Population density and shop size: matching the format to the neighbourhood

Population density decides how far your customers will come, and that decides how big a shop the area can carry. Where many people live on every square kilometre, a small shop within a short walk reaches enough of them, and rent per square metre is high, so small formats win. Where homes are spread out, people drive, a store must draw from a wide area, and a large format with parking makes sense. Density alone does not pick the site, but it tells you which format to look for in which part of a city.

This article is for retailers and service businesses that run more than one format: a convenience store and a supermarket, a kiosk and a flagship, a studio and a full club. It covers how density is measured, how to turn it into a format choice, a worked example in Lyon, the traps, and an AI employee that does the sums for every candidate.

How density is measured

The Demographics block returns people_per_km2 for any circle, area or hexagon: the residents divided by the land it covers. On a heat map of hexagons, each one carries its own density, so you can see where a city thins out. Population figures exist only in some countries, among them France, Canada, India, Thailand, Egypt and Australia; not the United Kingdom or the United States.

Density is a ratio, and ratios depend on the shape you draw. A 5 km circle around a riverside centre includes water, rail yards and parks, which pull the average down even if every street is packed. For format decisions, measure density over the area a customer would actually come from, not over a big round circle.

From density to format

The useful question is not "how dense is this area?" but "how many people live within the distance this format's customers will travel?" Each format has its own distance:

• A small convenience shop or a kiosk: a walk of five to ten minutes. Measure residents inside a Travel area on foot.
• A mid-sized supermarket: a short drive or a long walk, perhaps ten minutes by car in a suburb.
• A large store with parking: a drive of fifteen minutes or more. Travel area reaches up to 15 minutes by road today, so measure that and note that the real draw may be wider.
• A service people book, such as a gym or a clinic: somewhere between, depending on how often they come.

Then compare the residents in that reach with what your own working stores of the same format have in theirs. If your successful small shops each have 8,000 to 12,000 people within a ten-minute walk, a candidate with 3,000 is weak for that format, however dense it looks on the map. It might still suit a different format.

What the store counts hint at

Real counts show how much format varies from city to city. xMap's places database lists 1,332 supermarkets within 10 km of the centre of London, 190 within 10 km of the centre of Sydney and 82 within 10 km of the centre of Toronto (November 2026).

1,332Supermarkets in London

190Supermarkets in Sydney

82Supermarkets in Toronto

These numbers do not prove that density causes the difference. Each city also names its shops differently, and many small grocers in one city may be listed under another type in the next. But they are a warning against carrying one format from city to city: a centre where supermarkets sit a few streets apart is a small-format market, and a large store planned there would be fighting for space and for customers who walk.

A worked example in Lyon

A made-up case. A grocery chain with a small city format and a large suburban format is looking at three units in and around Lyon. For each it measures residents inside a 10-minute walk and a 10-minute drive. The figures are invented, to show the method.

Made-up figures, for illustration only.

The old-town unit has plenty of people on foot but would be hard to reach by car, so it is a small-format site. The retail park has almost no one within walking distance, so a small shop there would fail, but the drive-time area is large enough for the big format. The inner suburb sits between and could take either; the chain would then count the supermarkets already inside each reach with Nearby places before choosing.

Mistakes to avoid

• Averaging density over a circle that crosses a river, an airport or a port, and reading the low number as a quiet market.
• Treating density as spending. A dense district of students and a dense district of families buy very different baskets; read the age groups too.
• Forgetting daytime people. A business district has few residents and many workers, so a small format can thrive there on footfall that density does not show.
• Comparing a walking area in one city with a driving area in another.
• Assuming a format that works in one country works the same way in the next.

Let an AI employee size each candidate

1. My data holds the candidate units with their positions.
2. Travel area draws the reach for the format you are testing, for example a 10-minute walk.
3. Demographics counts the residents and the density inside each reach.
4. Nearby places counts the stores of your kind already inside it.
5. A Calculated field works out residents per existing store, and Sort and limit puts the best first.
6. Email sends the ranked list. Duplicate the flow with a 10-minute drive for your large format.

The inbox receives one list per format, each unit with its residents, density and competitors, so the property team sees at a glance which units to view for which format. For gaps rather than candidates, start from Neighbourhoods without a grocery store; for delivery formats, Where to put the next dark store.