CompetitorsInsightsSite SelectionStrategy

What Isaac Newton Knew About Your Trade Area

By September 17, 2015 November 7th, 2017 No Comments

Trade area analysis has changed dramatically in recent years and continues to evolve. The analysis we undertake on behalf of our clients was not possible only a few years ago, and what we do today will be primitive by the standards of the future. In the midst of this upheaval, I think it will be an interesting journey to step back in time and see how trade area analysis has evolved.Apples_from_Red_Apple_Farm,_at_the_Lexington_Farmer's_Market,_Lexington_MA

One early and popular method for trade area analysis was the gravity model. First put forth by William Reilly in 1931, the gravity model was developed as a heuristic to study consumer behavior. The principle underlying all the gravity models is that trade areas have different levels of attractiveness to consumers, and this attractiveness will draw consumers to them. The level of attractiveness for an area is inversely proportional to the distance a consumer must travel and is usually based on the number or variety of retail offerings in an area, but can also be based on population when comparing trade between cities.

There are a few different models (here, here, and here) we could use, but the intuition behind them is the same and that is what we are trying to get at here. I am borrowing this example of the Huff Model from Principles of Retailing, a comprehensive text on retail management.

You are a consumer with two shopping options as show below:

Shopping Center        Square Footage       Distance(miles)
A                                      20,000                                 3
B                                      40,000                                 4

The attraction for each store can be calculated as follows:

Attraction of Center A = 20,000/(3^2) = 2,222
Attraction of Center B = 40,000/(4^2) = 2,500

The probability that you will visit either center is the center’s attractiveness expressed as a percentage of the total attraction you face, that is:

Probability you visit Center A = 2,222/(2,222 + 2500) = .47 = 47%
Probability you visit Center B = 2,500/(2,222 + 2,500) = .53 = 53%

We find that even though store B is a longer trip, you are more likely to go there because the bigger center, with its larger selection of offerings, makes for a more attractive destination.

In the mid-twentieth century, cities were very centralized, so this kind of analysis did a fair job describing the behavior of consumers. Suburban sprawl and the rise of shopping malls have complicated the process by spreading out hundreds of shopping destinations across metropolitan regions. Consumers also increasingly shop near where they work so models based on where consumers live my give you very misleading results. Finally how we measure an areas’ attraction is important to our results, but highly subjective. Choosing the wrong metric will render the model useless.

Thankfully many of these problems are solved by the next tool of the trade we will discuss…. Regression.

Marc Smookler

About Marc Smookler

Marc Smookler has founded 6 companies—2 of which have been acquired and 3 of which are market leaders in their respective spaces—the leading brick-and-mortar retail analytics company (IdealSpot.com), a leading online retailer (SakeSocial.com), and a cutting-edge marketing services platform (Written.com). Marc’s companies have generated over $300M in lifetime revenues and sold over 150,000 products worldwide.

Leave a Reply

Your on-demand data and mapping platform

With our easy-to-use, all-in-one platform, anyone can get actionable location intelligence within 5 minutes of signing up.

Request a demo