On the "right" problem formulation: The effects of varying problem-related information on solver participation in crowdsourcing

Inviting firm-external contributors to provide solutions to defined problems has become increasingly popular to foster innovation. A key concern in broadcast search and tournament-based crowdsourcing is to receive a high... [ view full abstract ]

Inviting firm-external contributors to provide solutions to defined problems has become increasingly popular to foster innovation. A key concern in broadcast search and tournament-based crowdsourcing is to receive a high number of solution proposals. More solution submissions correlate with a higher diversity of solution proposals. If the quality of contributions made by individual participants was totally random, getting a larger crowd of contributors would be the only way to increase the odds for successful crowdsourcing. Compared to the manipulation of other organizational factors to boost submissions, there is only little research on the potential influence of the problem formulation. We propose that the participation to a broadcasted problem is dependent on whether or not the task description contains (1) need-related information and (2) information about the solution requirements. In addition, we explore whether (3) the type of evaluators designated to review the submitted ideas (user community vs. company experts) relates to participation. We test our propositions with a large-scale between-subjects experiment inviting the members of a firm-hosted community for sound products to provide innovative solutions for improving the comfort of headphones. The results show that the number of submitted problem solutions varies largely depending on the problem formulation while keeping the problem itself and its context constant. The most appealing and stimulating problem statement in this experiment almost doubled the number of submissions than the least activating problem description. Our findings suggest that adding information on solution requirements and involving the community in the idea evaluation process may represent “cheap” but effective means in order to increase solver participation in crowdsourcing.