Using Market Basket Analysis to Estimate Potential Revenue Increases for a Small University Bookstore

Market basket analysis (MBA) is a widely used technique for identifying affinities among items that customers purchase together. MBA metrics are support, confidence, and lift. We show that support and confidence may include misleading information about the nature of the affinity, and that lift is the most useful metric. Starting with the MBA, we use the product affinities to predict ways to increase revenues, and we estimate the magnitude of the possible increases as a function of customer price sensitivity and affinity saturation level. We also point out limitations of the MBA and suggest ways to overcome them. For the case of a small university bookstore, we identify pairings of items that have revenue-increasing potential. Depending on the customers’ price sensitivity and affinity saturation level, revenues could be increased by as much as $10,000 or as little as $100 for a $ 410,000 starting level. In particular, we identify pairings where customer price sensitivity might be overcome (one-time situations, for example graduation-related purchases). This case study is the first to provide an actual magnitude of the estimate of potential revenue increases.