Sourcing for Online Marketplaces with Demand and Price Uncertainty

Abstract

Our paper is motivated by a manufacturer that sells a single seasonal product through multiple retailers competing on an online marketplace, such as Amazon marketplace. Selling price uncertainty and evolution of forecasts of price and demand are key features of the online marketplace. Sourcing choices are differentiated by cost and available lead times -- delaying to a shorter lead time is more expensive, but yield more accurate information about future selling price and demand. Thus, ahead of the season, each retailer faces a continuous-time decision problem about when to place an order with the manufacturer and in what quantity. The manufacturer is interested in knowing the ordering pattern of the retailers in order to plan its production schedule. We consider two sourcing strategies varying in the flexibility of order timing: an optimal pre-committed ordering time strategy and an optimal time-flexible ordering strategy. Under the latter strategy, the optimal order time follows a double threshold policy in the selling price variable: it is optimal to order if the price is within a certain band, and wait otherwise. By characterizing both strategies, we identify scenarios when time-flexible ordering can be beneficial for the retailer, the manufacturer, and the supply chain as a whole. The predictions of our model are consistent with the experience of a large U.S. manufacturer that motivated our study.

Keywords: Supply chain management, Stochastic inventory theory, Optimal stopping.