SMU SOE Seminar Series (February 26, 2026): Optimal Pricing Schemes in the Presence of Social Learning and Costly Reporting

TOPIC:

ABSTRACT

A monopolist sells a new product of unknown quality to sequentially arriving, short-lived buyers with unit demand. As it's costly for buyers to report their realized utility, the monopolist must offer a bonus to incentivize disclosure. We work within an exponential-bandit framework which allows the arrival of both conclusive good and bad news, and characterize the monopolist's optimal joint policy of dynamic pricing and bonus, which switches among distinct phases. In the special case of no bad conclusive news, the monopolist will set zero bonus and halt learning when public belief is sufficiently high, but doing so is generally never optimal. Moreover, a polarity dependent manipulation--paying bonus only for verifiable good news--can improve social welfare when the arrival rate of conclusive bad news is close to zero.

Keywords: Social Learning, Costly Reporting, Dynamic Pricing.

JEL: D83. C02. C61. C73.

Click here to view the paper.