Why two shows in one sector report attendance so differently

An exhibitor sales director puts two competitor shows side by side in a deck. Same vertical, same continent, halls that look about the same size on the floorplan. One publishes 15,000. The other publishes 21,300. The obvious reading is that the second show is 42 per cent bigger and the rate card should reflect it.

That reading is usually wrong, and the question of why two shows report differently almost never has an answer about audience size. Both organisers can be counting carefully, auditing honestly and publishing figures they can defend, and still produce numbers that cannot go on the same chart. The divergence lives in the base each one chose.

Where does the gap actually come from?

UFI's Auditing Rules for the Statistics of UFI Approved Events, dated June 2021, carry the counting definitions in Annex 2, and they define three separate totals from the same floor.

The first is visitors. A visitor is counted once for the whole run regardless of how many days they attend. The second is total visits, which Annex 2 gives as visitors plus repeat visits, where a repeat visit is each controlled return after the first and no more than one is counted per person per day. The third is total attendance, which the standard defines as unique visitors plus exhibitors' staff plus speakers plus media representatives.

Three totals, one show, all of them arithmetically correct. Nothing in the standard says which one you have to publish. Annex 2 says only that figures related to either visits or visitors are accepted by UFI, and that the applicable term must be clearly understood in the context of its application, including in media information and promotional material.

That last clause is doing a lot of work and almost nobody complies with it.

One show floor, three defensible totals

Take a three day industrial show and count it properly.

Across the run, 15,000 distinct people came through the visitor entrance with an access document. Of those, a large share came back on a second or third day, producing 6,800 controlled repeat visits after the first-day count. On the exhibitor side, 4,100 people worked the stands, 260 spoke on the conference programme, and 340 were accredited media.

Unique visitors is 15,000. Total visits is 15,000 plus 6,800, which is 21,800. Total attendance is 15,000 plus 4,100 plus 260 plus 340, which is 19,700.

The spread between the smallest and largest of those is 6,800 people, or 45 per cent of the visitor figure. Every one of the three is a real count of a real population and every one of them would survive an audit. Which of them lands in the media kit is a marketing decision taken by somebody who may never have read Annex 2. The ratio between visits and visitors deserves its own treatment, and P7 has it.

Putting two shows on the same base

Now go back to the deck with 15,000 against 21,300.

Suppose the second show, asked directly, says its 21,300 is total visits. Suppose it also says the underlying unique visitor count was 14,700, giving 6,600 repeat visits. Restated on the same base as the first show, the comparison is 15,000 against 14,700. The second show is 2 per cent smaller, and the sales director's deck had it 42 per cent bigger.

Work the other direction and the error is just as large. If the first show's 15,000 turns out to be unique visitors and it also has 4,100 exhibitor staff, 260 speakers and 340 media, its total attendance is 19,700. Against a competitor publishing 21,300 total visits, the honest statement is that the two figures count different things and the gap between them is unmeasured until somebody supplies a common base.

The practical procedure is short. Ask each show for the population its number counts. Write both numbers down as unique visitors if that is the base you care about, and mark any figure you could not restate as unavailable. An unavailable figure in your tracker is more useful than a converted guess, because it stops somebody quoting it in six months.

Show length drives more of the gap than audience does

There is a mechanical reason the visits base flatters some shows and not others, and it has nothing to do with how good either show is.

Annex 2 caps repeat visits at one per person per day. A one day show therefore has no repeat visits at all by construction, so its total visits and its unique visitors are the same number. A four day show with an audience that comes back has a visit factor above one, and the longer the run the further the two bases separate.

Put two shows of identical size next to each other. Show C runs for one day and counts 9,000 unique visitors, so its total visits is also 9,000. Show D runs for four days, counts the same 9,000 unique visitors, and sees enough return traffic to reach 14,400 total visits, a visit factor of 1.6. Published on the visits base, Show D looks 5,400 people bigger, which is 60 per cent, and the entire difference is the calendar.

This is why a sector where show lengths vary produces the widest reporting spread. Machine tool and print shows running four or five days sit at one end, single day regional buying events at the other, and the published numbers line up by duration before they line up by anything else. When you see a sector where the biggest published attendance figures all belong to the longest shows, that is the first hypothesis to test.

The counting rule that produces this is also the one that keeps it honest. Because only one visit per person per day may be counted, a show cannot inflate the visit factor by scanning people repeatedly at the door, and the factor stays interpretable as a measure of return behaviour.

What is actually on the certificate?

The audit certificate is the only artefact in this process that records the base as a matter of obligation.

UFI's Auditing Rules set out what a Standard Audit Certificate must contain: the total number of visitors and or visits as defined in the Calculation Standards and Definitions, the breakdown of origin of those visitors or visits into international and national, the total number of exhibitors with the same origin breakdown, and total net exhibition space with separate figures for indoor and outdoor. The certificate must also reference the date of the auditing rules used, and state whether the visitor registration system was manual or electronic.

The UFI Approved Event programme page describes the same three audited quantities: visitor numbers split international and national, net exhibition space, and exhibitor counts split international and national, with visits countable alongside visitors provided this is clearly indicated on the certificate (UFI, 2026). Certificates have to reach UFI headquarters within six months of the event closing.

So the certificate answers the base question by construction. It also answers the second question people forget to ask, which is whether the exhibitor number you are comparing counts contracting companies or every brand in the catalogue. How to get hold of a competitor's certificate and work through it line by line is P28's subject, and the distinction between an audited figure and an organiser's own published one is P27's.

Which comparisons survive the restatement

Some comparisons hold up under all this and some collapse, and it is worth knowing which before you spend a week on the exercise.

Net exhibition space in square metres survives well. It is one of the three audited figures, it has a single published definition, and it does not have a visits variant. Exhibitor counts survive reasonably, once you have established that both sides are counting direct exhibitors. Visitor counts survive only when both shows disclose the base or both are audited.

Anything derived from attendance divided by something else inherits the whole problem and then hides it. Visitors per exhibitor computed from one show's total visits and another's unique visitors will differ by the visit factor alone, and the resulting number looks like an engagement measure while measuring the length of the show. On the worked example above, visitors per exhibitor on a 620 exhibitor floor is 24.2 using unique visitors and 35.2 using total visits, which is the same show and a difference that would change a pricing decision.

Where this stops

Restating both figures on one base tells you which show is larger. It says nothing about which show is better, and the base question can quietly become an excuse to avoid the harder work.

There is also a coverage limit. Not every show carries a certificate. UFI's own auditing rules record 921 events approved by UFI as of March 2021, against a global market that the UFI Global Exhibition Industry Statistics of April 2026 puts at around 32,000 exhibitions a year. For most competitor shows there is no certificate to ask for, and the honest position is that their number is unverified and stays that way in your tracker.

The last limit is that a common base does not fix a bad count. Two shows can agree perfectly that they are both publishing unique visitors and still differ on how completely their entrances are controlled, which is a measurement question rather than a definitional one and needs the audit rather than the vocabulary.

This week, take the three competitor figures you quote most often, and for each one write the population it counts and the document you learned that from. Where you have to write "unknown" in either column, that is the figure to stop repeating, and reading the published counting standards first will tell you which questions to ask to fill it in.