Who Ate the Lunch? A Belief Propagation Algorithm for Identifying Refund Hunters

For online platforms such as UberEats and DoorDash, a central challenge in customer service is the lack of ground truth: when a customer reports that an order was never received, it is difficult for support agents to determine if the driver kept the food, a passer-by took the delivery, or the customer is falsely claiming a missing order. This fundamental uncertainty often results in platforms shouldering refunds and appeasement costs without holding either side of the market accountable. In this work, we propose a belief propagation (BP) algorithm for identifying strategic customers and drivers, considering both their frequency of platform use and the trustworthiness of the participants they had interacted with. In large markets, we show that our proposed BP algorithm approximately recovers the correct types on the driver side, and achieves the highest statistical power on the customer side. Additionally, there is an exponentially large gap (with respect to the number of orders per participant) in the ratio of false positive rates between our algorithm and a naive benchmark that only relies only on the average past refund rate of each participant. Extensive experiments on both synthetic data and data from our industry collaborator — a major Southeast Asian platform — demonstrate that the BP algorithm provides substantial and robust accuracy improvements.