A sponsorship visibility adjustment factor separates walking past from actually noticing

The sponsor report claimed 19,240 impressions on the registration hall banner. Every one of those 19,240 was a badge scan at a door, which means the number is a count of people who walked through an opening, and nothing in it says anybody looked up.

A sponsorship visibility adjustment factor is the piece that closes that gap, or at least labels it. It is a multiplier between zero and one, applied to gross exposures, estimating the share that plausibly ended in somebody noticing the asset. Out of home measurement has run on exactly this distinction for years, which makes it the obvious place to borrow from.

Two different quantities, and organisers report the first

The distinction has names in outdoor media, and the names are worth adopting because they force the question into the open.

Geopath's published glossary defines circulation as "A measurement of traffic volume in a market", adding that "Circulation only estimates the number of people with an opportunity to see an Out of Home display". Likelihood to see is the second quantity: "The portion of the OTS (Opportunity to See) audience who are likely to see an ad." The glossary records that "Out Of Home is the first medium in the US to move from reporting OTS audiences" onward from that base.

Almost every exhibition sponsorship report in circulation today is an opportunity-to-see report wearing an impressions label. The count is a door scan or a hall entry, the word is impressions, and the price was set as though the second quantity had been measured. Naming the two separately costs a line in a report and removes an argument you would otherwise have in a renewal meeting.

What the index between them actually is

The bridge in out of home is the Visibility Adjustment Index. Geopath's glossary defines it as "A ratio or the percentage of a unit's total OTS audience (circulation) who are likely to notice an ad", and states that "VAIs are derived from Geopath's visibility study."

The inputs are more interesting than the definition. Geopath's best practices, standards and protocols document, published by its Futures Council in August 2019, lays out what the visibility adjustment takes into account: "Observed dwell time, degrees off-center (at optimal view), and apparent size (at optimal view)". The same table frames the whole thing as three questions. "How large does the media APPEAR within the audience's field of view? WHERE is the media within the audience's field of view? How much TIME does the audience have to see the media?"

Those three questions transfer to an exhibition hall without modification. A gantry banner is large and overhead and passed in four seconds. A sign above a coffee queue is smaller and directly in front of somebody standing still for two minutes. The academic side of this has been worked on too: Wilson and Casper published a study of location and visual saliency in capturing attention to outdoor advertising in the Journal of Advertising Research in 2016, in volume 56, issue 3, pages 259 to 273, with one author writing from the Traffic Audit Bureau for Media Measurement.

How much does distraction move the number?

More than most factor-setting exercises assume, and there is a clean experiment to point at.

Schmälzle, Lim, Cho, Wu and Bente published a study in PLOS ONE in 2023 that put participants in a virtual reality drive past a series of roadside billboards while an eye tracker recorded every fixation. In the free-viewing condition the eye tracker recorded a mean of 52.8 billboard fixations per participant. When the same drive was run with a concurrent counting task to occupy attention, the mean fell to 21.8, a difference the authors report as significant at p below 0.001.

The memory measures moved with it. Free-viewing participants recalled an average of 6.45 billboards, a recall rate of 32 per cent, and recognised 14.6, a recognition rate of 73 per cent. Under the counting task those fell to 2.95 recalled, 15 per cent, and 7.1 recognised, 35 per cent. The paper also reports that a billboard being looked at at least once raises the likelihood it will be remembered by a factor of five to twenty, depending on condition and memory measure.

Read that across to a show floor and the implication is direct. An attendee hurrying to a meeting they are late for is the counting-task condition. An attendee standing in a coffee queue with nothing to do is the free-viewing condition. Halving of attention is roughly what the experiment produced, and a single visibility factor applied across a whole hall is asserting that this variation does not exist.

Scoring your own hall

You cannot run an eye-tracking study on your own show floor, and you do not need to in order to produce something better than a constant. Score the assets against the same attributes the outdoor model uses.

Four attributes, each scored from zero to three. Apparent size at the typical viewing distance. Position in the field of view, where overhead and eye level score above floor level and above behind-the-shoulder. Dwell available, meaning the seconds a person spends within the viewing zone. Clutter, scored inversely, meaning how many competing messages sit in the same field of view.

Sum the four for a maximum of twelve, then scale against a ceiling. Using a ceiling of 0.60, the factor is the score divided by twelve, multiplied by 0.60.

A banner above the coffee queue scores 2 for size, 3 for position, 3 for dwell and 1 for clutter, which is 9. Nine twelfths of 0.60 is 0.45. An aisle sign scores 1, 1, 1 and 0, which is 3, and three twelfths of 0.60 is 0.15. A hall entrance gantry scores 3, 3, 1 and 1, which is 8, giving 0.40. A keynote stage backdrop in front of a seated audience scores 3, 3, 3 and 3, giving the full 0.60.

The rubric is arbitrary in its details and consistent in its application, which is the trade worth making. Two people scoring the same hall should land within one point of each other on most assets, and where they do not, the disagreement is about something real and is worth ten minutes.

Why does the factor need a ceiling below one?

Because even the best placed sign in the building is missed by some of the people standing in front of it.

The Schmälzle figures give a sense of the ceiling's neighbourhood. Under ideal conditions, free viewing with no competing task, 73 per cent of billboards were recognised afterwards and 32 per cent recalled unprompted. Those are memory measures taken after the fact, so they cannot be read as noticing rates and cannot be lifted across directly. What they establish is that the honest ceiling is well below one, and that a factor of 0.60 for the single best asset in a hall is a defensible position, and calling it pessimistic would be the harder argument to make.

Setting the ceiling at 1.0 would mean asserting that every person passing under the keynote backdrop registers the sponsor's logo. Nobody believes that, and putting it in a report invites the sponsor to test it.

Publishing the assumptions is the point

A visibility factor is a judgement, and a judgement that is written down and applied consistently is worth having. The same judgement applied silently is a way of moving a number.

Publish the rubric, the score for each asset and the ceiling, in the same document as the impression count itself. Then a sponsor who thinks 0.45 is generous can substitute 0.30 and see the effect, which is a much better conversation than the one where they suspect the factor was chosen after the answer.

Be equally explicit about what you did last year. If the factor for an asset moves from 0.35 to 0.45 between editions, say why, and expect the question, because a ten point move on a factor changes the reported impressions by 28.6 per cent with no change in the hall at all.

Where this stops

The factor is calibrated against nothing. That is the honest limit and it does not have a workaround at exhibition scale.

Out of home derives its index from a visibility study. An organiser scoring a hall on four attributes is applying a structure borrowed from that work without the underlying measurement, and the output should be described as an assessment, with the word rate reserved for quantities somebody measured. The value is in consistency across assets and across editions, which lets you compare a coffee queue banner to an aisle sign fairly, and it does not make either figure true in absolute terms.

The second limit is that the factor multiplies through into every number downstream of it. A judgement of 0.45 against 0.30 changes the impression count by half and therefore changes the cost per thousand you publish by the same proportion, which is why the factor deserves more scrutiny than the traffic count everybody argues about. It is also the largest single contributor to the uncertainty around the final figure.

Score five assets from your last edition on the four attributes this week, without looking at the impression numbers you already published. Then compare the factors your rubric produces with the ones implied by last year's report. Where the two disagree by more than about ten points, you have found a figure somebody chose without scoring it, and it is worth fixing before it reaches the sponsor's view of their own performance.