Tolga Aydinliyim, Baruch College, CUNY
Michael S. Pangburn, University of Oregon
Elliot Rabinovich, Arizona State University
Min Choi, Arizona State University
Unlike in a traditional store environment where inventory is directly visible to customers, Internet retailers have more flexibility in exposing consumers to inventory information. It is common for online sellers to provide only binary information to shoppers, indicating a product is either “in stock” or not. In this paper, we investigate whether an online retailer which optimally prices its product should divulge (precisely) its stocking level, or merely indicate whether the product is in stock. By appropriately designing an inventory information disclosure policy, an online retailer may potentially leverage consumers’ heterogeneous sensitivity to stock-out risk and thus enhance profits.
We employ a two period analytic model to address an online retailer’s contrasting inventory disclosure policy options. In the second period we assume the firm attempts to clear out any remaining stock at cost and therefore the firm must earn its profits during the primary (first) selling period. We permit two types of consumers: one segment that rationally infers availability when the firm does not divulge its stock level, and the others who rely on some prior belief distribution regarding inventory. If the firm’s stock level is either shown or believed to be sufficiently high, then consumers may strategically defer their purchase to the second period, however at the risk of potentially not obtaining the product. Intrinsic to the model is the assumption that consumers’ purchase decisions are sensitive to stocking quantities, which we support using data collected from an online book sales website (StrandBooks.com) that divulges stock levels to its customers.
Using the analytic model, we derive the threshold stocking level at which the inventory sharing and masking policies and the corresponding optimal prices yield equivalent expected profits. Below that level, sharing is optimal; above that level, masking is optimal. We also show that, if the inventory decision is endogenous then the retailer should optimally set and disclose a low stock level to ensure a first period sell-out. This structure suggests to managers the potential for increasing profits by tailoring disclosure tactics to a product’s stocking level. To explore this issue further, we collected data from Amazon.com, which provides a richer (real-time, multi-product) context for studying the impact of inventory information on sales. Amazon follows a hybrid disclosure policy that masks inventory information at high stock levels but divulges that information at low stock levels. We show that the sales rate under disclosure tends to be higher than under masking, thus demonstrating that Amazon is, at least in many cases, able to enhance sales rates by selectively opting to disclose detailed inventory information only when inventories are sufficiently low.
Aydinliyim, T.; Pangburn, M.; Rabinovich, E.; Choi, M*. “Is the Cue ‘In Stock’ Always Effective? Inventory Information Disclosure Tactics to Leverage Stockout Risk.” Working Paper