EURO 2025 Leeds
Abstract Submission

703. Probabilistic Services on a Ride-hailing Platform

Invited abstract in session WA-30: Shared Mobility Optimization I, stream Shared Mobility Optimization.

Wednesday, 8:30-10:00
Room: Maurice Keyworth 1.05

Authors (first author is the speaker)

1. Di Wu
School of Management Science and Engineering, Central University of Finance and Economics

Abstract

Motivated by Didi's simultaneous call service option, we study probabilistic pricing on a ride-hailing platform offering high- and low-type services. Riders selecting this option do not know in advance which service they will receive but are charged based on the service they ultimately get. Using a rational expectation framework, we characterize the optimal probabilistic pricing strategy. Our results show that the introduction of probabilistic pricing allows the platform to segment riders more effectively and better balance the supply and demand, mitigating the congestion level in the market. Interestingly, more demand from the low-type riders can be fulfilled, albeit at a higher price for the low-type service. Moreover, the platform's optimal pricing scheme depends on riders' aversion to congestion and the platform's cost of offering the probabilistic service. Remarkably, providing probabilistic service can be a win-win-win policy for riders, drivers, and the platform when riders are not highly averse to congestion and the computational cost of the probabilistic services is not too high. Our findings provide practitioners with valuable guidelines for designing probabilistic services in ride-hailing markets.

Keywords

Status: accepted


Back to the list of papers