Sponsorship sell through rate is the number your rate card is hiding

The slide said sponsorship sell through was 71 per cent. Somebody in the room asked what the denominator was, and the honest answer turned out to be that nobody had written down how many units were offered, so the 71 per cent had been reconstructed afterwards from the assets that happened to sell.

Sponsorship sell through rate is worth computing properly, and computing it properly means two ratios per asset rather than one percentage per show. Units sold over units offered tells you about demand. Revenue invoiced over the card value of what sold tells you about price. Collapse them into one number and you lose the ability to tell those two apart, which is the only reason to calculate either.

Two ratios, and the four cases they produce

Call the first one unit sell through and the second one value capture. Every asset lands in one of four places, and each place has a different instruction attached.

High units, high capture. The asset cleared and nobody argued about the price. That is an underpriced asset, and the fix is to raise the card price next edition and watch what happens to the unit ratio.

High units, low capture. The asset cleared, but only after the sales team gave away a third of the card price on every deal. Your card number is fiction. Reset it near the achieved price and defend it, because the discount is what is doing the selling and everybody involved already knows it.

Low units, high capture. The people who bought paid what you asked, and there were fewer of them than you offered. That is a supply problem rather than a price problem, and it belongs with the sizing decision rather than with the rate card.

Low units, low capture. Nobody wanted it and the few who did paid under the odds. Either the asset needs proof it has never had, or it comes off the card.

The reason this matters is that the middle two cases look identical in a single blended percentage and require opposite actions. One says add price. The other says remove inventory.

Working both ratios across eight assets

Take a register of eight assets, with illustrative figures for one edition.

Work two of the rows by hand so the method is clear. The aisle signs offered 46 faces and sold 12, which is 26.1 per cent on units. The twelve that sold carried a card value of 12 times 850, or 10,200, against 9,180 invoiced, which is 90 per cent on value. So exhibitors who wanted an aisle sign paid close to the asking price, and there were only twelve of them against 46 faces on offer.

The hall entrance banners are the opposite. Four offered, four sold, so 100 per cent on units. Card value of 4 times 6,500 is 26,000, invoiced 17,680, so 68 per cent on value. The real clearing price for that banner is 17,680 divided by 4, which is 4,420, and the card says 6,500.

Why does the show level number mislead?

Add the whole register up and you get several defensible numbers that say different things.

Units offered total 61 and units sold total 23, so show-level unit sell through is 37.7 per cent. Card value of the units that sold comes to 99,600 against 83,360 invoiced, so value capture is 83.7 per cent. Total invoiced against the full card value of everything offered, which is 160,100, comes to 52.1 per cent.

Three numbers, one dataset, all correct. Whichever one appears on the board slide will be the one that flatters or alarms most, and none of them tells anyone what to do on Monday.

The unit ratio is worse than useless here, because a single asset dominates it. Remove the 46 aisle sign faces and the remaining fifteen units include eleven sold, which is 73.3 per cent. The edition did not change. One line item with a high unit count and weak demand pulled the whole show's headline down by 35 points, and the reason is arithmetic rather than commercial.

Any asset you have chosen to slice into many small units will do this. That is a good argument for computing the ratio per asset, and a bad argument for slicing assets differently to make the number look better, which is a temptation worth naming out loud before somebody acts on it.

How does the denominator quietly get rewritten?

Units offered has to be recorded before the selling cycle opens. If it is reconstructed afterwards, it will be reconstructed from what sold, and the ratio becomes meaningless in a way that is very hard to spot later.

The failure looks like this. Sales opens with 46 aisle faces on the plan. In week fourteen, operations cuts hall 4's signage because of a rigging clash, so eight faces disappear. Nobody updates the sales system. At the end of the edition somebody counts the faces that were physically installed, finds 38, and computes 12 over 38, which is 31.6 per cent rather than 26.1 per cent.

Neither number is a lie. They answer different questions, and only one of them was fixed before the outcome was known. Freeze units offered at the moment the card is published, keep an audit trail of any change with a date and a reason, and report against the frozen figure with the changes disclosed underneath.

The register produced by a proper inventory audit is where that frozen count should live, because it is the only place that holds unit counts for assets which sold nothing at all. The charging lounge row above contributes zero revenue and one unit of denominator, and it disappears from any calculation built from invoices.

What the market says about your strongest assets

CEIR's 2023 Exhibit and Sponsorship Sales Approaches report asked organisers which offerings they found effective, and the branded items category came out at the top of the whole list. Lanyards were rated effective by 84 per cent of organisers, ahead of exhibition giveaways at 77 per cent, registration badges at 74 per cent and hotel keycards at 73 per cent. An email blast to the registration list scored 83 per cent, printed banners on premise 83 per cent, show signage 81 per cent, and mobile app ads 72 per cent.

Set that against your own two ratios. If your lanyard is clearing at 100 per cent of units and 100 per cent of card value, the broader market view and your own data agree that it is the strongest asset in the building, and your card price is the only thing failing to say so.

An asset that the wider market rates highly and your own value capture rates poorly is a sales problem rather than a product problem. An asset that scores badly in both places should probably go.

Treat the CEIR ranking as a prior and your own two ratios as the evidence. The ranking is what organisers across the industry say works, gathered in one survey, and it will be wrong for particular shows in particular verticals. A hotel keycard sponsorship rated effective by 73 per cent of organisers means very little at a show where 70 per cent of attendees commute in daily and never see a hotel room. Your own value capture on that asset, computed over two editions, beats the survey every time. The survey is most useful for the assets you have never tried, where you have no evidence of your own and a starting view is better than a guess.

Where an asset has never been offered, both ratios are undefined, and the question becomes how to price something with no history at all.

Where this stops

Both ratios are silent about what the sponsor got.

An asset can clear 100 per cent of units at 100 per cent of card value for three consecutive editions and still be the thing that loses you the account, because the sponsor bought it, paid for it, and never found out whether it worked. Sell through measures your commercial process. It measures nothing about delivery, and a reporting layer that pushes both to the same exhibitor and sponsor view should keep the two clearly labelled.

The second limit is that value capture assumes a card price worth measuring against. For bespoke deals assembled from scratch there is no card line to divide by, and forcing one into existence for the sake of the ratio produces a number that describes the accounting rather than the market.

Take your last edition's contract list and add one column: how many units of that asset were on the card when selling opened. Where you cannot answer from records rather than memory, that asset has no sell through rate this year, and fixing it costs one field in the contract system before the next cycle starts.