← All Labs

Footfall / Store Visit Measurement

Flagship
Measurement
01

Customer question

Did our advertising drive people into our stores?

02

What this solution does

It compares how often people who saw the ad actually walked into a store versus people who didn't see it — the gap is the estimated effect of the campaign on foot traffic.

Specifically, it answers: Did exposure to a campaign increase the rate at which people visited physical stores, relative to people who weren't exposed?

03

How it works

Location or exposure data identifies who was shown the campaign, then visitation to defined store locations is measured for both an exposed group and a control group; the difference in visitation rate is the estimated lift.

04

Data requirements

  • Defined store locations / geofenced polygons
  • Exposed population tied to campaign delivery data
  • Comparable control population
  • Location/visitation event data for both groups
05

Discovery questions

How are store locations geofenced, and how precise is that boundary?

Imprecise geofences can misattribute nearby foot traffic to the wrong store or miss genuine visits.

How is the control group selected to avoid bias?

If the control group differs systematically from the exposed group (e.g. lives further from stores), the lift estimate will be biased.

What baseline visitation rate exists before the campaign?

Lift is only meaningful relative to an accurate pre-campaign baseline.

06

Build an experiment

07

Generate synthetic data

08

Data readiness

Generate a dataset in the previous step to run readiness checks.
09

Run model / simulation

Generate a dataset first to run the model.
10

Results

Run the model in the previous step to see results here.
11–12

Business ⟷ Technical interpretation

The exposed group visited stores at a rate of exposedRate%, compared to controlRate% for the control group — an estimated lift of estimatedLift%. This suggests the campaign meaningfully increased in-store visits, though 'visited after seeing the ad' is not the same as 'visited because of the ad.'

Shown as a template — the highlighted tokens are filled in once the model has been run.

13

What would you do next?

  • Check whether lift holds consistently across different store formats or regions
  • Pair with an incrementality or holdout test if a more rigorous causal estimate is needed
14

Solution & vendor landscape

Implementation approaches

Specialist location/footfall measurement vendorPlatform-native store visit measurementIndependent measurement partner using mobile location panels
See the full Vendor Landscape →