Inventory Auditing and Replenishment Using Point-of-Sales Data

Spoilage, expiration, damage due to employee/customer handling, employee theft, and customer shoplifting usually are not reflected in inventory records. As a result, records often report phantom inventory, that is, units of good not available for sale. We derive an optimal polynomial-time algorithm for the auditing and replenishment problem with phantom inventory, under the assumption that inventory is known immediately after replenishment. Moreover, we numerically show that a service-level threshold policy performs comparably to the optimal.