Footfall is the easiest number to produce and the least useful one to receive. It measures the location, not the work. Put the same counter in the same mall on the same Friday with no activation running at all and the number barely moves.
So we stopped leading reports with it.
The four numbers, in order
Every activation we run reports the same funnel, and the report is built around the middle of it:
- Passed — the location's number. Context, not performance.
- Stopped — chose to break stride. The first number the work is responsible for.
- Tried — took the sample, played the game, entered the draw. Intent.
- Converted — bought, subscribed, or came back. What the client is paying for.
The two that tell you whether the activation worked are stop rate (2 ÷ 1) and try-to-convert (4 ÷ 3). One measures whether the thing was interesting. The other measures whether the product delivered on it.
A high stop rate with a low conversion is a creative success and a product problem. A low stop rate with a high conversion means you built the right thing in the wrong place. Footfall tells you neither.
How we count, and why it is boring on purpose
The method matters more than the tooling, and the method is deliberately dull:
- One counter per gate, one tally per promoter, reconciled at the end of every shift. Not at the end of the campaign, when nobody remembers Tuesday.
- Stopped is defined before the campaign starts. Ours: stationary, facing the unit, for three seconds or more. Any definition works as long as it is written down and the same on day one and day nine.
- Try is counted by stock movement, not by a promoter's estimate. Units out, minus units returned, minus damages.
- Convert needs the client. This is the number we cannot produce alone, and we say so in the proposal. Without a till, a code, or a redemption, the last row of the funnel is a guess — and a guessed conversion number is worse than an absent one.
Where the benchmark comes from
The obvious question is what a good stop rate is. We will not print a table of bands here, and we are suspicious of anyone who does: stop rate on a supermarket aisle and stop rate on a mall atrium are different games, and a number lifted from someone else's category tells you nothing about yours.
A benchmark has to come from one of three places, in this order of preference:
- Your own prior campaigns, if the funnel was counted the same way. Usually it was not, which is worth knowing early rather than at the readout.
- Your category, where a distributor or a retail partner will share it.
- The first campaign itself. If neither of the above exists, say so in the proposal and treat run one as the baseline — a measured number from your own aisle beats a borrowed one from anybody's.
What matters far more than the band is that the definition is fixed before day one and does not move afterwards. A report saying "4.8%" is only meaningful next to the sentence explaining what was counted.
The uncomfortable consequence
Reporting this way means some campaigns come back with a number that is not flattering, and it means saying so out loud at the readout rather than leading with the footfall figure and hoping nobody asks.
That is the commitment, and it is worth being explicit about it before a campaign rather than after: if the stop rate on a format you love comes back at 1.9%, you will hear 1.9%. A low number that identifies a fixable placement problem is worth more than a five-figure footfall total that identifies nothing — the unit was on a wall it should not have been on, and that is a change you can make next week.
Measure the thing you can change.