge sharing. Contrary to prior literature, market uncertainty does not appear to significantly influence coopetition design for radical innovation.