Optimizing Driver Incentives for Ride-Sourcing Services
Optimizing Driver Incentives for Ride-Sourcing Services (Prof. Xu)
Ride-sourcing platforms such as Uber and Lyft implement various types of driver incentives, but a typical scheme can be generalized into the following target-based format: “A driver will receive a certain amount of monetary reward from the platform if she finishes a given number of orders within a given period”. Such incentive schemes aim to prolong drivers’ labor hours with generous reward setups. Platforms allocate millions and billions of funds to sustain the incentive program. However, as a long-time practice of the platform, these target-based incentives were designed through heuristic techniques via trial-and-error. In this project funded by Didi Chuxing, one of the largest global competitors in the ride-sourcing market, our research team from George Washington University and the University of Michigan aims to evaluate the effectiveness of the schemes currently in place, and to develop optimization models to seek methodological and design improvements.