Service Differentiation through Priority Matching by Ride-Sharing Platforms

Ride-sharing platforms that match riders with drivers have become one of the most celebrated examples of disruptions caused by information technology (IT). A rider’s willingness-to-pay (WTP) for the platform’s service is affected by the expected waiting time, which in turn depends on the match made by the platform (“match quality”). While some riders may have low sensitivity to match quality (“low segment”), others may be highly sensitive (“high segment”). Without knowing an individual rider’s sensitivity, ride-sharing platforms often use priority matching to offer differentiated services, allowing riders to self-select the service that best suits them. While such a strategy is akin to quality-differentiated products in conventional markets, the ride-sharing context exhibits unique features. For instance, one-to-one matching of riders and drivers imposes constraints on the match qualities the platform can offer. Moreover, any differentiation in driver wage that accompanies service differentiation on the rider side can induce drivers to anticipate a possibly higher future wage (”forward-looking behavior”), which can lead to supply-side cannibalization analogous to well-known demand-side cannibalization resulting from rider self-selection. As a consequence of these features, service differentiation generates novel implications for ride-sharing platforms. We find that a platform would enhance the match quality intended for the high segment and degrade the match quality intended for the low segment to mitigate demand-side cannibalization, whereas a conventional firm would degrade only the quality intended for the low segment. If drivers are not forward-looking, the platform would also differentiate wages when it offers differentiated services. However, if forward-looking drivers wait to obtain the higher wage, the platform may find it optimal to offer a uniform wage to drivers even though it may offer differentiated services to riders. Moreover, the platform could degrade the match quality intended for the high segment and enhance the match quality intended for the low segment to mitigate supply-side cannibalization. We further find that the demand-side and supply-side cannibalization have qualitatively similar same-side effects, but they have asymmetric cross-side effects.