Participants versus visitors terminology and why the broader word creeps in

Somebody in marketing asks whether the show can say participants this year instead of visitors. The reason given is that it reads better in a sponsorship deck and the competitor uses it. Nobody objects, because participants versus visitors terminology sounds like a style question and the number in the box does not change that week.

Three editions later the participant figure has grown eleven per cent and the visitor figure has not moved. No fraud, no error, no decision anybody would recognise as a decision. The bucket simply got wider every time somebody had a reason to put another group in it.

What does the standard actually define?

UFI's Auditing Rules for the Statistics of UFI Approved Events, dated June 2021, is unusually precise about this, and the precision runs in one direction only.

Annex 2 defines a visitor as a person attending the exhibition, counted once for the entire duration of the fair regardless of how many visits they make. It defines what a visit is, caps it at one per person per day, and lists who is excluded from the visit calculation: staff of the exhibition venue and organiser, staff working for service providers, staff of exhibiting companies, speakers during the event, and media representatives.

It then defines the widest total it recognises. Total attendance, in the standard's words, "is equal to the unique visitors plus the exhibitors' staff plus the speakers plus the media representatives" (UFI, 2021). Notice what is absent from that list. Contractors and service provider staff are excluded from visits and are not added back into total attendance either, so they sit outside every certified figure.

What Annex 2 never defines is a participant. The word does not appear as a counting term anywhere in the document, and there is no clause telling an auditor what to check if you publish one. That absence is the whole mechanism. A term with no definition cannot be audited, cannot be compared, and cannot be wrong.

The four ways the bucket widens

Watch a participant figure over several editions and the additions arrive in a predictable order, each one defensible on its own.

• Exhibitor staff. The first and largest addition, usually justified on the grounds that stand personnel are genuinely at the show. Whether they belong in a published attendance figure is a question with two sides and P8 argues it out.
• Speakers and the conference programme. Added when the conference grows enough to have its own budget line and its own director who wants credit for the audience.
• Contractors and build crew. Added when somebody notices the badge system already issues them badges, and the data is sitting right there.
• Accredited media and influencers. Added last, usually when a press office is trying to demonstrate reach.

None of these arrive with a memo. Each one is a small extension made by a person solving a local problem, and the cumulative effect is invisible because the label never changed.

Four editions of quiet drift

Here is the pattern with numbers on it. One show, four consecutive editions, publishing a participant figure throughout.

In 2023 the show published 15,200 participants, being 12,000 visitors plus 3,200 exhibitor staff. In 2024 the conference director asked for speakers to be included, so 12,050 visitors plus 3,260 exhibitor staff plus 250 speakers gave 15,560. In 2025 the operations team added contractors, and 12,010 visitors plus 3,300 exhibitor staff plus 265 speakers plus 900 contractors gave 16,475. In 2026 accredited media went in, and 11,950 visitors plus 3,340 exhibitor staff plus 280 speakers plus 940 contractors plus 390 media gave 16,900.

The published headline moved from 15,200 to 16,900, which is 1,700 more, or 11.2 per cent growth across three editions. The visitor base moved from 12,000 to 11,950, which is 50 fewer, or a fall of 0.4 per cent. A show that lost audience reported double digit growth, and every individual number in the sequence is a true count of a real group of people.

The 2026 figure is worth one more line of arithmetic. Under Annex 2, total attendance for that edition would be 11,950 visitors plus 3,340 exhibitor staff plus 280 speakers plus 390 media, which is 15,960. The published participant figure of 16,900 exceeds the widest total the standard defines by 940, because contractors sit outside it. The show has drifted past the edge of the vocabulary without anyone noticing there was an edge.

A one column test for drift in your own history

You can find this in your own numbers in about twenty minutes, and it does not need the badge detail that produced the original figures.

For each edition, divide the published headline by the visitor count you can defend, whether that is an audited figure or a badge collection count you trust. Put the ratio in a column beside the years and look at it.

On the four editions above, the ratio runs 15,200 over 12,000, which is 1.27; then 15,560 over 12,050, which is 1.29; then 16,475 over 12,010, which is 1.37; then 16,900 over 11,950, which is 1.41. A show whose format did not change has no reason to move that ratio by fourteen points in three years. Real changes in audience behaviour, such as more people returning on a second day, move the ratio slowly and in ways you can explain from the calendar. Definitional changes move it in steps.

The test has one useful property beyond detection. It gives you the size of the correction before you have reconstructed anything. If your ratio has climbed from 1.27 to 1.41 and you want to restate the current headline on the original composition, the first approximation is the published figure divided by 1.41 and multiplied by 1.27, which on 16,900 gives 15,222. That is close enough to open the conversation about whether last year's growth story was real.

Where the ratio has been flat for five editions, you have a portfolio problem you do not have, and you can stop.

Why is the broader word so attractive?

It is worth being honest about the incentives, because pretending this is only sloppiness will not fix it.

A bigger number sells more space. An exhibitor deciding between two shows reads the headline before the footnote, and a 16,900 beats a 11,950 in a thirty second comparison. Sponsorship packages are priced against reach. Venue negotiations reference footfall. Internally, a growing figure is easier to present than a flat one, and the person presenting it did not choose the label.

There is also a genuine argument for the broader count, which is that an exhibitor's stand is worked by people who meet contractors, speakers and journalists as well as buyers, so total attendance describes something real about the density of the hall. That argument is fine. It stops being fine when the broader figure is published alone, under a word nobody has defined, in place of a figure that has a definition and could have been audited.

The evidence from published industry data runs the other way. UFI's Global Exhibition Industry Statistics of April 2026 reports around 32,000 exhibitions worldwide with 4.7 million exhibiting companies and 318 million visitors, keeping the two populations in separate lines. Nobody adds them together into a global participant number, because the sum would not mean anything.

Publish both numbers, labelled

The fix costs one extra line in the media kit and one extra column in the post-show report.

Publish the visitor figure, defined as unique people counted once across the run. Publish the wider figure beside it if you want one, and write out its arithmetic in the same place, naming each population and its count. Where the wider figure matches the Annex 2 definition, call it total attendance and say so, because a term with a published definition is worth more to a buyer than a term without one.

Then freeze the composition. The rule that makes this stick is that any change to what goes into the wider figure has to be applied backwards across every edition you publish for comparison. A team that has to restate four years of history in order to add contractors will usually decide that contractors were not that important.

Setting the classification at the point each badge is created is what makes the restatement possible at all, and the mechanics of that belong to D12. Comparing your restated figures against another show's is P12's problem and needs the certificate rather than the label.

Where this stops

Publishing both figures fixes your own reporting. It does not fix the market you compete in, and it can cost you in the short run, because a competitor still quoting an undefined participant number will look larger than you on a slide where neither footnote is read.

The second limit is that the standard's own boundary is a judgement rather than a law of nature. Annex 2 puts contractors outside every total and accredited media inside total attendance, and I can construct arguments for moving either. What the standard gives you is a line drawn in public that both sides of a negotiation can read, and that is worth more than a better line drawn privately.

The third is retrospective. Once a portfolio has published undefined participant figures for several years, the history is often unreconstructable, because the badge classes that would let you decompose the old numbers were never stored. In that case the honest move is to publish the visitor series from the earliest edition you can rebuild and mark everything before it as not comparable.

This week, take your last four published headline figures, write out the arithmetic behind each one as a sum of named populations, and check whether the same populations appear in all four. Where the composition changed, the published counting definitions will tell you which of your figures had a defined name available all along.