Counting unique attendees versus visits without inflating your headline number

A three-day show. One buyer comes on all three days, parks in a different car park each morning, and goes out for lunch twice. Your entrance readers log seven events for her. Your marketing figure counts her three times. The audit certificate will count her once. Unique attendees versus visits is the difference between those answers, and each of the three is defensible under a different definition.

The standards settle this cleanly, and the settlement is more useful than most organisers expect, because it gives you a ladder with named rungs instead of an argument about which system is right.

What does the standard actually count?

UFI's Calculation Standards and Definitions, which sit as Annex 2 to UFI's Auditing Rules dated June 2021, define four things in sequence.

An entry or admission is each individual who enters the site during hours officially open. A visit is the action of a person who, on any official open day and hour, enters the event with an access document, and the standard attaches a hard cap: a person may only be counted once per day. A visitor is a person attending the exhibition, counted only once for the entire duration regardless of the number of visits. A repeat visit is each additional visit after the first that can be controlled, again capped at one per day.

The arithmetic that binds them is stated outright. Total visits equals total visitors plus total repeat visits.

FKM, the German association that has published verified exhibitor and visitor data since 1966, uses the same rule from the other side of the standard, defining the exhibition visitor as a person attending during the officially open hours and adding that "The person is only counted once for the entire duration of the trade fair, regardless of the number of visits."

Two bodies, two documents, one rule. Nothing in either allows a headline built by adding up daily totals.

The ladder from a reader event to a certified figure

Take a three-day show and walk a real file down the rungs.

Your entrance readers produce 31,800 read events inside the declared opening hours. That is the raw layer, and it is the only one the hardware knows about.

Collapse to one per person per day and you have 22,500 visits. The 9,300 events that disappeared are 29.2 per cent of the raw file, and they are almost entirely lunch, a second hall with its own reader, and people leaving to take a call. That percentage is a property of your building. A single-entrance show in one hall might lose 8 per cent at this step. A four-hall campus with separate readers can lose more than a third, and neither number says anything about the audience.

Collapse again to one per person across the whole show and you have 15,000 visitors. Repeat visits are 22,500 minus 15,000, or 7,500.

One detail bites at this step and it is worth deciding before the auditor asks. The cap is one visit per day, so you need a definition of a day. For a show open 09:00 to 18:00 that is obvious. For a show with a networking evening that runs to 00:30, a naive calendar-date split turns one person's single evening into two visits, and a file of 22,500 can pick up several hundred extra rows from nothing more than a clock. Define the day as the declared opening block rather than the calendar date, and document the choice.

The visit factor is 22,500 divided by 15,000, which is 1.50. Fifteen thousand people attended an average of 1.5 days out of three available, so the audience used half the days it could have used.

Now the publication question. Printing 22,500 as attendees overstates your unique audience by 7,500 people, which is 50 per cent of the real figure. Printing 15,000 to an exhibitor who wants to know how much traffic passed the stand understates the opportunity, because that exhibitor genuinely had 22,500 chances to be walked past. Both readers exist and they want different rungs.

Which figure goes on the certificate?

Either. The standard is explicit that figures relating to visits or to visitors are accepted, with one condition attached: the definition of the applicable term must be clearly understood in the context of its application, and UFI extends that requirement to statistics communicated in declarations, media information and promotional information.

The certificate enforces it structurally. The UFI Standard Audit Certificate that UFI publishes today, the form marked version 08/21, lays the visitor block out as visitors counted once, plus repeat visits, equalling total visits, with national and international rows across all three, and carries a note that either the visitor figure or the total visits figure is compulsory.

That layout is the most useful thing on the form. An organiser who completes all three cells has published the visit factor whether they meant to or not, because a reader can divide. An organiser who completes only one has made a choice, and it is a choice worth making deliberately rather than defaulting to whichever number the registration platform puts on its dashboard.

My own preference is to certify visitors and report visits alongside as a named secondary figure. Visitors is the harder number to produce, since it requires a person key that survives across days, and the difficulty is the point. A show that can produce it has an identity layer that works, and every other question about its audience gets easier. Visits is trivially derivable once you have visitors, and the reverse is not true.

The visit factor moves for two opposite reasons

This is where the ratio earns its place in a report, and where it misleads people who track it alone.

Say next edition the same show records 22,500 visits again and 13,500 visitors. The factor has risen from 1.50 to 1.67, which reads as improved engagement in every dashboard that treats a rising number as good news. What actually happened is that the unique audience fell by 1,500 people, or 10 per cent, and the remaining audience came more often.

The opposite case is just as easy to misread. Visitors rise to 16,800 while visits hold at 22,500 and the factor falls to 1.34. That looks like disengagement and it is a 12 per cent gain in unique audience, usually caused by successful recruitment of first-time attendees who come for one day.

The only safe treatment is to publish both components with the ratio, never the ratio alone, and to read a change in the factor as a question rather than a result. If your reporting layer collapses them into a single engagement metric, it will tell you a story with the sign flipped roughly half the time.

Where the underlying scan feeds disagree before you even get to this arithmetic, that is a separate problem with its own method, covered in reconciling badge scan counts across systems, and the stand-level version of the same distinction sits in unique visitors against total scans.

Digital and hybrid editions have their own rule

The standard was updated for hybrid formats and the rule it lands on is stricter than most platforms report.

For an event offering both a digital and a physical component, UFI requires counts to be provided separately: unique visitors attending only the digital event, unique visitors attending only the physical event, and where available an unduplicated count of both sets, which the standard calls the net visitor number. Adding the digital total to the physical total is exactly what the separation exists to prevent.

There is a second clause that removes a large number people like to quote. Counting takes place only during the event's official opening hours when synchronous interaction with exhibitors can happen, and UFI's internal rules state that asynchronous access to content outside those hours does not meet the definition of a visitor and should not be included in the visitor count. On-demand replays in the fortnight after the show are a content metric with a legitimate audience of their own, and they sit outside the audited figure.

Where this stops

The ladder gives you comparable numbers within your own show and it does not make two shows comparable.

A show certified on visits and a show certified on visitors will publish figures that differ by a factor of anywhere between 1.1 and 2.5 for the same real audience, driven by show length and buying rhythm more than by anything the organisers did. Comparing a three-day certified-on-visits show with a two-day certified-on-visitors show is arithmetic with two different units, and the certificate is the only place the unit is written down. Reading a competitor's document properly means finding that line first, which is one of the things an audit certificate will and will not settle.

The deduplication also depends entirely on your person key. If badges are reissued without linking to the original record, or if a hosted buyer group checks in under one identifier, your visitor count is too high and your visit factor is too low, and no auditor will catch it unless the check-back happens to draw one of those records. The counting rule is only as good as the identity behind it.

This week, take last edition's entrance log and compute all three rungs: raw reads inside declared hours, distinct person-days, and distinct persons. Put the two collapse percentages next to them. Then look at which of the three numbers your last media kit printed, and whether the page said which one it was. That single check tells you whether your published figure sits on the same base as the registered count everyone quotes internally, and whether your headline currently includes populations that belong on a separate line from visitors. The definitions themselves are short enough to read in an afternoon, and they are the foundation of every other number in the standards pack.