Huff Model

The Huff Model, developed by economist David Huff in 1963, is a spatial interaction model that calculates the probability of a consumer at a given location patronizing a specific retail outlet. It is derived from the gravity modelGravity ModelThe gravity model predicts the flow of consumers between origins and destinations based on the attractiveness of a de... but focuses specifically on competitive retail environments where consumers choose among multiple store options. The Huff Model has become one of the most widely used and academically validated tools in retail location analysis.

The model calculates the probability that a consumer at origin i will visit store j as: Pij = (Sj / Tij^b) / Σ(Sk / Tik^b), where Sj is the attractiveness of store j (typically measured by square footage, product assortment, or a composite score), Tij is the travel time from origin i to store j, b is the distance-decay parameter, and the denominator sums attractiveness-to-distance ratios across all competing stores k. The result is a probability value between 0 and 1 for each origin-store pair. Multiplying these probabilities by the population and spending potential at each origin yields estimated sales for each store.

The distance-decay parameter b is critical to model accuracy and varies by retail category. Convenience-oriented categories (grocery, pharmacy) typically have high b values (consumers are very sensitive to distance), while destination categories (furniture, specialty retail) have lower values. Analysts calibrate b using observed customer data—such as loyalty card addresses or mobile device visit patterns—to ensure the model reflects actual behavior in the study market.

The Huff Model is used to delineate probabilistic trade areas (showing the probability surface around each store rather than hard boundaries), forecast sales for proposed new locations, estimate the impact of competitor openings on existing stores, and model the cannibalization effects of adding outlets to an existing network. Site selectionSite SelectionSite selection is the analytical process of evaluating and choosing optimal physical locations for new stores, facili... analysts use it to compare scenario outcomes—for example, testing whether a 40,000 sq ft store at Site A will outperform a 25,000 sq ft store at Site B.

The Huff Model accounts for competition explicitly, producing more realistic trade areas than simple distance rings or drive-time zones. Its probabilistic output acknowledges that consumers do not belong exclusively to one trade area—they split their visits among alternatives based on relative attractiveness.

The model assumes that consumers evaluate all alternatives objectively and that attractiveness can be adequately captured by a small number of variables. It does not inherently account for brand loyalty, pricing differences, or qualitative factors like store ambiance. Extensions and hybrid models address some of these limitations by incorporating additional variables and behavioral data. The Huff Model endures because it provides an elegant, practical solution to the core question of retail location analysis: where will customers shop, and how much will they spend? Its probabilistic framework and competitive structure make it an indispensable tool for site selectionSite SelectionSite selection is the analytical process of evaluating and choosing optimal physical locations for new stores, facili... and network planning.