This talk explores simple menus in robust screening. In the first part, we address a robust selling problem where a seller, uncertain about the buyer’s valuation distribution, aims to sell one item. While robust screening offers stronger guarantees than deterministic pricing, it often requires complex menus with infinite options. Our research introduces simple mechanisms with finite menus, balancing performance and implementation simplicity. We propose a tractable framework for various ambiguity sets, including support, mean, and quantile, and derive optimal mechanisms and performance ratios for different menu sizes. We show a modest menu size can deliver similar benefits to those achieved by the optimal infinite-menu mechanism. Notably, even a two-option menu significantly improves the performance ratio over deterministic pricing.

The second part discusses multi-item mechanism design, where characterizing the optimal mechanism is challenging, even when the seller knows the buyer’s valuation distributions. A key result in the literature shows that separable mechanisms (where items are sold independently) are robustly optimal when only marginal distributions are known. However, separate selling may fail to capture a substantial portion of the optimal revenue. To improve performance while retaining simplicity, we introduce “semi-separable mechanisms,” where allocation and payment rules for each item depend solely on that item’s valuation, but leverage joint distributional information. We prove that semi-separable mechanisms achieve the optimal performance ratio among all incentive-compatible and individually rational mechanisms when only marginal support information is available. Our framework also extends to scenarios where sellers have information on the aggregate valuations of product bundles.