Effects of learning and uncertainty on crowdsourcing performance of solvers: insights from performance feedback theory

Purpose In crowdsourcing contests, the capabilities and performance of individual workers (solvers) determine whether seeker firms can obtain satisfactory solutions from the platform. It is noted that solvers may learn such skills in crowdsourcing from doing (experiential learning) or observing (vicarious learning). However, it remains unclear if such learning can be materialized into improved performance considering the unique settings of crowdsourcing contests. The study aims to understand how experiential learning and vicarious learning enhance solver performance and under what conditions. Design/methodology/approach The model was tested using survey and archival data from 261 solvers on a large contest platform in China. Findings Results support the premise that experiential learning and vicarious learning separately and jointly enhance solver performance. Moreover, perceived task uncertainty strengthens the effect of vicarious learning but weakens the effect of experiential learning, whereas perceived competition uncertainty weakens the effect of vicarious learning. Originality/value The current study enriches the understanding of the impacts of experiential learning and vicarious learning and offers a more nuanced understanding of the conditions under which solvers can reap the performance benefits from learning in crowdsourcing contests. The study also provides practical insights into enhancing solver performance under perceived task uncertainty and perceived competition uncertainty.