Glossary

Market basket analysis

Finding which products are frequently bought together, to inform bundling, placement, and recommendations.

Market basket analysis examines transaction data to find which products tend to be purchased together, revealing associations that aren't obvious from looking at any single product's sales in isolation — the retail example most often cited is discovering that two seemingly unrelated items sell together more than chance would predict.

The standard technique uses association rules, scored with three measures: support (how often the item combination appears across all transactions), confidence (how often the second item is bought given the first was), and lift (how much more often the pair occurs together than would be expected if the two items were independent — a lift above 1 signals a real association). This differs from product affinity, a broader term for any measure of which products relate to each other, which may come from browsing behavior or content similarity rather than co-purchase data specifically.

Market basket analysis matters for store layout, bundle pricing, and the "customers also bought" logic behind many recommendation systems. The main pitfalls are mistaking a strong association for a causal relationship (two items may sell together because of a shared promotion, not an inherent affinity), and generating rules from too little data, which produces associations that look statistically strong but don't hold up out of sample.

Last reviewed September 22, 2026

In the index now

Related terms

Product affinity A measure of how strongly two or more products relate to each other in customer interest or purchase behavior. RFM analysis A segmentation technique that scores customers on how recently, how often and how much they buy. Clustering An unsupervised technique that groups data points so items within a group are more similar to each other than to items outside it. Recommendation system A system that predicts and ranks the items a specific user is most likely to want, such as products, content, or actions.