Footfall counts the people who walked past, which is why it gets counted: it is the easiest number to collect. The two that predict whether an activation worked are how long the people who stopped stayed, and how many came back with somebody else.
Footfall measures the venue, not the work
Footfall is a property of where the activation stood. Put the same stand in a wider aisle and the number goes up without a single thing about the experience changing. That is the tell: a metric that moves when you change nothing is measuring something other than what you did.
It survives because it is easy and because it is large. A big number in a recap deck feels like a result, and nobody in the room has an incentive to ask what it excludes.
If the number would be the same with an empty stand in the same aisle, it is not measuring the stand.
The two that actually predict
Dwell time is the first. It is the closest available proxy for whether the thing you built held anyone, because attention is the only currency a visitor actually spends. A stand that stops two hundred people for twenty seconds did not stop anybody.
Return with a companion is the second, and it is the harder one. Somebody who walks back with a colleague has done the only thing that matters on a show floor: they have recommended you inside the building, to a person who was not going to visit. That is word of mouth measured at the moment it happens rather than surveyed afterwards.
What to instrument, and when to decide it
Measurement fails for the same reason every time: nobody owned it in the brief. It is rarely expensive and almost always absent, so it gets discovered in the recap, when the only evidence left is a photograph of a crowd.
| Question | Why it has to be settled in advance |
|---|---|
| Who records participation | If it is the host team, it is part of their script and their headcount. If it is the system, it has to exist before the build is signed off |
| What counts as a participation | A completed turn, not a person who touched the rig. The definition decides the number and it cannot be chosen afterwards |
| Where the data lands | A leaderboard on the stand is not a report. Something has to export |
| What the baseline is | A number with nothing to compare it to is a number nobody will act on next year |
Why a leaderboard is a measurement instrument
A timed activity records itself. Every turn produces a row: who, when, how well. That is participation, dwell and repeat behaviour captured as a by-product of the thing being fun, rather than as a clipboard somebody has to remember to fill in.
It is also the reason the return metric is available at all. A visitor who has a time on a board has a reason to come back and beat it, and a reason to bring the colleague they want to beat.
Questions people ask
- How do you measure a brand activation?
- With dwell time and return visits, not footfall. Dwell time shows whether the experience held the people who stopped. Return visits, especially when somebody comes back with a colleague, show that the activation generated a recommendation inside the venue itself.
- Why is footfall a bad metric for activations?
- Because it is a property of the location rather than of the work. The same stand in a wider aisle reports a higher number without anything about the experience changing. A metric that moves when you change nothing is measuring something else.
- When should measurement be decided?
- Before the build is signed off. Measurement is rarely expensive and almost always absent, because it is nobody's line in the brief. Decided afterwards, the only evidence left is a photograph of a crowd.