Why footfall data arrives days later, and how to work with it

Footfall for any given day reaches xMap Flows about ten days after that day. So when Mobility Footfall reads "the last 7 days", it means the 7 days that end roughly ten days ago, not this week. When even those days are late, the block uses the newest days that exist and tells you exactly which ones in footfall_from, footfall_to and footfall_note.

This article explains where the delay comes from, shows what a window really covers on a given date, and sorts the decisions footfall serves well from the ones it cannot serve. It is for analysts and managers who want to trust the figures and say plainly where they stop.

Where the delay comes from

Mobility data of this kind is built from a large panel of devices. Before any figure is released it has to be gathered, checked, stripped of anything that could point to one person, grouped into small areas of the map and hours, and scaled from the panel up to the whole population. Each of those steps waits for the day to be complete and for late readings to come in. The result is a figure you can compare between places and weeks, at the cost of a few days.

The delay is the price of two things you want: that nothing about any person ever leaves the data, and that a Tuesday is measured the same way as a Saturday. A figure published the same evening would be rougher and easier to misread.

What a window covers on a given day

Here is how the windows read for a flow run on 2 November 2026. The window ends about ten days earlier, on 23 October. The data itself starts on 17 October 2026, so for now a long window cannot reach further back than that.

A run on 2 November 2026; the days count from the window’s end, about ten days before the run.

Every month that passes, the longer windows fill up. By late November the last 30 days will be complete, and by late January the last 90. Until then footfall_days tells you how many days a figure stands on, which matters most for Trend: it compares the last 7 days with the 7 before and needs at least four days of data in each.

When days are later than usual

Sometimes the newest days arrive later than the usual ten. Rather than return nothing, the block looks back for the newest days there are and uses those. It never hides this. The place carries a note such as: the days asked for have not arrived yet, so these are the newest 3 days there are, with their dates. A place with no data at all says so, and is not charged.

A made-up example: a retailer runs the block every Monday for its stores in Jakarta and Buenos Aires. One Monday, the Jakarta stores show their usual 7 days, while the Buenos Aires stores show 4, with a note naming the dates. The report is still useful; the reader simply knows the Buenos Aires figures rest on fewer days that week.

Decisions the delay does not hurt

Most uses of footfall are about patterns that hold for weeks or months, and a ten-day delay changes nothing for them:

• Choosing a site or signing a lease. You are betting on years, and the last few weeks are as good a guide as this week. See checking footfall before you sign a lease.
• Opening hours and rotas. The shape of a street’s day changes slowly, with seasons and new buildings.
• Ranking places against each other. Every place is measured over the same days, so the order is fair.
• Spotting a slow change. A street that loses footfall four weeks in a row is telling you something, even if each week is ten days old.

Decisions it cannot serve

• Today’s crowd. For what is happening now, look at live data: Traffic Flow measures road speeds at the moment the flow runs, and Weather forecast looks ahead.
• A one-off event last weekend. You will see its effect, but only about ten days later.
• Daily alerts. A flow that warns you every morning about yesterday’s footfall would always be ten days behind. Run footfall flows weekly or monthly instead.

Setting up a flow that respects the delay

1. Use a Schedule set to every week or every month, never every day.
2. Choose the longest window that answers your question; 30 days smooths out holidays and weather once it is full.
3. Keep footfall_from, footfall_to and footfall_note in the output, so every reader sees the dates.
4. Ask xMap LLM to mention when a place stands on fewer days than the others, so nobody over-reads it.
5. Send the result with an Email or a Make a report PDF.