Home · Measurement
The currency problem
Why two European audience numbers can't simply be added
Each major market runs its own out-of-home measurement, built on its own definitions. The figures look like the same unit. They are not, and adding them invents an audience that does not exist.
Out-of-home is measured nationally, by joint-industry bodies that media owners, agencies and advertisers fund together so that everyone trades on a number nobody privately controls. Britain has its system, France another, Germany another, and so on across the continent — each a serious, audited piece of research, and each built on its own survey base, its own travel data and its own rules for turning "a person passed this site" into "a person had an opportunity to see this ad." Those rules are where the trouble starts. The headline figure from one country and the headline figure from another are denominated in different currencies that happen to share a symbol.
What an "impression" quietly assumes
Every out-of-home audience number is the end of a chain of reasonable adjustments, and the chain differs by market. Raw passage counts get discounted for visibility — whether a panel could actually be seen given its position, angle, illumination and the speed of the passer-by. They get adjusted for direction of travel, because half the people on a street are facing the wrong way. Some systems apply a "likelihood to see" model derived from eye-tracking; others use a flatter visibility factor. The upshot is that one country's "impression" might bake in a stricter attention assumption than another's, so two sites with genuinely identical footfall can post different audiences purely because their national methodologies discount differently. None of this is dishonest. It is simply local, and locally it is consistent — which is exactly why it does not survive being added across a border.
The practical damage and how to limit it
The error shows up the moment someone builds a pan-European reach figure by summing national audiences, or compares cost-per-thousand between markets as if the thousand meant the same thing. It does not. A defensible pan-regional plan keeps each market in its own currency, compares within markets rather than across them, and treats any single "European reach" headline as a presentation convenience rather than a measured fact. When a seller offers one continental number, the right question is which national methodologies were merged and how the joins were handled — a question good sellers answer readily because they had to make those choices deliberately.
Digital out-of-home complicates this further, because screens generate play logs — hard records of what showed, when and where — that sit awkwardly next to survey-based audience models. The industry is slowly reconciling the two: census-level play data anchored to panel-based audience research, so a screen's logged plays can be expressed as estimated impressions in the national currency. The direction is sensible, but it is happening at different speeds in different countries, which means yet another axis on which two markets fail to match. For the operator running an estate rather than buying media on one, the measurement that matters is simpler and more concrete — proof of play, uptime, and what actually rendered on each screen — and that operational reporting lives in the management platform, not in the audience survey.
The rule of thumb
Treat each national measurement system as a well-built instrument calibrated for its own country and uncalibrated for any other. Compare like with like inside a market; resist the arithmetic that adds across markets; and ask, every time a single European figure appears, what was assumed to manufacture it. The discipline is unglamorous and it is the difference between a media plan that survives scrutiny and one that quietly overstates its own reach.