How comp and guest badges appear on an audited attendance certificate

The auditor asks a question that sounds administrative. How do you classify a comp. Three people in the room give three answers, and one of them says it depends on which system you ask, which is the correct answer and the wrong thing to say to an auditor in October.

Comp and guest badges are the least standardised part of most registration files and one of the easiest to get right, because the standard settles inclusion cleanly and leaves you only the classification work.

Do complimentary badges count in an audited visitor figure?

Yes, where the person turns up.

UFI's Calculation Standards and Definitions, Annex 2 to the auditing rules dated June 2021, list who goes into the calculation of visits: persons who purchase their tickets at the event, persons who purchased a ticket before the event and who attend, persons with a complimentary invitation or a reduced tariff ticket and who attend, and persons attending with a season ticket.

The free and reduced categories sit in the middle of that list with the same standing as a paid delegate. Price has no bearing on countability. A trade visitor admitted on an invitation code is a visitor. A student on a reduced tariff is a visitor. A guest admitted on an exhibitor's allocation is a visitor, since they hold a complimentary invitation and they attended.

Two limits follow from the same clause. A comp that was issued and never used contributes nothing, because attendance is the operative condition throughout the list. And a badge issued to somebody on the exclusion list stays excluded whatever it cost, so a free badge handed to an exhibitor's own sales manager is exhibitor staff and sits outside the visitor figure, a boundary covered in where exhibitor staff belong.

Where they appear on the certificate, which is nowhere

Here is the part that changes how you should think about the whole question.

The UFI Standard Audit Certificate, the version UFI publishes in 2026 and marks 08/21, carries net indoor and outdoor exhibition space, exhibitors split national and international, and a visitor block of visitors counted once, plus repeat visits, equalling total visits, each split national and international. There is no field for badge type. No field for ticket price. No field for the free share.

So a show where three quarters of admissions were free and a show where none were produce certificates with identical structure and identical credibility. The classification work you do is invisible on the document, and it still matters, for two reasons.

The first is that classification decides who lands in the visitor figure at all, and the exclusion list is long enough that a sloppy taxonomy moves thousands of people. The second is that your own reporting is the only place the composition is visible, so if you want an exhibitor to trust the headline, the split has to come from you voluntarily.

Work the example. A certified figure of 16,800 visitors, made up of 11,010 standard registered trade visitors, 3,900 admitted on complimentary invitations, 1,250 on one-day passes and 640 hosted buyers. The free population is 3,900 of 16,800, or 23.2 per cent. Hosted buyers are 3.8 per cent. One-day passes are 7.4 per cent.

Every one of those people is legitimately inside the certified total. An exhibitor told only the 16,800 and an exhibitor told the four-line split will make different renewal decisions, and the second one will trust you more the following year. Publishing the split is a choice about what kind of relationship you want, and the certificate neither helps nor hinders it, which is one of the things a certificate is not built to settle.

A badge taxonomy the auditor can follow

Settle this in writing before fieldwork, with one row per badge type and four columns.

The type name, exactly as it appears in the registration platform. Not the marketing name for the campaign.

Who issues it. Organiser, exhibitor allocation, association partner, media partner, speaker programme, venue. The issuing route is what makes a type auditable, because it points at a list somebody maintains.

Its standard bucket. Inside the visitor count, inside total attendance but outside visitors, or outside both.

The evidence. Which extract proves the badge existed and which proves it was used.

A show with five badge types can hold that table in a page. A show with nineteen types, which is common once every sponsor negotiation has added one, cannot, and the excess types are almost always duplicates created because nobody could find the existing one. Consolidating the taxonomy is worth doing for the audit and worth more for everything else.

One rule to apply while consolidating: classify by the person's role at the show and let price fall where it lands. Price is a commercial decision that changes every edition. Role is what the standard's categories describe, and a taxonomy keyed to role survives a pricing change without breaking your series.

The allotment side of exhibitor guest passes, meaning how many each exhibitor should get and what that does to your audience mix, is a separate commercial question covered in the exhibitor guest pass allotment.

Hosted buyers and the group check-in problem

Hosted buyers are visitors under the standard and they are the population most likely to be undercounted, for an operational reason that has nothing to do with definitions.

Hosted programmes run coaches. A coach arrives, a programme coordinator walks to a dedicated desk with 40 lanyards, and one badge gets scanned. Do that 16 times across three days and your access log holds 16 records where 640 people walked in. The auditor reviewing the entrance data finds 16 admissions and no evidence for the rest, and the certified figure is 624 lower than reality.

The fix is unglamorous and it works. Issue a badge per person, scan every badge at the point of entry even when the group is walking in together, and keep the hosted list as its own extract with names, countries and programme dates so the auditor can reconcile the programme against individual admissions. A handheld reader at the coach door costs less than the credibility of a number that is quietly a couple of per cent low.

Keep the hosted list separate from the comp list too. They are different populations with different evidence: a hosted buyer has an application, an approval and usually a travel booking, while a comp has an invitation code and nothing else. Merging them in the extract throws away the stronger evidence.

Why do comp routes quietly lower your international share?

This is the trap that costs organisers something measurable, and it comes from data collection rather than from the standard.

UFI's Calculation Standards and Definitions determine nationality on the basis of the visitor's place of residence, which must be stated on the visitor registration form with the address abroad. The certificate splits every visitor figure into national and international.

Comp routes are exactly the routes that skip the long form. An invitation code redeemed at a kiosk with a name and an email captures no address. A guest pass forwarded by an exhibitor to a customer captures whatever the customer types in twenty seconds. When the extract reaches the auditor, those records have no residence, and a record with no evidence of foreign residence is counted domestic.

Put numbers on it. If 3,900 of your 16,800 visitors came through comp routes and 12 per cent of them were genuinely resident abroad, that is 468 international visitors with no evidence behind them. Against a total of 16,800 those 468 are 2.8 percentage points of international share, and the threshold for a UFI Approved International Event is 5 per cent foreign visits or visitors. Losing 2.8 points to a missing form field is the kind of self-inflicted problem that only shows up once the certificate arrives.

Adding one mandatory country field to every comp route is a small piece of work with a direct effect on a figure that has a label attached to it.

Where this stops

The standard tells you what counts. It cannot tell you whether the person holding the badge is the person it was issued to, and comp routes are where transfer happens most.

An invitation forwarded twice, a guest pass used by a colleague, a one-day pass shared across two days by two people. Each of those produces a record the auditor will accept, because the evidence is genuine and the identity behind it is not testable from a registration file. The scale of it is unknowable from the inside, and the honest mitigation is to make transfer harder at the routes where it matters rather than to model it afterwards.

There is a second limit worth admitting. Classification decisions change the measured composition of your audience without anybody attending differently, so a taxonomy change between editions breaks the series in a way the certificate will not flag. If you consolidate nineteen badge types into six, restate the prior edition on the new taxonomy before you publish a comparison, in the same way you would after any change to how the registration base is defined.

This week, export last edition's badges grouped by type with a count against each, and put the standard's three buckets in a column beside them. Any type you cannot place, and any type with fewer than a couple of hundred badges against it, is a candidate for deletion before the next edition opens. That single page is what your auditor will work from, and it is the cheapest entry point into the wider set of published definitions.