Skip to content

Show calendar comparability across editions when a date moves

Standards and researchUpdated 2026-08-238 min read

In short

Show calendar comparability across editions rests on the interval between editions rather than the calendar year they land in. A move from March to October leaves nineteen months between editions, and reporting that gap as year on year growth can turn a 21 per cent decline into a 25 per cent rise on the same two numbers.

The portfolio roll-up for 2026 shows net exhibition space down 6.9 per cent across eight shows. The finance lead wants to know which show fell over. The answer is that none of them did, and one of them simply is not in the 2026 column, because it moved from December to January and its next edition lands in 2027.

Show calendar comparability across editions is the discipline of fixing the comparison window before computing anything, and it matters most in the years when a date moves, which is exactly when nobody has the appetite for a methodology conversation.

Editions are the unit, calendar years are an accident

A show is an edition. It opens, it runs, it closes, and the next one happens whenever the venue, the sector calendar and the competitive set allow. The calendar year is a reporting convention layered on top, and it fits until a date moves.

The standards side of the industry already works this way. UFI's auditing rules, published in June 2021, attach audit obligations to editions and never to years. At a minimum every other edition of an approved event must be audited, except for events running once every three years or less often, where each edition must be audited. UFI (2021) even works the example: an event taking place once every two years must be audited once every four years.

The submission clock runs on the same basis. Certificates must reach UFI within six months of completion of the relevant event, and the standard certificate carries the instruction on its signature page in plain terms, asking for the audit at the latest six months after the end of show. A show ending in November and a show ending in February have deadlines seven months apart, and neither of them is a financial year end.

The nineteen month year, worked through

Take a show that ran in March 2025 and then moved to October 2026. March to October across two calendar years is 19 months.

The March 2025 edition finished with 19,800 registrations. The October 2026 edition finished with 24,700. Divide one by the other and registrations are up 24.7 per cent, which is the number that will reach a slide if nobody intervenes.

Now hold the window constant. Multiply 24,700 by twelve nineteenths and you get 15,600 registrations per twelve months of selling. Against 19,800 for the prior edition, that is a fall of 21.2 per cent.

The same two numbers, read across the same two editions, produce plus 24.7 per cent and minus 21.2 per cent. The sign flips, and the magnitude is large in both directions. Every downstream decision that touched that figure, budget, headcount, hall commitment, was made on whichever version happened to be computed.

The annualised figure is the more defensible of the two for a growth question, and it carries an assumption worth naming, which the next section does. What it definitely is not is optional. Publishing 24.7 per cent with no interval attached asserts something about a twelve month period that never happened.

Why does a calendar year roll up drop a show entirely?

Because grouping by calendar year asks each show to supply exactly one edition per year, and a moved show can supply two or none.

Three cases turn up in practice. A show moving from December 2025 to January 2027 has no edition inside calendar 2026 at all. A show moving from January 2026 to December 2026 has two editions inside calendar 2026, eleven months apart. A show moving from October 2025 to March 2026 has one edition in each calendar year with only five months between them.

The first case is the dangerous one because the loss is invisible. Work it through on a portfolio of eight shows holding 214,000 square metres of net exhibition space in calendar 2025. One of those, at 18,400 square metres, runs in December 2025 and next runs in January 2027. Calendar 2026 therefore contains seven shows.

Those seven held 195,600 square metres in 2025 and hold 199,300 in 2026, which is growth of 1.9 per cent. The calendar year comparison puts 199,300 against 214,000, a fall of 6.9 per cent.

A portfolio that grew nearly two per cent reports a seven per cent decline, and the entire gap is a scheduling decision taken eighteen months earlier by somebody who was solving a venue availability problem. The fix is to compare each show's latest edition against its own previous edition and sum the changes, which produces the 1.9 per cent, and to state separately that one show contributed to 2025 and will next contribute in 2027.

The two-editions-in-one-year case inverts the same error. A portfolio can post double digit growth because one show ran twice, and the following year it will post a decline of similar size for the same reason.

Annualising, and what the adjustment assumes

Multiplying by twelve over nineteen assumes demand accrues at a constant rate across the interval, and it does not.

Registration for a trade show is front-loaded toward the final weeks, exhibitor space sells against a rebooking window anchored to the previous edition, and both curves are shaped by the show's own calendar rather than by elapsed time. Nineteen months between editions does not give you nineteen twelfths of a normal year's demand. It gives you one normal selling cycle plus seven months during which most of the audience had nothing to register for.

That makes the annualised figure a floor-and-ceiling exercise rather than a point estimate. Compute it three ways and quote the range. Straight annualisation at twelve nineteenths gives 15,600. Treating the extra seven months as contributing nothing gives the raw 24,700 against 19,800, or plus 24.7 per cent. Treating them as contributing at half rate gives an effective window of 15.5 months, so 24,700 times twelve over 15.5, which is 19,123, or a fall of 3.4 per cent against 19,800.

Minus 21.2, minus 3.4, or plus 24.7. Stating all three with their assumptions is more useful than picking one, and the honest headline is that the edition is roughly flat to down with a wide band around it. Where the underlying registration curve needs proper rebasing instead of a scaling factor, the method is the day-relative alignment problem, which belongs with A6.

Which of the three you lead with should follow from the question being asked. A hall commitment decision wants the floor, because over-committing space on the back of a 24.7 per cent figure that turned out to be a nineteen month artefact is an expensive mistake to unwind. A marketing effectiveness review wants the middle scenario, since the campaign did run for longer than usual and some of the extra registrations are genuinely attributable to it. A board pack wants all three and a sentence saying the series is interrupted.

What does the audit certificate already record for you?

More than most organisers realise, which makes it a useful source of truth when the internal spreadsheet has drifted.

The UFI standard audit certificate, in the version UFI currently publishes, asks the organiser on its first page for the dates of the exhibition being audited, from and to. It then asks for the dates of the next session if known, the number of times the exhibition has been organised previously, and the dates of the most recent previous editions, each with a box indicating whether that edition was audited.

That is the interval, the cadence and the audit history on one page, filled in by the organiser and countersigned by the auditor. A certificate series therefore carries the date moves in it whether or not anyone maintained a separate record, and it is the artefact to reconcile against when two internal systems disagree about when the 2023 edition ran. Distinguishing that signed record from a figure that only ever appeared in a press release is the prior skill.

Keep one more column of your own next to it: the reason the date moved. Venue availability, a sector calendar clash, a merger with a co-located show, a public holiday. The reason predicts whether the interval effect is one-off or the start of a new cadence, and it is the first thing anyone will ask when the annualised number goes in a board pack.

Where this stops

None of this recovers demand that was never expressed. A nineteen month gap loses a cohort of buyers who needed a show in the missing autumn and went somewhere else, and no divisor brings them back into the file.

Annualisation also breaks down entirely for shows whose audience is tied to a fixed external cycle: a harvest, a fiscal year end, an academic calendar, a regulatory deadline. Move such a show by seven months and you have changed which buyers can attend, which makes the two editions different products rather than the same product measured over different windows. In that situation the correct report says the series is broken, names the edition where it broke, and starts a new baseline, which is the same discipline as versioning the definitions themselves so older editions stay readable.

There is also a limit on how much of this a reader will absorb. A pack carrying three annualisation scenarios per show is unreadable. Pick the scenario you believe, put the interval in months next to the percentage, and keep the other two in the appendix for the person who asks.

Do the small version this week. Add one column to the show-level table in your portfolio report, headed months since previous edition, and fill it in for every show. Any row that reads something other than twelve is a row where the year-on-year figure next to it needs a sentence, and finding those rows is the whole of the measurement standards job here.

Questions people ask about show calendar comparability across editions

What happens to year on year comparisons when a show changes its dates?
The comparison window changes with it. Two editions nineteen months apart accumulated nineteen months of demand, and reporting the difference between them as annual growth overstates it. Annualising by multiplying the later figure by twelve over nineteen gives a rough correction, and stating the interval in months next to the percentage is the minimum honest disclosure.
Can a portfolio roll up lose a show when dates move?
Yes, and silently. A show that runs in December one year and January two years later has no edition inside the intervening calendar year, so a roll up grouped by calendar year contains one fewer show without any line indicating it. The portfolio total falls with no operational cause behind the fall.
Does an audit certificate record the interval between editions?
It records the components. The UFI standard audit certificate asks the organiser for the dates of the edition being audited, the dates of the next session if known, the number of times the exhibition has been organised previously, and the dates of the most recent previous editions with a yes or no on whether each was audited. The interval is a subtraction away.

Related reading

All standards and research articles