Defining an exhibit space sell through rate you can compare year to year

A show director puts a slide up in a portfolio review. Sell through, 95 per cent, against 90 per cent the year before. Someone in finance asks how much of the improvement is the hall change. The room goes quiet, because the honest answer is most of it, and nobody present can prove that either way inside a fortnight.

An exhibit space sell through rate is the first number anyone asks for and one of the least stable numbers an exhibition business produces. It is a fraction, and almost all the trouble sits in the bottom half of it.

The denominator moves and nobody logs it

Everyone agrees the numerator is net square feet sold. Contracts signed, area committed, add it up. There is some argument about whether to count space given away and whether to count a contract that has not paid, and those are real questions with boring answers.

The denominator is where the number goes soft. Sellable net square feet is a design output. It comes out of the floorplan, which changes between the day sales opens and the day the show builds. Aisles get widened after a venue walk. A feature area gets added because sponsorship sold something. A hall gets dropped because the forecast came in light. Registration moves from the concourse into the hall, taking 4,000 square feet with it.

Each of those is a reasonable operational decision. Together they mean the denominator on the final plan has almost nothing to do with the denominator the sales team was working against in January, and if you compute sell through at the end from the final map, you have measured a fraction whose bottom half was quietly edited to suit the top half.

That is not a hypothetical failure. It is the ordinary way a floorplan behaves, and the incentive runs one direction. Nobody rebuilds the map to add unsold inventory in May.

Fix the denominator at plan freeze

The fix is procedural and takes about an hour a year. Pick a date, before sales opens, when the floorplan is signed off well enough to sell against. Call it plan freeze. Record sellable net square feet on that date, per hall, and never touch it again for that edition.

Every subsequent measurement of sell through uses the frozen number. If you widen an aisle in April and lose 3,000 square feet, sell through does not improve. You have sold the same area against the same plan and given up inventory, and both facts stay visible.

The frozen figure also gives you a second, more interesting metric almost for free. Track final sellable against frozen sellable and you have plan drift, which tells you how much inventory the operating team removes from the sales team every year. Two per cent is normal. Eleven per cent is a conversation.

Store it with the show, alongside the rest of your exhibitor analytics and not in a spreadsheet on someone's laptop: edition, hall, frozen sellable net square feet, freeze date, and the name of the person who signed it. The last field matters more than it looks, because the only defence against a denominator argument in November is a record of who agreed to it in January.

The arithmetic, and what one wall does to it

Take a hall planned at 140,000 sellable net square feet at freeze. Sales closes the edition at 126,000 net square feet sold. Sell through is 126,000 divided by 140,000, which is 90.0 per cent.

Now suppose 8,000 square feet at the rear never sold, and in April the operations team moved a hall divider to close off that bay, using it for storage and staff catering. Final sellable is 132,000. Compute sell through from the final map and you get 126,000 divided by 132,000, which is 95.5 per cent.

Five and a half points of improvement, no additional contract, no additional revenue. The show sold exactly the same amount of space in both worlds. One version of the report says the floor performed well, the other says it left 14,000 square feet on the table, and the difference is a decision about where to put a wall. Putting a figure on what that unsold space costs you is F40's subject.

Run that comparison on your own last three editions and you will usually find at least one year where the gap is worth more than a point. It is worth doing once, precisely because the result embarrasses the number rather than the people, which makes the fix easy to land.

What should count as sold?

Having frozen the denominator, be equally boring about the numerator, because the same drift happens there in slower motion.

Four categories cause every argument. Contracted but unpaid space, which is sold until it is not, and roughly three to seven per cent of it will fall out depending on your deposit terms. Barter and contra space, given to media partners and associations, which occupies inventory and produces no space revenue. Association-run pavilions where one buyer takes a block and subdivides it, which is one contract and forty exhibitors. Feature areas built by an exhibitor at their own cost in exchange for space.

Pick a rule for each, write it down in one sentence, and apply it to five editions of history so the series is consistent. My preference is to count contracted space including unpaid, count barter as sold with a flag, count a pavilion block once at its full area, and treat exhibitor-built features as sold. The specific choices matter far less than the fact that they are the same choices next year.

Where you want the alternative view, report sell through on paid contracts as a second line rather than replacing the first. Two clean series beat one negotiated series.

Anchor it to something outside your building

Internal consistency is necessary and does not tell you whether 90 per cent is good. For that you need an external series measured the same way for a long time.

The CEIR Index Dashboard, released by IAEE and CEIR in April 2025, tracks four metrics across 14 industry sectors: net square feet of exhibit space sold, professional attendance, number of exhibiting companies, and gross revenue, with history back to 2000 and forecasts running to 2027. The metric it uses for space is net square feet sold, which is the same numerator you just defined. That is the point of naming it. If your internal numerator matches the index's numerator, your show's space performance can be read against its sector rather than against your own optimism.

CEIR's Q4 2024 Index results, released in March 2025, put the total index at 95.6, a 6.0 per cent year on year improvement and 4.4 per cent below pre-pandemic, with 37.0 per cent of completed 2024 events surpassing their pre-pandemic performance, up eight points on 2023. A show reporting 90 per cent sell through in a sector that is flat is a different story from the same 90 per cent in a sector running hot.

The forward view is worth holding in mind too. UFI's 35th Global Exhibition Barometer, published in July 2025 from 386 companies across 58 countries and regions, found 34 per cent expecting rented space in their own country to grow by more than 5 per cent in 2025, 48 per cent expecting it to stay inside plus or minus 5 per cent, and 12 per cent expecting a fall of more than 5 per cent. Flat is the modal outcome. A sell through number that jumps six points in one edition is more likely to be a definition change than a market change, and the first thing to check is the denominator.

Why report sell through as a curve?

Sell through as a single end-of-cycle number is a grade. It arrives too late to do anything with.

The version that earns its place on a weekly report is sell through against days to open, plotted against the same curve for the previous two editions, using the frozen denominator for each. At 180 days out you were at 61 per cent last edition and you are at 57 per cent now. That is four points of pacing gap, which on a 140,000 square foot hall is 5,600 square feet, which at a 30 dollar rate is 168,000 dollars of space revenue currently absent.

The curve also exposes the shape of your sales year, which is usually more lopsided than anyone admits. Most shows do half their space in two windows: the onsite rebooking period and the six weeks after the floorplan opens to the general market. If the first window underperforms, the curve tells you in week two and you have nine months. The closing number tells you in November and you have nothing.

Keep the plot in absolute square feet as well as percentage. Percentages hide the size of the hole, and the sales team argues about square feet. Converting the gap into money is a rate question and belongs to F37.

Where this stops

A frozen denominator makes sell through comparable across editions of the same show in the same hall. It does nothing to make it comparable across two different venues, and treating it as though it does will produce a portfolio ranking that mostly measures architecture. A hall with a heavy service corridor and a large permanent registration area will show a lower sell through of gross area and can be running at full occupancy of everything it is legally allowed to sell. That normalisation problem is real, it has its own pair of ratios, and it belongs to F23.

The second limit is more awkward. Sell through is bounded above by a number you chose. An organiser who freezes a conservative plan, holding back a bay for a feature area that may never be built, will post better sell through than a colleague who planned every sellable foot into the map. Both are honest. Only one looks good. The defence is to publish plan drift beside sell through so a reader can see how much inventory was removed after freeze, and to keep an eye on shows whose frozen denominator falls year on year while their hall does not.

Neither limit is a reason to stop reporting the metric. Both are reasons to report it with the frozen sellable figure printed next to it, so a reader can reconstruct the fraction without asking.

Start this week by opening the last three editions of one show and writing down four numbers per edition: sellable net square feet at plan freeze, sellable net square feet on the final map, net square feet sold, and the freeze date. If you cannot find the first or the fourth for any edition, that is the gap, and it costs an afternoon to close for the edition you are selling now.