You live the same distance away from two cities and decide to go shopping – which do you go to; the one with the most stores, right? What if you didn’t live the same distance away from each city – how far would you

have to be away from the larger one, in order to decide that it’s easier just to go to the smaller, closer one?

That question is the crux of William J. Reilly’s 1931 Law of Retail Gravitation. He asserted that:

the Break Point (BP) [from city p₂] is equal to the Distance (d) between two places, divided by the following: 1 plus the Square Root of, the size of Place One (p₁) divided by the size of Place Two (p₂). (slightly paraphrased from the previous Wikipedia link)

The “size” of a place is a little arbitrary, but you could use figures such as the total square-footage of retail space, or the total number of unique stores. Either way, this data is fairly difficult to track down.

Luckily, ubiquitous stores like Starbucks are a good indicator of the larger retail environment, so we can use their UK Store Locator to count the total number of Starbucks stores in each city, and use this as a good estimate for ‘Retail Size’.

Obviously the Break Point figure, in miles, will be larger for the cities that are further away from London, such as the Scottish cities. Therefore a more interesting statistic is the Break Point distance in terms of percentage of the total distance (the green line in the graph above). This demonstrates the relative ‘pull’ of each city, influenced by the number of Starbucks (which, remember, we’re using to represent the total number of stores). Looking at these figures, we can see how the lack of Starbucks in Liverpool is causing a relatively low ‘retail pull’ towards the city, whereas the additional Starbucks stores in Manchester give it almost the same relative ‘pull’ as the much farther away Edinburgh.

Just because I had the data, here’s a bonus graph of People per Starbucks for each city:

Looks like Liverpool and Birmingham need a few more stores.