Two-stage Pandora ’ s Box for Product Ranking

In online platforms, consumers face an abundance of options that are displayed in the form of a position ranking. Only products placed in the first few positions are readily accessible to the consumer, and she needs to exert effort to access more options. We study how platforms with different business models should rank products to maximize their profit. We provide empirical evidence from a large online platform that products placed in higher positions have a favorable chance of being selected regardless of their utility. Further, the externality that high-positioned products impose on low-positioned ones substantially increases with their utility. Motivated by our empirical evidence, and building upon the seminal work of Weitzman (1979) (also known as the Pandora’s box), we develop a two-stage sequential search model; in the first stage, the consumer learns the intrinsic utility of products, already known to the platform, and forms a consideration set. While in the second stage, she inspects the products in her consideration set to learn the additional idiosyncratic utility she derives from a product. We develop optimal or FPTAS ranking algorithms under two diametric business models: one only concerned with maximizing consumer welfare a suitable objective for platforms seeking consumer’s long-term engagement and one concerned with maximizing short-term revenue. Further, somewhat surprisingly, we show that ranking products in a decreasing order of their intrinsic utilities does not necessarily maximize consumer welfare. Such ranking may shorten the consumer’s consideration set, due to the externality effect, leading to insufficient exploration.