Managing Innovation Adoption in Organizations

In organizations, experimentation with innovations is often decentralized across multiple units (e.g., functional areas, locations, subsidiaries, etc.). Because adopting innovations is risky, each unit may prefer to delay adoption and wait for information about the experience of others. Such delay could conflict with broader organizational learning because innovations may not be sufficiently tested early, slowing or even preventing adoption. Theoretically, in this dynamic social learning environment, the central planner (corporate manager) can reduce the benefit of information free-riding and help increase collective experimentation via binding and non-binding recommendations that discourage future innovation adoption. This paper uses lab experiments to investigate how human decision makers behave in these novel strategic interactions and their ability to resolve information free-riding. We find that subjects respond to the intervention mechanisms in the direction predicted by the theory; however, the magnitude of their response is insufficient to generate the expected benefits of the interventions. In addition, we find that subjects acting as the corporate manager who selects the information mechanism appear to anticipate the suboptimal adoption decisions of subordinates and frequently choose recommendation rules that do not discourage innovation adoption.