How to count the businesses in a neighbourhood, and get a number you can trust

To count the businesses in a neighbourhood, define the neighbourhood as a shape or a point and distance, choose the kinds of business that matter, and ask a places database how many of those lie inside, leaving out places marked closed. The number is only as good as those three choices, so write them down: "cafes and bakeries, open, inside the boundary we drew on 4 November" is a count someone else can repeat; "about 40 cafes around here" is not.

This is for people who need one number for one area: a shop owner sizing up a street before signing a lease, a council officer describing a district, an investor comparing two neighbourhoods, a franchise team checking a territory. It covers how to set the boundary, which kinds to count, how to read the result, a made-up worked example, the common errors, and how to repeat the count automatically.

Step one: say where the neighbourhood is

People rarely agree on where a neighbourhood ends. Pick one of three ways and stick to it.

• A drawn boundary. Draw the area on a map, following streets, a river or a railway. Best when the neighbourhood has a clear shape.
• A point and a distance. A corner or an address with a walking distance around it: 250 m for a single street, 500 m for a short walk, 1 km for a district centre. Easy to repeat and to compare.
• A travel time. The area reachable in, say, ten minutes on foot. Closer to how customers move, but harder to explain.

For comparing several neighbourhoods, the same distance around each centre is the fairest choice, because every area then has the same size.

Step two: say what counts as a business

"Businesses" is too wide to be useful. A neighbourhood count normally answers a narrower question, such as how many competitors, how many possible customers, or how many shops on the high street. Choose the kinds of place deliberately:

1. Name the kinds you care about, such as Cafe, Bakery and Coffee shop, rather than a single broad word.
2. Decide whether chains count. A landlord may want all shops; an independent cafe owner may care more about other independents.
3. Leave out places marked closed, unless closures are the point.
4. Keep the same list of kinds for every area you compare.

How to read the number

A count is most useful next to another count. On its own, "28 cafes within 500 m" says little. Next to "11 cafes within 500 m of the other site", it says the first street is far more crowded. Next to the residents of the same area, where population figures exist, it becomes people per cafe.

Keep scale in mind too. In xMap's places database for November 2026, there are 3,111 cafes within 10 km of the centre of London, 2,573 in Sydney and 622 in Toronto. A neighbourhood count is a small slice of numbers like these, and small slices move with every opening or closure, so a change of two or three is not a trend.

3,111Cafes within 10 km, London

2,573Cafes within 10 km, Sydney

622Cafes within 10 km, Toronto

A worked example (made up)

Say someone planning a bakery in Montreal is choosing between two streets. They count Bakery, Cafe and Coffee shop within 500 m of each corner, open places only. Street A has 9 bakeries and 21 cafes; street B has 2 bakeries and 14 cafes. Then they count every kind of place to get a sense of activity: 410 places around A, 260 around B. B has fewer direct rivals and still a fair amount of activity, so it goes on the shortlist. These numbers are made up; the method is the point.

Where counts go wrong

• Edges. A business across the road from the boundary is outside a drawn area but may be a direct rival. Look at a slightly wider circle before deciding.
• Names of kinds. One database calls it Coffee shop, another Cafe. Count both if you mean both.
• Closed places. Counting places that have shut makes a street look busier than it is.
• A big area searched from its middle. One search from the centre of a wide area finds what is near the centre first. For a large district, search it in parts.
• Long lists cut short. A single search lists at most the 500 nearest places. In a dense centre, a wide circle can hold more; read the total the search reports rather than the number of places it lists.

How to repeat the count with an AI employee

For a one-off question, Ask the map answers in plain words, such as "how many bakeries are within 1 km of this address". To count several neighbourhoods the same way every month, set up an AI employee in xMap Flows:

1. Schedule: the first of each month.
2. My data: a small table of your neighbourhoods, one row per centre with its latitude and longitude.
3. Nearby places: Where set to Each incoming place, Within 500 m, the kinds you chose, Pass on set to Each incoming place, with counts. Each neighbourhood goes on with its count, the nearest one, the average rating and how many belong to a chain; with two kinds or more, a count of each.
4. Sort and limit: the most crowded first.
5. Google Sheets: one row per neighbourhood per month, so the counts build into a series.
6. Email: a short table each month.

What lands each month is a table of your neighbourhoods with their counts side by side, and a sheet that shows how each one moves over time. To draw a single neighbourhood instead, use Area of interest with a drawn shape and switch on Cover the whole area in Nearby places when it is wider than about 3 km.