International visitor classification in audits and the residence rule behind it

Two weeks after the show closes, the auditor sends a query. Your international visitor figure is 5.2 per cent and they want to know which field it came from. Somebody on the registration team pulls the export and finds two candidate columns: the address the visitor typed for themselves, and the country of the company they said they worked for. The two produce different answers, and nobody wrote down which one the last edition used.

International visitor classification in audits has one rule, and it is narrower than most registration teams assume. It also differs from the rule applied to exhibitors on the same certificate, which is the part that catches people out.

Two populations, two rules, one certificate

UFI's auditing rules, published in June 2021 and still the version the current standard audit certificate references, set the visitor test on residence:

The "Nationality" is determined on the basis of a visitor's place of residence and must be stated on visitor registration forms with mention of the address abroad. Foreign or international visits are those visits by visitors of places of residence other than the country in which the event takes place.

The exhibitor test in the same document runs on a different field entirely. UFI (2021) states that "the nationality of an exhibitor is determined by the address of the exhibiting company/organization as written on the contract with the organizer".

So a certificate carries an international split for visitors computed from where people live, and an international split for exhibitors computed from where a company signed. Those are both defensible, and they are not the same rule, and the certificate does nothing to tell a reader that two different tests were applied to the two halves of the page.

The host country is fixed by where the event happens. For hybrid and physical events UFI (2021) makes the country hosting the physical event the host country, and for digital-only events it names the country supplying the most visitors as the domestic one, with everyone else counted as international.

Then there is a footnote at the bottom of the visitor rule that decides more outcomes than the rule itself. EU countries are counted as individual nations.

What does place of residence mean on a registration form?

It means the address the visitor gave, and it means the registration form has to ask for one.

UFI's audit process rules are specific about the evidence: to justify the number and percentage of international visitors, each international visitor should be registered. That is a per-person requirement, and it is the reason a show that collects country only for the people who tick a box cannot certify an international share.

There is one documented escape. Where individual registration is genuinely impossible, UFI (2021) allows an organiser to seek authorisation to establish the foreign percentage through sampling, carried out by a market research company and certified by an independent audit company, with the sampling methodology named on the certificate. For a public show it is acceptable to count only foreign professional visitors against total professional visits. Both routes leave a visible trace on the certificate, so a reader can tell a counted figure from a surveyed one.

Other national bodies land in the same place by their own wording. UNIMEV in France and BDO in Portugal both treat a visitor whose address is in another European or third country as foreign, in the definitions UFI printed in its Euro Fair Statistics for 2022, published November 2023. CENTREX establishes foreign visitors through pre-registration, registration cards carrying evidence of foreign origin, or representative visitor surveys.

Why does the 5 per cent threshold make the rule choice expensive?

Because the label depends on it, and the two candidate rules straddle it.

Take a show in the Netherlands with 24,000 unique visitors. Under residence, 1,248 gave an address outside the Netherlands, which is 5.2 per cent, just clear of the 5 per cent floor UFI sets for foreign visits or visitors in an approved international event.

Now suppose the registration team had mapped the employer's country instead. Of those 1,248, some number are people who live in Belgium or Germany and work for Dutch-registered employers, say 190 of them, and they would drop out. Coming the other way, Dutch residents working for foreign-registered employers would be added, say 60. The count becomes 1,248 minus 190 plus 60, which is 1,118, and 1,118 over 24,000 is 4.7 per cent. Below the line.

Same file, same show, same people through the door. One column choice, and the event holds or loses its international designation. UFI (2021) does allow a grace period of sorts, since an event that stops meeting the criteria is reclassified only after two consecutive audit reports, which gives an organiser one edition to fix the field rather than the label.

The EU footnote is worth its own arithmetic. Of the 1,248 foreign residents, suppose 1,040 live elsewhere in the EU and 208 live outside it. Treating the EU as one bloc would give 208 over 24,000, or 0.9 per cent, and the show would be nowhere near the threshold. Counting EU states individually, as UFI requires, gives 5.2 per cent. That single footnote is worth 4.3 percentage points to this show, and it is the difference between a label and no label.

The exhibitor rule runs on the contract address

The exhibitor side has its own arithmetic and its own escape hatch, and the escape hatch is where multinationals live.

A show has 612 exhibitors. Contracts signed by entities registered outside the host country number 55, which is 9.0 per cent, below the 10 per cent floor UFI sets for direct foreign and multinational exhibitors combined.

Sitting alongside them are 23 exhibitors that are local subsidiaries of foreign groups. On the contract address rule every one of those is domestic, because a local subsidiary signs with a local registered address. UFI (2021) then provides the route out: where the address does not represent the true nationality, the organiser may accept the nationality the exhibitor gives on a self declaration, and where doubt exists "the exhibiting company's/organization's headquarters should provide a declaration stating that the subsidiary is acting on its behalf". A declaration by the organiser alone is explicitly refused.

If 14 of those 23 supply the declaration, the international exhibitor count becomes 69 out of 612, or 11.3 per cent, and the show clears the exhibitor threshold. The other nine stay domestic because nobody chased the paperwork.

Note the shape of that. The classification is decided by a document the exhibitor has to produce, on a deadline that runs inside your audit window, and the organiser cannot decide it unilaterally. UFI's own criteria treat direct foreign exhibitors and multinational exhibitors as two named categories that should both be indicated clearly on the certificate, so the split is visible to anyone reading a competitor's certificate properly.

What the auditor will actually ask for

Four things, in roughly this order.

• The field. Which column in the registration export holds the residence address, and whether it was mandatory.
• The completeness. How many registrations have a usable country value, and what you did with the blanks. Blanks counted as domestic will inflate your domestic share and depress your international one.
• The exhibitor declarations. Copies of any self declarations and headquarters letters supporting a subsidiary counted as international.
• The consistency with last time. Whether the same rule produced last edition's figure, since a jump in international share with no change in audience is usually a change in field mapping.

That fourth question is the one worth pre-empting. If the rule changed, say so before the auditor finds it, and restate the prior edition on the new basis in your own reporting so the trend line stays readable.

The blanks question deserves a worked number attached to it rather than a policy. On a file of 24,000 registrations, 1,900 rows with no usable country value is 7.9 per cent, larger than the entire international share the certificate is trying to report. Of the 22,100 rows that did answer, 1,248 are international. Assign every blank to domestic and the reported share is 1,248 over 24,000, or 5.2 per cent. Exclude the blanks from both numerator and denominator and it is 1,248 over 22,100, or 5.6 per cent. Impute them at the answered rate and you add 107 international registrations for a count of 1,355, which lands on that same 5.6 per cent while changing the number of people the certificate claims. Three treatments, two rates, and the one you picked belongs in writing next to the number.

Where this stops

Residence is self reported and nobody checks it. The rule is auditable only as far as the form captured an address, and a visitor who types their office address because the form asked for a business address has been classified by their employer's location while the certificate says otherwise.

That gap is largest exactly where it matters most, among senior international buyers who register through an assistant or a hosted buyer programme, and it runs in one direction: toward under-counting international residence for people whose registration was completed by somebody else. There is no correction for it that an auditor would accept, because the correction would be an estimate and the certificate carries counts.

The honest response is to make the form ask for a home country separately from a company country, keep both, and report the international share on residence while holding the employer column for your own commercial analysis. Two fields cost one line on the form and settle the question permanently, and they also let you compute both versions when the national rules you are comparing against differ or when a different body signed the other show's figures.

Before the next registration build freezes, open the form and check whether residence country is a mandatory field with a country list rather than a free text box. If it is optional, make it mandatory for this edition, and record the date you changed it so the measurement standards note in next year's report can explain the step in the series.