Site selection for a supermarket: catchment, format and the trucks at dawn

Choose a supermarket site by first deciding the format, then counting the households that can reach it in the trip that format relies on: a 5 to 10-minute walk for a small city store, a 10 to 15-minute drive for a large one with a car park. Count the grocers already inside that trip by tier (discount, mainstream, premium), and confirm that a delivery truck can reach the loading bay every morning. A supermarket lives on weekly repeat visits, so the trip has to feel easy every single week.

This article is for grocery chains, independent grocers, franchisees of a grocery brand and property teams who lease to them. It covers how formats change the catchment, what three cities' figures say about chain grip, a worked example in Toronto, the usual mistakes, and an AI employee that screens sites.

Format first, then the catchment

A rule of thumb, not a measurement.

Draw the catchment as a travel time along the roads with Travel area, by car or on foot, up to 15 minutes or 10 km for now. A circle would count homes on the far side of a river or a motorway that would never cross it to buy milk.

How much of the market the chains hold

The grip of the big chains decides whether a newcomer competes with a few large operators or with many small shops. xMap's places database, November 2026:

• London: 1,332 supermarkets within 10 km of the centre. Of the 437 nearest the centre, 296 belong to a chain and 141 are independent; the biggest chains there are Tesco (110) and Sainsbury's (101). The average rating of those 437 is 3.6.
• Sydney: 190 supermarkets within 10 km, of which 75 belong to a chain and 115 are independent; the biggest are Coles (34) and Woolworths (24). Average rating 3.9.
• Toronto: 82 supermarkets within 10 km, of which 28 belong to a chain and 54 are independent; FreshCo (10), Longo's (8) and Sobeys (7) lead. Average rating 4.1.

Two things follow. In central London a new store is almost always next to a chain; it has to win on format, range or hours. And in all three cities supermarkets are rated lower on average than the bakeries, gyms or dentists counted the same way, so a store that is clean, stocked and fast at the till can stand out quickly on reviews.

Reading competitors by tier

Not every supermarket competes with yours. A discount store, a mainstream chain and a premium grocer draw different shoppers from the same street. List the stores in each catchment with Nearby places, then sort them into tiers. A Place classifier or xMap LLM step can do the sorting from each store's name and category, and you correct the few it gets wrong. A catchment full of discount stores can still be open to a premium grocer, and the reverse.

A worked example in Toronto

A made-up case. A mid-size grocer with a strong fresh range is choosing between two leases in Toronto. The figures below are invented, to show the reading.

Made-up figures, for illustration only.

Midtown has more people, but six mainstream rivals and a loading lane where a truck blocks the road each morning. Lakeshore has fewer people today, only one direct rival, two towers coming and a clear bay. The grocer chooses Lakeshore and plans the opening range around the families moving into the new towers.

Mistakes to avoid

• Counting people the format cannot reach. A city store's customers walk. Do not count everyone in a 15-minute drive.
• Treating every supermarket as equal. Compare with stores in your own tier first.
• Forgetting deliveries. A store that can only take trucks after 10:00 will run out of fresh food before the evening rush.
• Ignoring what is coming. New housing can change a catchment within two years. A gap today may be filled before you open, or a quiet site may fill up with families.

Doing it with an AI employee

An AI employee can screen candidate leases and keep watching the market around the one you choose. The blocks, in order:

1. My data: the candidate sites.
2. Travel area: 10 minutes by car, or 10 minutes walking for a city format.
3. Demographics: residents in each catchment, where the country has figures.
4. Nearby places: supermarkets and grocers in each catchment, with chain or independent and rating.
5. xMap LLM: put each rival in a tier, then score each site for your format, using only these figures.
6. Make a report: one page per site for the property committee.

The committee gets a ranked report with households, rivals by tier and the score with its reason. For the watching afterwards, New homes with no supermarket yet flags new estates, and Discount, mainstream and premium grocers, city by city keeps the tier map up to date.