Why population data is missing in some countries, and what to use instead

Population data is missing in some countries because a street-level grid of residents has to be built from detailed census counts, and those counts are not released at that level everywhere, are released under terms that do not allow this use, or have simply not been turned into a grid yet. In xMap Flows the result is plain: the Demographics block returns residents for places in covered countries and says when a place is not covered. Among the covered countries are France, Canada, India, Thailand, Egypt and Australia. The United Kingdom and the United States are not covered today.

This piece is for teams whose sites span several countries, and for anyone who ran Demographics on a place and got nothing back. It explains where a population grid comes from, how to tell a gap from an empty area, and what to measure instead so that a plan does not stall.

Where a population grid comes from

A national census counts people by address and publishes totals for areas: districts, wards, blocks. To make a grid of small hexagons, those totals are typically spread over the land, guided by where buildings are, and adjusted so each area still adds up. The finer the published areas, the better the grid. xMap's grid uses hexagons of about 0.1 km², so it needs fine inputs.

That chain can break at three points:

• The counts are too coarse. Some statistics offices publish only regional totals. Spreading a region's people over thousands of hexagons would be guesswork, so it is better left out than shown as fact.
• The terms do not allow it. Census tables may be free to read but not free to redistribute in a commercial product, or only under conditions that a grid cannot meet.
• The grid has not been built yet. Covering a country takes work: matching the census to its areas, checking the result. Coverage grows over time.

None of this says anything about how many people live there. A gap is a gap in the data, not in the population.

How to tell a gap from an empty area

The difference matters, because a quiet suburb in a covered country and a busy city in an uncovered one can both come back with low numbers if you are not careful.

1. Test-run the flow on one place in each country you care about. A search that finds no population data says so in the step, and costs nothing.
2. Look at the step's words, not just the figure. A place outside the grid gets 0 and a note; a real empty area, such as a port or a park, still has neighbours with people in them.
3. Widen the circle once. If 3 km around a city-centre address still shows no one, it is the country, not the street.

What to measure instead

When residents are not available, you are rarely left with nothing. The question behind the residents count is usually one of three: how much life is around a place, how many people could need what I sell, or how crowded is the competition. Each has a stand-in.

How busy the area is

Mobility Footfall measures people-hours of presence around a place: everyone there, residents, workers and visitors, scaled to the population. It covers most of the world from mid-October 2026, though not yet Australia, and each day arrives about 10 days late. It does not replace residents, because it counts presence rather than homes, but for a shop that lives on passing trade it can be the better measure anyway.

Places that follow people

Some places exist because people live nearby: primary schools, supermarkets, pharmacies, nurseries, places of worship. A count of these within 1 km, from Nearby places, rises and falls with the residential population. It is an indirect signal, so use it to compare areas inside one city, never across countries with different habits.

Housing and owners

In the United States, where the grid is missing, property records fill part of the gap: Property changes follows new owners and sales in the counties it covers, and the property listings dataset in Regional data sources shows homes for rent or for sale around a point. A street where many homes changed hands this year is one where new households are arriving.

Your own customers

The most reliable stand-in is often already in your files: where your current customers live or order from. Upload them to My data and count them by area. It measures your demand, not the population, which is what most plans really need.

A worked example

A made-up case: a chain of children's swim schools has pools in Toronto and London and wants to rank possible new sites in both cities. In Toronto, Demographics gives the children aged 0 to 9 and 10 to 19 within 3 km of each site. In London it gives nothing, because the United Kingdom is not covered.

So the team runs two rankings, never one. Toronto sites are ranked by children living nearby. London sites are ranked by a different score: the number of primary schools within 2 km, the footfall around the site on Saturday mornings, and how many of the chain's current London customers live within a 15-minute drive. The two lists are read side by side, and the board is told which measure each city used.

Mistakes to avoid

• Reading a 0 as nobody. Check the note first.
• Typing in old census totals by hand without saying where they came from. If you must, keep them in a separate column with their year and source written beside them.
• Treating footfall as residents. People-hours of presence include everyone passing through, and a busy station can have very few residents.
• Using a stand-in from one country to compare against residents in another.

Building it into an AI employee

A flow can handle mixed countries for you by choosing the measure per place.

1. My data brings in the candidate sites from every country.
2. Demographics adds residents within the distance you choose. Places outside the grid come through with 0 and a note.
3. A Branch splits them: population above 0 on one side, 0 on the other.
4. On the second path, Nearby places counts schools or supermarkets around each site, and Mobility Footfall adds people-hours where it has data.
5. Each path ends in its own Sort and limit and the same Email, so the inbox holds two ranked lists, each labelled with the measure it used.

For market entry work, the Entering a new country: people, rivals and couriers template is a good start; check its population step on one place first.