FlowsBook a callSign inCreate free accountSign up
Footfall and mobility

Footfall in small towns: low numbers, thin coverage and honest reading

Measured presence works in small places too, but the figures are smaller and noisier. How to read them fairly, what the gaps mean, and when not to trust a reading.

Part of the guide Location intelligence automation

By the xMap team · · 4 min read

Footfall in small towns can be measured, but it has to be read differently from a big city. The numbers are lower, a single week can swing a lot, some days may have no data, and a few spots may have none at all yet. Read it with longer windows, rank places against similar towns rather than against cities, and always check how many days stand behind each figure.

This article is for anyone choosing between rural sites or serving them: telecom and banking teams placing shops or agents, health and aid programmes planning mobile services, retail chains moving into secondary towns, and local councils. It explains why small places behave this way, what to check, a made-up example, the mistakes, and a flow that only reports what it can stand behind.

Why small places are harder to read

Mobility Footfall comes from a panel of devices, scaled up to the whole population, and is reported as people-hours of presence. In a city centre the panel is large, so the scaled figure is steady from day to day. In a town of a few thousand, far fewer devices stand behind each number. A coach party, a market day or a funeral moves the average more than it would downtown, and a quiet Tuesday can come back with very little.

So small does not mean wrong, but it does mean the error bars are wider. The cure is the same as in any survey with a small sample: more days, comparisons with similar places, and caution with fine differences.

What to check before trusting a number

FieldWhat to look for
footfall_daysHow many days in the window had data. Fewer than about two thirds of the window: read with care.
footfall_from, footfall_toWhich days were used. The newest days arrive about ten days late, and the data starts on 17 October 2026.
footfall_noteSays when a place has no data yet. That is not zero people, and it is not charged.
footfall_peak_hourA sensible peak (market morning, school run) is reassuring; a peak at an odd hour on few days is a warning.
footfall_trendOn small numbers a week-on-week change of a fifth may be noise.

The fields that tell you how much weight a small-town figure can carry.

A made-up example

Nothing below is a real reading. A mobile operator wants one new shop among six towns in Portugal's interior, each with a few thousand people. The first run, with a 7-day window, ranks them in an order that changes completely the next week. The team switches to the last 90 days and widens the circle to 400 m, which adds up the blocks around each town's square. Two towns now sit clearly above the rest on most runs, one of them with all its days filled and a peak at the Saturday market. The other has data on fewer than half its days, so the team keeps it as a candidate but does not rank it first.

A health programme in Kenya uses the same approach on highway towns, comparing each only with towns of a similar size along the same road. A credit union in Quebec ranks its village branches against one another and ignores the gap between them and Montreal.

Mistakes that lead to bad decisions

  • Comparing a town with a city. The city wins every time and teaches you nothing. Compare small with small.
  • Reading a 7-day window. Use 30 or 90 days in small places.
  • Treating no data as empty. A place without figures yet is unknown, not deserted.
  • Splitting hairs. Two towns a tenth apart are tied. Look for clear gaps.
  • Keeping the 200 m circle out of habit. In a small town the square, the main street and the bus stop may sit in different blocks; 300 to 500 m around the centre gives a steadier figure.

A flow that is honest about its own data

  1. Schedule: once a month.
  2. My data: the town centres you are comparing, grouped by size.
  3. Mobility Footfall: 400 m, last 90 days, with Footfall, Busy hours and Busiest days.
  4. Branch: towns with enough days of data go one way, the rest another.
  5. Sort and limit on the well-covered side: highest average day first.
  6. Nearby places: what each town already has, such as a pharmacy, a bank or a phone shop.
  7. Email: the ranked towns, and a separate short list of towns with too little data to rank.

The inbox gets two lists, which is the honest answer: these towns can be compared, and these cannot yet. Over a few months the second list shrinks as more days arrive. Templates in the same spirit: Rural towns with people and no phone shop and Villages with no health centre nearby. The telecom and non-profits pages show more, and the location intelligence automation guide covers the wider picture.

Questions

Is footfall reliable in a small town?

It is usable with care. Figures rest on fewer devices, so use a 30 or 90 day window, check footfall_days, and compare only with places of a similar size.

What does it mean when a town has no figures?

That there is no data there yet. It is not the same as zero people, and the place is not charged.

Should I widen the circle in a small town?

Often yes. 300 to 500 m around the centre adds up the blocks around the square and main street and gives a steadier average than a single block.

Is every country covered?

Most of the world is, from 17 October 2026, but not yet Australia.

Templates that do this

Read next

More on footfall and mobility · Every article · Templates

Put it to work

An AI employee in xMap Flows does this on live data, on its own timetable. Start from a template or describe it in plain words.