1 Information Spillovers and Strategic Behaviors in Open Innovation Crowdsourcing Contests: An Empirical Investigation

Designing optimal contests is crucial to the success of online crowdsourcing markets that seek to attract a diverse crowd of high quality participants. While there has been some theoretical work examining the design of different types of contests, there are hardly any empirical studies on the effectiveness of different types of contests for crowdsourcing. This study analyzes data from a large online crowdsourcing market for design tasks, to examine the strategic behavior of participants faced with a trade-off between two competing forces in open innovation contests. Open innovation contests typically, make information about contestants and their submissions visible to all participants, thereby allowing later contestants to imitate or learn from earlier entries, favoring later submissions. On the other hand, early entries are more likely to get valuable feedback from the contest holder – feedback that can help these early entrants to significantly revise their solutions and resubmit. Timing of entry, thus, becomes an important strategic choice for contestants. Our findings show that open innovation contests are subject to significant informational spillovers relating to submissions that benefit later entrants. More interestingly, we find that the benefits of informational spillovers from feedback provided to early entries depend on the type of feedback provided. We contrast these findings with the outcomes of blinded contests that prevent informational spillovers. Our findings have interesting implications for the design of innovation contests, as well as for market makers, and for contest holders in online crowdsourcing markets.