Our study examines the effect of virtual teams on worker productivity and retention on an online platform. Hailed as the future of work, the gig economy provides flexible, low-barrier jobs for millions of workers globally. However, a lack of both organization identity and social bonds contributes to the high attrition rate experienced by gig platforms. To test the impact of virtual teams, we use a large-scale natural field experiment with 27,790 drivers on a global ride-sharing platform to organize drivers into teams that are randomly assigned to one of three experimental conditions. Treated drivers receive either their team or individual ranking, whereas those in the control condition receive individual performance information without social comparison. We find that treated drivers are significantly more productive than those in the control condition. We further find that drivers in the team leaderboard treatment continue to work longer hours on the platform three months after the end of the experiment. Lastly, we find that those identified as laggards within a virtual team benefit the most from a team contest.

Chair: Subhasish Chowdhury, Co-Chair: Kai A. Konrad