« Back to Results
Market Design for Online Platforms
Friday, Jan. 4, 2019
2:30 PM - 4:30 PM
Economic Science Association
University of Michigan
Toward an Understanding of the Economics of Apologies: Evidence from a Large-scale Natural Field Experiment
We conduct a nationwide field experiment involving 1.6 million ride-sharing consumers and combine it with theory to deepen our understanding of the economics of apologies. First, apologies are not a panacea: the efficacy of an apology -- and whether it may backfire -- depends on how and when the apology is made. In some cases sending an apology is worse than sending nothing at all. For firms, caveat venditor should be the rule when considering apologies. Second, money consistently speaks louder than words. We find consistent evidence that the best form of apology is to include a coupon for a future trip.
The Design of Feedback Revision Rules - An Experimental Study
Online platforms often employ feedback systems and allow to revise submitted feedback under certain conditions. We experimentally study potential strategic problems in the feedback revision process and possible improvements. We document the negative effects on informativeness and market-enhancingness of feedback when feedback revision rules allow for strategic behavior. We propose small improvements to the feedback revision process that aim to address these problems. Our laboratory experiment indicates that the improved system mitigates and partly reverses the negative effects of wrongly designed revision options.
Team Competition and Driver Productivity in Ride-sharing: A Natural Field Experiment at Didi
While the sharing economy benefits workers with autonomy and flexibility, the lack of a strong identity from work and bonds with colleagues can lead to low engagement with the platform. To address this issue, we run a field experiment to create teams for drivers at Didi. We vary the recommender algorithm and find that drivers are most likely to join a team when the recommendation is based on hometown similarity. Once teams are formed, they enter into a week-long contest. We find that both randomly-formed teams and teams of age similarity have the highest productivity increase compared to the placebo.
University of Pittsburgh
University of Rochester
D4 - Market Structure, Pricing, and Design