Abstract
This paper considers the situation of a retail store subject to inventory inaccuracies stemming from execution problems. We assume that inventory inaccuracies are introduced by misplacement type errors that occur within the store, i.e. the whole quantity of products that is ordered and received from the supplier is not available on shelf to satisfy consumers’ demand either because the replenishment process from the backroom to shelves is prone to errors (e.g. products are lost during this transfer, products are forbidden in the backroom, products are put on other shelves than where they should be, etc.) or products are misplaced on other shelves by consumers during their visit to the store. We consider a Newsvendor model that captures this issue in a simple way: For a given quantity of products ordered from the supplier, only a random fraction is available for sales. We compare three approaches. In the first approach, the retailer is unaware of errors in the store. In the second approach, the retailer is aware of errors and optimizes its operations by taking into account this issue. The third approach deals with the case where the retailer deploys an advanced automatic identification technology (such as the Auto-ID system based on the Radio Frequency Identification (RFID) technology) to eliminate errors. In particular, we provide insights on the relative benefit of implementing the RFID technology (moving from approaches 2 to 3) compared to the benefit of optimizing the system in presence of inaccuracies (moving from approaches 1 to 2). We also provide an analytical expression of the cost of the RFID tag which makes its deployment cost effective.