The challenges of scanner data

Data capture at point of sale, using laser scanning technology, is now commonplace in retail outlets in developed economies. The volume of data collected is enormous (a single U.K. grocery superstore carries over 15,000 different lines, each with a unique 13 digit barcode and will sell 600,000 items per week). It has been recognized that, where point of sale data can be translated into meaningful information, there is scope for efficiency savings and exploitable competitive advantage. In U.K. grocery retailing the uses of scanner data have concentrated on operational aspects such as improved ordering systems and staff scheduling. There have, however, been many recent studies, predominantly in the U.S.A., which have examined, using scanner data, the relationship between marketing variables (price, display, advertising, coupons, and other promotions) and customer purchasing behaviour. This paper presents an overview of these studies and identifies the opportunities for operational researchers to further the understanding of such relationships through modelling and data handling.