Improved Approximations for Posted Price Mechanisms, Second Price Auctions and Free-Order Prophets

We study revenue maximization through sequential posted price mechanisms (SPMs) in single-dimensional settings with n buyers, and independent but not necessarily identical value distributions. We construct our SPMs by considering two simple pricing schemes: one that imitates Myerson's mechanism via taxation principle, and the other that posts a uniform price. We design a factor revealing LP that crisply captures the approximation factor of this SPM. In the H-units setting, our SPM yields the first improvement in approximation over the correlation-gap-based approximation factor from over eight years ago (Yan 2011). In the position auction setting, our SPM yields the first higher-than 1-1/e approximation factor. In the 1-unit setting, our SPM yields better than the current best known approximation factor for small values of the number of buyers n. Our results on SPMs immediately imply improved performance guarantees for the equivalent free-order prophet inequality problem. In the recent surge of results that show an improvement over 1-1/e for SPMs, ours is the only technique that generalizes beyond the 1-unit setting, and yields the first improvement over long established approximation factors in several settings.