The event ROI calculation an organiser can defend line by line

A group finance analyst asks for the ROI of your show and gives you a spreadsheet template with two cells in it. One says revenue. The other says cost.

The trouble starts about ninety seconds later. Does revenue include the sponsorship that group sold centrally against the portfolio? Does cost include the four people in the audience marketing team who work on this show and two others? Does it include the share of the group's data platform? Every one of those questions has a defensible answer in both directions, and the difference between the two ends is larger than the number most organisers are arguing about.

The way out is to stop looking for a single event ROI calculation and start producing a specific one with its boundary written on the front.

Contribution over direct show cost, and nothing else in the first line

The measure I would put at the top is contribution divided by direct show cost, where direct means the cost would not exist if the show did not run.

That is a deliberately narrow definition and it has one property that makes it worth defending: everything in it is attributable without an allocation rule. No apportionment, no percentage of a shared team, no share of head office. Every line is a cost somebody incurred because this show took place, and if the show were cancelled it would disappear.

Run it. Gross revenue of 12.5 million, made of 9.4 million exhibit space, 2.3 million sponsorship and 0.8 million registration and delegate income. Direct show cost of 5.6 million. Contribution is 12.5 minus 5.6, which is 6.9 million. As a margin on revenue that is 6.9 divided by 12.5, or 55.2 per cent.

The 5.6 million breaks into venue and space rental of 1.98 million, operations and build of 1.18 million, marketing and audience acquisition of 1.28 million, sales commission of 0.42 million, show team staff cost of 0.54 million, and content and speakers of 0.20 million. Those six lines are the entire denominator and each has an owner who can be asked about it. The 1.28 million of audience acquisition is the line most exposed to attendee analytics, since it is the only one whose output is measured in people.

That figure is not unusual for a show of this size. CEIR's Performance Benchmark Playbook, whose second edition was released in July 2026 and covers business to business exhibitions with 200,000 net square feet or more of paid exhibit space, reports an average net profit margin of 55 per cent for shows at that scale, with 80 per cent of organisers reporting profitability and a median gross revenue of 12.5 million dollars. A show landing at 55.2 per cent against that benchmark is squarely normal, which is a useful thing to be able to say out loud before anybody starts optimising.

Which costs belong in the denominator?

Four cost categories cause almost every disagreement, and each has a right answer that depends on what the number is for.

Portfolio marketing spend that touches several shows. Put the directly attributable part in, meaning campaigns, media and list rental bought for this show, and leave the shared brand and platform spend out of the direct figure. If your marketing team cannot separate those, the separation is the project, and it comes before any ROI figure is worth quoting.

Sales team cost where a seller carries three shows. Commission is direct because it is paid on this show's contracts. Base salary is not, unless the seller works on nothing else. Splitting base salary by revenue share produces a number that moves when a different show has a good year, which is the definition of an allocation you should not put in a headline.

Central overhead. Out of the direct figure, always, and reported separately. If group overhead attributable to this show is 1.1 million, then show contribution is 6.9 million and business unit profit after overhead is 5.8 million, giving 46.4 per cent. Both are true, they answer different questions, and publishing only the second makes shows look worse than the CEIR benchmark for reasons that have nothing to do with the show.

Capital and one-off build for a new feature area. Out of the annual ROI, into its own case, because a three-year fit-out charged entirely to one edition destroys the comparability of your series. Measuring a feature area properly is D20's subject and the double counting it warns about is real.

Two numbers, both published, one on top

The practical answer to the group analyst with the two-cell template is to give them two numbers and refuse to give them one.

Show contribution margin, at 55.2 per cent, is the operating measure. It is the one a show director can be held to, because every line in it is under their control or their team's. It is comparable across shows in a portfolio and across years within a show.

Business unit margin after allocated overhead, at 46.4 per cent, is the ownership measure. It is what the show is worth to the company. It is not a fair basis for judging a show team, because a change in the group's overhead allocation policy moves it without anybody doing anything.

Print both, in that order, with the allocation rule for the second stated in one sentence. What kills credibility is publishing one of them under a label that implies the other, and then having the definition discovered by somebody outside the team.

Is ROI a margin or a return?

There is a genuine ambiguity in the word ROI that is worth surfacing, because it is the reason two competent people get different answers from the same file.

Return on investment, read literally, is a return divided by an investment. Contribution over direct cost, which is 6.9 over 5.6, gives 123 per cent, and that is a legitimate ROI in the strict sense: for every pound of direct cost, the show returned 1.23 pounds of contribution on top.

Contribution over revenue, which is 6.9 over 12.5, gives 55.2 per cent, and that is a margin. It tells you how much of the revenue you kept.

Both get called ROI in event conversations, constantly, and they differ by more than a factor of two. The CEIR benchmark quoted above is a net profit margin, so it belongs next to the 55.2 and not next to the 123. Pick one, label it correctly, and put the other in brackets if you want it. What you cannot do is quote your 123 per cent against somebody else's 55 per cent, which happens more often than it should.

The revenue side is not the easy half

Cost gets all the attention in this exercise and revenue quietly causes as much trouble. How that revenue moved, split into volume, rate and mix, is the show revenue bridge analysis in D14, and this section is only about getting the total into the numerator honestly.

Timing is the first issue. Space revenue for the next edition starts arriving during show week, which is the point of on-site rebooking. If your show closes in October and your financial year closes in December, some of what the sales team achieved at the show lands in a different period from the show. Report the show on a show basis, matching all revenue and cost to the edition, and reconcile to the financial period separately. Two sets of books is the right answer here and it is less painful than it sounds.

Barter and contra is the second. Media partnerships, venue trades and sponsor in-kind arrangements have a nominal value that somebody has agreed. Put them in at the value you would have paid in cash, put the same amount in the cost side, and the margin is unaffected while the revenue line stops being fiction. Recording only the revenue half of a contra deal inflates both revenue and margin, and it is the single most common defect I have seen in a show P and L.

Bad debt is the third and it is the one that gets forgotten. A show reporting on contracted revenue rather than collected revenue is reporting a number that will be revised downwards in February, quietly, after the board has seen the first version.

Where this stops

Contribution over direct cost measures the show as an operation. It does not measure the show as an investment, and the gap between those matters when somebody is deciding whether to keep running it.

Nothing in this calculation values the brand, the database, the rights to the market position, or the option to launch an adjacent show off the back of it. A show at 30 per cent contribution margin that owns a market and feeds three other products can be worth more than one at 60 per cent that does not. The ratio will never tell you that, and a portfolio decision made on the ratio alone will be wrong in a predictable direction. Scoring an edition against what it was actually asked to achieve is return on objectives for events, which is D18's.

The second limit is that direct cost is not as clean a category as the definition implies. A venue booked on a three-year deal across two shows, a shared build contract, a technology platform bought for the portfolio and used mainly by one event: each of those is genuinely partly direct and partly not, and any split is a judgement. My rule is to make the judgement once, write it down, and never revisit it mid-series, because a definition that improves every year produces a trend that measures your accounting rather than your show.

The third is that a single point estimate hides how much of this rests on assumptions. Change the overhead rule, the contra valuation and the treatment of sales salary, all defensibly, and the same show reports anywhere from about 44 to about 57 per cent. Running sensitivity analysis on event ROI and showing the range is D21's argument, and it is the right way to present this to a board.

Take your last edition's P and L, mark every cost line as direct or allocated, and total the two. If the allocated column is more than about a fifth of the direct one, your ROI conversation is mostly a conversation about allocation policy, and it should be held with the finance team rather than the show team.