How to compare footfall across stores in a network, fairly
A city-centre store always wins on raw footfall. How to group stores, pick fair measures and compare footfall across stores in different cities and formats.
Part of the guide Location intelligence automation
To compare footfall across stores in a network, measure every store with the same circle, the same days and the same kind of days, then judge each one against stores of its own format and against its own past, not against the whole list. Raw people-hours mostly tell you which stores sit in the busiest parts of town, which you knew already. The useful comparisons are the shape of each store's day, its change over time, and how much trade it turns out of the presence around it.
This is for retail, banking, pharmacy and food operators with ten or more sites who want a ranking they can act on: which stores to visit, which to staff up, which to question. It uses Mobility Footfall, which measures how busy the area around each place is in people-hours of presence, and your own sales figures from a table.
Why a single ranked list misleads
Put a flagship on a shopping street next to a suburban store and the flagship will show several times the footfall, every week, whatever either manager does. Sorting the network by footfall per day then rewards location, not performance. It also hides the stores that matter most: a suburban branch whose footfall fell after a road layout changed looks fine in a list topped by city-centre stores, because it was never near the top.
Footfall also counts hours, not people. A store beside a hospital or a university, where people stay for hours, scores higher than one on a street people walk through quickly, even if the second sees more passers-by. That is a feature when you compare like with like, and a trap when you do not.
Group the stores first
Before any number is read, give each store a format in your stores table. Most networks fall into a handful of groups:
- City-centre street stores, open to walk-in trade all day.
- Mall units, where the centre itself draws the crowd.
- Transit stores, at or beside a station, with sharp peaks.
- Suburban or neighbourhood stores, serving people who live nearby.
- Roadside and retail-park stores, reached mostly by car.
Rank inside each group. A mall unit is compared with mall units, a station store with station stores. If a group has only two or three members, compare each store with its own past instead.
Three measures that travel well
- Change on its own past. The trend field compares the last 7 days with the 7 before. Read monthly, the same store against itself is the fairest comparison there is, because the location is held still.
- The shape of the day. Busy hours gives each store its busiest hour and how the day splits between morning, midday, evening and late night. Two stores with the same total can have very different days, and the shape tells you when to staff and when to open.
- Trade per unit of presence. Divide a store's sales (from your own till data, in a table) by its footfall. A store that turns little of a busy street into sales is a question for the manager; one that does well on a quiet street is a lesson for the others.
A worked example (made up)
Say a phone retailer has stores in Lisbon, Seoul and Johannesburg and wants to know which deserve a visit from the regional manager this month. The figures below are invented to show the method, not measured.
| Store | Format | People-hours a day | Change on prior week | Sales per 1,000 people-hours |
|---|---|---|---|---|
| Lisbon, Baixa | Street | 52,000 | +3% | 41 |
| Lisbon, Benfica | Neighbourhood | 11,400 | -9% | 63 |
| Seoul, Gangnam | Street | 58,500 | +1% | 22 |
| Seoul, Mapo | Neighbourhood | 13,200 | +2% | 58 |
| Johannesburg, Sandton | Mall | 37,800 | +5% | 35 |
Made-up figures for illustration only.
Sorted by footfall, Gangnam leads and Benfica sits near the bottom. Grouped and compared fairly, the story turns. Among street stores, Gangnam turns about half as much of its footfall into sales as Baixa, so it gets the visit. Among neighbourhood stores, Benfica sells well on what it has, but its footfall fell 9% in a week: worth a call to ask whether a road closure or a new rival explains it before treating it as a trend.
Mistakes to avoid
- Different circles for different stores. A 500 m circle in a dense centre takes in several streets; 200 m around a mall covers the mall. Pick one radius per group and keep it.
- Mixed windows. Thirty days for one store and seven for another is not a comparison. Run the whole network in one block, with one window.
- Treating no data as zero. A store with "No footfall data here yet" in footfall_note is unmeasured, not empty. Leave it out of the ranking. Footfall does not cover Australia yet.
- Weekends that differ by country. The block follows each place's own weekend, Friday and Saturday in most of the Gulf, Saturday and Sunday elsewhere. Compare weekends with weekends, not Saturdays with Saturdays.
- Reading a week as a verdict. One week can move with holidays and weather. Look for a change that holds over several weeks.
Doing it with an AI employee
An AI employee in xMap Flows can run this comparison every month and put the result where the team works. It belongs with the other repeatable jobs on location intelligence automation.
- Start with a Schedule on the first Monday of each month.
- Add My data with your stores table: name, address or position, format, and last month's sales.
- Add Mobility Footfall on each incoming place, 200 m, the last 30 days, returning Footfall, Busy hours and Trend.
- Add a Calculated field that divides sales by footfall_daily, per 1,000 people-hours.
- Add xMap LLM to read each store against the others in its format and flag the three that most need attention, with a reason.
- End with Make a report or an Email to the regional managers.
What lands in the inbox is a short list of stores to look at, each with its format, its footfall, its change and why it was picked, plus the full table for anyone who wants it. Add Nearby places before the AI step if you want it to see new rivals around a store whose footfall moved.
Questions
Can I compare a mall store with a street store?
Only on change and on the shape of the day. Their totals measure different things: the mall's draw and the street's passing trade.
Why not just rank stores by footfall per day?
Because it ranks locations, not stores. The busiest districts stay on top whatever the stores do.
How often should a network comparison run?
Monthly is a good rhythm for decisions, with a weekly report for spotting sudden drops. Footfall is about 10 days behind, so a daily run adds little.
Does footfall tell me how many people entered the store?
No. It measures presence in the area around the store in people-hours, not entries. Your door counters or tills answer that question.