EXPRESS: Pricing and Information Provision in Service Platforms with Heterogeneous Customers

Abstract

Service platforms frequently employ virtual queues where customers have differing sensitivities to wait times and cannot see their precise positions in line. In these settings, providers often disclose waiting-time information to influence customers' decisions on whether to join or balk. A natural benchmark is to report the long-term average waiting time, yet whether this deterministic disclosure policy is optimal remains unclear when both pricing and customer participation are endogenous. This study develops a stylized theoretical framework to explore the joint structure of pricing and waiting-time information disclosure for service providers managing diverse customers in a virtual queue. The provider decides on both the service price and the waiting-time disclosure policy within a rational-expectations framework. Our findings reveal that the optimal disclosure strategy is captured by scaled Rademacher distributions, meaning at most two distinct announcement values are required. This reduces a complex infinite-dimensional problem to a more manageable finite-dimensional optimization task. Leveraging this framework, we derive the optimal strategies for pricing and disclosure, identifying scenarios where variability in announced waiting times is beneficial. Our analysis shows that noise is advantageous when customers can handle uncertain wait-time signals or initially perceive the service as quick. The optimal level of noise correlates with system congestion, offering the greatest benefits when there is a notable disparity between perceived and actual congestion levels. This provides significant advantages when disclosing waiting times with noise. These findings offer actionable insights for service platforms, guiding them in creating cohesive strategies that encompass pricing, information disclosure, and service configuration, ultimately enhancing customer experience and operational efficiency.