Trade Area Models

Buffers

Buffers create trade areas by calculating a specified distance from a store to the area that surrounds it. They assume that all customers falling within the buffer's boundary will shop at the store.
Euclidean Drive time
Variable distance Fixed distance
Distance multiplier

Buffers in depth: A buffer around a store creates an area of a specified distance away from a store. Distance can be calculated using different methods, such as Euclidean (straight-line) distance, or a distance measured in travel time. They are relatively simple to communicate, but can be difficult to understand when there are overlapping trade areas because they are a deterministic model, meaning customers can only shop at one store. Buffer analysis can be made more powerful by incorporating customer demographic variables like income, education level, and age, so that stores can be sure to locate near their target customers. A buffer's size can be based on a retail analyst's gut feeling, customer surveys, or other quantitative methods.
Learn more about buffers

Huff Model

The Huff Model calculates the likelihood of an individual using a store based on the store's square footage and their distance from the store. Don't forget you can resize stores by right clicking and dragging.
Attractiveness exponent: 2

Distance decay exponent: 2

Maximum distance threshold:
50 km

Huff Model in depth: The Huff Model is a form of spatial interaction model that calculates the probability any given customer will choose to shop at a store based on a combination of the store's attractiveness (commonly square footage or product offerings), the distance to that store, and the combined attractiveness of other competing stores. Often, a distance decay exponent is used to better model the influence of distance decay on customer decisions, and the ideal value for this model parameter varies based on the industry. Similarly, the attractiveness value amplifies the store's customer "pull" based on its relative size. These model parameters would typically be validated using empirical data from store surveys, like asking customers their postal codes at checkout. Some versions of the Huff Model include a maximum distance threshold denoting a distance that customers will not consider travelling beyond.

The Huff Model's final output includes the probability that each customer will use each store, based on the customer's location and the stores' attributes (in this web app, customer's locations are represented by their home census tract). The probabilities aim to model customer decision making, and highlight that individuals may shop at more than one store. Store forecasting can be calculated using census income and population data to calculate the expected store revenues by multiplying the probability of each customer (or census tract) to use the store, their average income, and the proportion of their income that is typically spent on goods from the type of store.
Learn more about the Huff Model