How weather and footfall go together, and how to check it for your own sites

Weather and footfall move together at many sites, but in directions you have to measure rather than assume. A wet Saturday can empty an open-air market and fill an indoor mall a kilometre away; a hot week can do the reverse. Nobody can give you a rule for your street. What you can do is keep a weather log next to a footfall log for each of your sites, week after week, and see what your own figures say.

This article is for retailers, restaurant groups, mall and market operators, and marketing teams who plan promotions or staff by the forecast. It explains what each kind of data gives, how to line the two up, a made-up example, the traps in reading them, and an AI employee that keeps both logs for you.

Two kinds of data on different clocks

Weather forecast looks ahead: up to 16 days of temperature, rain, wind and UV for each place, from national weather services. It does not look back. If you want to know what the weather was on a day, record it on that day.

Mobility Footfall looks back: people-hours of presence around each place over a window of days, plus the busiest hour, the busiest day and how the day splits. Each day arrives about 10 days later, so its window always ends about 10 days ago. The figures are averages over the window (7 days at the shortest), not one value per day.

So the two never arrive together. The way to join them is a log: one row a day with that day's weather at each site, written on the day, and one row a week with the footfall of a week that has since arrived. You then match the footfall week to the weather days it covers.

How to read the two side by side

• Compare like with like. Put weekends next to weekends and weekdays next to weekdays. Mobility Footfall's Day type does that split for you.
• Use the week's weather, not one day's. Because footfall is a weekly average, add up the rain (rain_total_mm) and take the highest temperature (temp_max) over the same days.
• Look at the shape as well as the size. A rainy week may not lower footfall but may move it: footfall_by_part shifts from evening to midday, or the busiest hour moves indoors to lunchtime.
• Wait for enough weeks. Four weeks tell you little. A season of weekends, with several wet and several dry, starts to show a pattern.
• Write down what else happened. School holidays, a public holiday, a big event or roadworks can move footfall more than the rain did.

A made-up example in Toronto

Invented figures, for illustration only. A clothing retailer has two stores in Toronto: one on an open shopping street, one inside an enclosed mall. Every evening a flow records the day's rain and top temperature at each store. Every Monday another flow adds the weekend footfall of the latest week that has arrived, 200 m around each store. After ten weekends the team sorts the rows by rain:

Invented people-hours for illustration; not real figures for Toronto.

In this made-up log the street store loses about a quarter of its footfall on wet weekends while the mall store gains a little. Ten weekends is a small sample and two of the wet ones fell in a holiday week, so the team treats it as a lead, not a law. The plan that follows is modest: when the Thursday forecast shows heavy rain for the weekend, move one person from the street store to the mall store and push the mall in that weekend's ads.

Mistakes to avoid

• Matching the wrong days. Footfall read on a Monday covers a week that ended about 10 days earlier. Match it to the weather rows for those dates, not to this week's forecast.
• Using a forecast as the record. Log the weather on the day, from a run the same evening. A forecast made a week ahead is often wrong.
• One store, one season. A pattern at one site in autumn may reverse in summer or at another site.
• Mixing weekdays and weekends. A rainy weekday and a dry Saturday are not comparable.
• Reading people-hours as shoppers. If rain keeps people sheltering in a doorway for longer, presence can rise while sales fall. See why footfall is not sales.

Let two AI employees keep the logs

The first AI employee writes the weather down every day:

1. Schedule: every day at 21:00, in the stores' time zone.
2. My data with your stores.
3. Weather forecast: Each incoming place, 1 day, with Temperature and Rain.
4. A Google Sheets step that adds a row per store: the date, rain_total_mm, temp_max and weather_today.

The second writes the footfall down every week and does the matching:

1. Schedule: every Monday at 08:00.
2. My data with the same stores.
3. Mobility Footfall: 200 m, 7 days, Weekends only, with Footfall and Busy hours.
4. A Google Sheets step that adds a row per store with footfall_daily, footfall_from, footfall_to and footfall_by_part.
5. xMap LLM on the whole list, asked to write one line per store on its latest week: the footfall, the dates covered and how the day split.
6. Email the note to the planning team.

What lands in the inbox is a short Monday note: each store's weekend footfall for the latest week that has arrived, and the dates to look up in your weather log. In the sheet, a column for the week's rain beside each footfall row lets you sort wet weeks from dry ones. To act on the next forecast, the templates weather-driven promotions per store and rainy weekend: staff the right stores turn the coming weekend's forecast into a note per store.