Inventory Policy Implications of On-Line Customer Purchase Behavior

In this paper we will examine some implications of online data for a classical operations management model, vis. the Economic Order Quantity model. Customer waiting behavior on individual orders (which occur during stockouts) forms the basis for evaluating the potential backorders. The potential attraction of reducing inventory holding costs must be balanced with the loss due to lost sales. We clearly delineate the conditions under which it is profitable to stock out every ordering cycle, and the conditions under which the traditional economic order quantity model still holds. In order to allow practical application of the model, we develop a number of different approaches to the problem of estimating the backorder function from available on-line transaction data.