My approach (N.B. it is based on my experience in Europe)
First, define the scope. Supermarkets in which country/region/city?
There are 3 main types of supermarket:
- small convenience stores
- medium-sized supermarkets in the city
- large-sized supermarkets in big malls (out of the city)
The convenience stores that I've been to were approximately x by y metres. On average the surface is "A"
The medium-sized supermarkets... and the average surface is "B"
The large-sized supermarkets... and the average surface is "C"
Assuming that the sample that I have is representative of the supermarket, I need to calculate a weighted average of A, B and C. In order to do so it is essential to know the scope of the analysis. I would make some assumptions about the density of each type of supermarket in the area, by distinguishing urban area (population), suburban area (population), countryside.
I can see two main drivers: surface (how big is the city?), population density (what is the number of inhabitants per sq km?) and preferences (do people in suburbs actually buy in small, expensive convenience stores?). You can use your empirical experience or data you know and show your business culture/business sense in this way.