Information Design and Pricing of Supply Competition in Lead-Selling Platforms

In this paper, we analyze a type of platform common in the service sector, where customers can freely post tasks, and service providers pay a fee to purchase job leads and unlock the ability to communicate with these customers. We adopt a Bayesian persuasion framework to explore how the platform can leverage its informational advantage about the supplier base and pricing to strategically influence the decisions of service providers, aiming to maximize its revenue. We begin the analysis by assuming that prices are exogenously given and characterize the structure of the optimal signaling mechanism. Specifically, the platform recommends all service providers to participate when the pool size is below a certain threshold. When the pool size exceeds this threshold, the platform randomizes its recommendations between two consecutive numbers, independent of the pool size. This behavior is driven by a unique feature of our model, where the beliefs of the platform and suppliers are non-identical: a supplier's knowledge of their own participation inflates their belief about the total number of service providers. Next, we endogenize the platform's pricing decision and consider the policy space where the signaling mechanism and state-dependent pricing are jointly optimized. Notably, within this space, we find that the optimal signaling mechanism with a single price remains optimal. This result implies the superiority of the optimal signaling mechanism over the commonly adopted contingent pricing policy, where the price charged by the platform depends on the number of interested suppliers. We show that the performance discrepancy between these two policies is particularly pronounced when supply competition is intense.