Agency Conflict in Crowdsourced Corporate Innovation Contests

We investigate crowd versus expert selection in the context of a recurring innovation contest at a global technology company. Textual analysis of 14,697 submitted ideas reveals that agency conflict among the two groups in the... [ view full abstract ]

We investigate crowd versus expert selection in the context of a recurring innovation contest at a global technology company. Textual analysis of 14,697 submitted ideas reveals that agency conflict among the two groups in the contest. Crowds of company employees are influenced by topic popularity and social concerns, while experts focus on strategic alignment. While experts exhibit less agency conflict than employees relative to stated corporate strategy, they are far less numerous and potentially more expensive than crowd resources. We identify hybrid mechanisms that balance the use of scarce expert resources with the potential selection biases of the crowd.