This dissertation is comprised of two essays at the intersection of Empirical Industrial Organization and Behavioral Economics. They explore how business decisions are more richly explained with additions in psychological insights commonly found among consumers. These behavioral biases can affect a firm's decision to participate and their supply in the market, ultimately impacting market competition, supply thickness, and equilibrium. In Chapter 1, I explore how retailers make entry and exit decisions in the context of an online marketplace. Using a rich panel of internal data from eBay on dedicated sellers, I analyze a feature of the platform requiring sellers to select among monthly contracts that differ in the listing fee schedules. I further exploit a regime change that introduced a monthly allowance of free listings of inventory, altering all contracts from a two-part tariff to a three-part tariff design. This design change was effective in attracting new users and sellers to the platform, encouraging experienced users to become high-volume sellers, and increasing total inventory listed in the marketplace. This is despite little changes in average costs of selling. I demonstrate that standard entry and exit models cannot explain this increase in supply and competition. Instead, I propose loss aversion as an additional factor impacting participation. Chapter 2 investigates how high-volume, experienced retailers value their products and make supply decisions. Using similar data from eBay and exploiting the same contract feature and policy change, I analyze both the sellers' contract choice decisions and the timing of product listings. By estimating a dynamic model of plan choice and listing decisions, I find that sellers have a limited learning period and hold biased beliefs on the option values of their products. Furthermore, they respond heterogeneously to dynamic incentives and future listing costs, leading to an uneven supply of products on the platform. Using the model estimates, I show that debiasing sellers would increase seller surplus by 6\% but decrease platform listing revenue by 26%. However, since many platforms generate the majority of revenue from percentage fees on the listings' values, they may prefer to pursue policies that increase total listings. By targeting specific sellers, such policies can increase aggregate listings on the platform by up to 4.8% at a loss of 5.5% in listing revenue.