Analog Model

An analog model is a location performance prediction technique that estimates a new site's expected revenue by finding the most similar existing sites in a brand's portfolio—its 'analogs'—and using their actual performance as a benchmark. The underlying assumption is that locations with similar geographic and market characteristics should produce similar results.

Analysts define a set of trade area and site attributes that drive performance—population density, median income, traffic counts, competitor count, store size, and format. Each attribute is measured for both the proposed site and all existing stores. A similarity algorithm (commonly Euclidean distanceEuclidean DistanceEuclidean distance is the straight-line distance between two points in a plane, computed using the Pythagorean theore..., cosine similarity, or k-nearest neighbors in the attribute space) identifies the handful of existing stores most closely matching the proposed site. The predicted revenue is typically a weighted average of the analog stores' actual sales, with weights inversely proportional to dissimilarity. Analysts review the analogs qualitatively to ensure they are contextually reasonable—a suburban drive-through store should not be matched to an urban walk-up location, even if their demographics are similar.

Retailers use analog models as a first-pass screening tool in site selectionSite SelectionSite selection is the analytical process of evaluating and choosing optimal physical locations for new stores, facili..., quickly estimating revenue potential before investing in more complex predictive models. Franchise systems use analogs to set franchisee revenue expectations. Real estate committees reference analog comparisons to contextualize revenue forecasts generated by regression or machine learning models. The analog model's transparency and intuitive logic make it a trusted foundation of retail site selection, especially when combined with more sophisticated quantitative approaches.