April 2008
Journal Article
Management Science

Inventory Record Inaccuracy: An Empirical Analysis

By: Nicole DeHoratius and Ananth Raman

Format:Print

Abstract

This study explores the systematic variation in inventory record inaccuracy (IRI) observed both within and across stores. Traditional inventory models, with a few exceptions, do not account for the existence of IRI and those that do treat record inaccuracy as random. Examining nearly 370,000 inventory records from 37 stores of one retailer, we found 65% to be inaccurate. That is, the recorded inventory quantity of an item fails to match the quantity found in the store. We identify factors associated with this inaccuracy that are stock keeping unit- (SKU) and store-specific. SKU-specific factors such as item cost, selling quantity, and method of distribution account for the observed variation in IRI within stores. Store-specific factors such as the density and variety of inventory observed at each store account for the variation in IRI across stores.