A Heuristic Algorithm for Estimating Retail Market Size Using POS Data

Abstract

This study considers the problem of estimating the market size for individual items in the retail grocery industry by sampling the point-of-sales (POS) data from a small number of stores. Using publicly-accessible data provided by each chain and the information from the retailers' financial reports (e.g., number of stores and total revenue), the sampled POS data can be translated into market size information for each item in each area. This research culminated in the development of an algorithm called the POS Market Estimation Heuristic Algorithm (PMEHA), which uses heuristic rules and sampled POS data to estimate overall market information for each area. This algorithm was employed in a prototype designed to calculate market size estimations. Through use of a simple cross analysis procedure, the results for different types of retailers-such as convenience stores, supermarkets, and hypermarkets-can be combined, thus making PMEHA useful for many different retail channels. The algorithm was successfully integrated into a new Retail Sales Information Sharing System (RSISS), and an innovative business model was created to insure that the RSISS could be run profitably once the system had been transferred from a non-profit context into the private sector.