Trade fair ROI is (attributed gross profit − fully loaded cost) ÷ fully loaded cost, measured over an attribution window you fix before the fair opens and counted only on leads that pass a written qualification test. Three decisions produce the number, and none of them is arithmetic: what counts as a qualified lead, how long you keep counting, and whether staff time is inside the cost. Use gross profit, never revenue — revenue makes a 12% margin look like a triumph. Almost every argument about a fair ROI figure is an argument about definitions, so write the definitions down first and freeze them.
Updated: 2026-08-29 · Reading time: 26 min · Guide
Two exhibitors leave the same hall on the same evening. One reports a 140% return, the other reports a loss. They had adjacent stands, similar products and almost identical order books. The difference is not performance. One of them counted revenue against invoices; the other counted gross profit against invoices plus thirty-six person-days of internal time. Neither is lying and only one of them can make a decision with the number they produced.
This guide builds the metric term by term, then runs one case from the first invoice to the closing balance, then puts the metrics that routinely mislead next to the ones that replace them. Every figure below belongs to a single illustrative company and is labelled as such. None of it is a benchmark, because there is no useful industry benchmark for this — your gross margin and your sales cycle decide what a good number looks like, and both live in your own systems. If you want to build the cost side first, the trade fair cost calculator does that part in the browser and hands you the denominator this page needs.
The arithmetic is the least interesting part. Written out, trade fair ROI is: attributed gross profit, minus fully loaded cost, divided by fully loaded cost. Multiply by 100 and you have a percentage. Every dispute you will ever have about that percentage is a dispute about two words — attributed and fully loaded — and neither of them is defined by the formula. They are defined by you, in writing, before the fair opens, or they are defined afterwards by whoever wants the number to come out a particular way.
The metric is built from four terms. Term 1, attributed gross profit: revenue from deals the fair can be shown to have started, multiplied by your gross margin, counted only inside the attribution window. The common error is using revenue instead of gross profit, which inflates the result by the inverse of the margin — at a 40% margin, revenue reports two and a half times the true return. Term 2, fully loaded cost: every invoice plus every internal person-day at a loaded day rate, including preparation and follow-up. The common error is counting invoices only, which typically leaves a meaningful slice of the real cost outside the calculation. Term 3, the attribution window: the fixed period, starting on the first day of the live event, during which a closed deal still counts as fair-sourced. The common error is choosing the window after seeing the results. Term 4, the qualification test: the written rule that decides which contacts enter the count at all. The common error is treating badge scans as leads. Assemble the four and ROI equals attributed gross profit minus fully loaded cost, divided by fully loaded cost.
Use gross profit rather than revenue and the reason becomes obvious the first time you run both. A company with a 12% gross margin that reports revenue against cost will produce a triumphant number for an edition that lost money. The inflation factor is exactly the inverse of the margin: at 40% you overstate by 2.5×, at 20% by 5×, at 12% by more than 8×. Nothing about the fair changed; only the word in the numerator did.
These three go in a one-page document, signed off by sales and marketing together, before the first invoice is paid. Not because governance is enjoyable, but because each of them can swing the final percentage by more than the fair itself can, and a definition written after the fact will always be written by the person who most wants a particular answer.
A badge scan is a record that a person walked within range of a reader. A business card is a record that someone was polite. Neither is a lead. Write a test with five criteria and apply it identically to every contact, at the stand, on the same day, while the conversation is still fresh:
Three of five is a common threshold and it is a choice, not a rule; what matters is that the threshold is the same for every edition you intend to compare. Expect the funnel from scans to qualified leads to be steep. We are not going to hand you an industry figure for it, because we have no survey to attribute and a number from a page like this one is exactly the kind of borrowed authority this guide argues against. What we will say is our own position: we have not seen the ratio come out better than five scans per qualified lead, and the worked example below lands at 7.6. If yours is much better, suspect the test before you celebrate the traffic. Record the reason for every rejection, because the rejection reasons are the most actionable data the fair produces — "wrong geography" is a targeting problem, "no budget" is a positioning problem, and they call for opposite responses.
The window is the single most abused parameter in fair measurement. Report at 30 days and a business with a nine-month sales cycle will conclude that fairs do not work. Report with no window at all and every deal signed for the rest of the decade drifts into the numerator. The workable rule: take the median sales cycle for the product line you are exhibiting — median, not mean, because one 30-month enterprise deal will drag a mean somewhere useless — multiply by 1.5, round up to a whole month, and write that number down before the fair.
Long windows create a real management problem: nobody will wait fourteen months for a verdict, and the pressure to shorten the window is enormous. Resist it, and solve the impatience with leading indicators instead — cost per qualified lead is available two weeks after the fair, weighted pipeline value within a month. Both are honest and neither is ROI. Calling weighted pipeline "ROI" in a board deck is the most common way an otherwise careful measurement turns into a fiction.
Fully loaded means invoices plus internal time. The invoices are easy and they are also the part everyone already counts: space, stand build, freight, flights, hotels, per diems, print, samples, shipping, insurance. The part that goes missing is the person-days, and they go missing because no one sends you a bill for them. Count preparation days, build and travel days, show days, and — the most forgotten of all — follow-up days in the three weeks after the fair, when someone has to actually work through the leads. Value them at a loaded day rate (gross salary plus employer contributions plus overhead, divided by working days), not at the net salary. As a planning heuristic, invoice-only costing leaves somewhere in the range of 15–35% of true cost outside the number; that is a range offered for sanity-checking, not a measured figure, and your own timesheets will beat it. The cost calculator walks the full list.
Two boundary questions come up every time. Sunk assets: a reusable modular stand bought three years ago should enter at its depreciated per-edition share, not at zero and not at full purchase price. Shared costs: if the same trip covers two fairs and four customer visits, split travel by days, and write the split rule down once so the next edition is comparable.
What follows is one illustrative company: a component supplier with an average order value of €18,000 and a gross margin of 42%, exhibiting on a 24 m² stand at a four-day international fair, with a median sales cycle of seven months. Every number is an assumption chosen to make the arithmetic legible. Substitute yours; the structure is the transferable part, the figures are not.
Cost side, fully loaded. Space and stand build: €21,000. Freight, samples and logistics: €4,800. Travel and accommodation for four people over four nights: €7,200. Marketing, print and giveaways: €2,500. Internal time, 36 person-days at a €300 loaded day rate — 12 days of preparation, 16 days of build, travel and show, 8 days of follow-up: €10,800. Fully loaded cost: €46,300, of which €10,800, about 23%, would have been invisible in an invoice-only count. Lead side. Badge scans and cards collected: 312. After removing duplicates, existing customers, students, competitors and job seekers: 174 new contacts. Passing the five-criterion qualification test: 41 qualified leads, a ratio of 7.6 scans per qualified lead. Cost per qualified lead: €1,129. Return side, measured at the close of a 12-month window fixed before the fair, because the seven-month median cycle times 1.5 gives 10.5 months, which the rule rounds up to 11 — and this company then extended it to 12 so the window would close with its reporting year, and wrote that extension into the same document as the rule, before the fair. Extending is defensible; shortening after you have seen the number is not. Closed-won inside the window: 9 of the 41 qualified leads, a 22% close rate. Orders placed by those 9 customers inside the window: 11. Attributed revenue: €198,000. Attributed gross profit at a 42% margin: €83,160. Net gain: €83,160 minus €46,300 equals €36,860. ROI: €36,860 divided by €46,300 equals 79.6%. Payback point, where cumulative attributed gross profit crossed the €46,300 cost line: month eight of the window.
Notice which single change would have destroyed this number and which would have flattered it. Report revenue instead of gross profit and the same edition returns 328% — a figure that is arithmetically correct and completely useless, because the company cannot spend revenue. Drop internal time from the denominator and ROI rises from 79.6% to 134%, which is how invoice-only costing quietly rewards editions that consume enormous amounts of staff attention. Shorten the window to 90 days and perhaps two of the nine deals had closed, turning a profitable edition into a reported loss of roughly two thirds of its cost.
Now run the identical structure for digital participation, where travel, freight and stand build are not reduced but absent. The cost side becomes the participation fee plus content production — photography, specification sheets, a demonstration video, translation — at roughly €1,800, plus internal time of about 19 person-days (8 preparation, 5 across the live period, 6 follow-up) at the same €300 rate, or €5,700. That is €7,500 of loaded cost before any fee at all.
Rather than typing a participation fee here — fees are set per edition and published on each fair's own page, such as Health onMESSE — the more useful move is to invert the formula and let it tell you what a fee may be worth. Assume the digital edition produces fewer but better-briefed conversations: 18 qualified leads instead of 41, at the same 22% close rate, giving 4 closed-won customers and 5 orders, €90,000 of attributed revenue and €37,800 of attributed gross profit. Break-even sits where fee plus €7,500 equals €37,800, so any all-in fee below roughly €30,300 returns something. To match the physical edition's 79.6%, the total cost must not exceed €37,800 ÷ 1.796 = €21,046, which leaves about €13,500 for the fee. That is the number to carry into a pricing page, and it is derived from your own margin and close rate rather than from anybody's brochure.
A physical stand stops existing when the hall is cleared. A published exhibitor profile does not: it stays indexed, stays searchable inside the platform, and keeps receiving enquiries between editions. That is a genuine structural difference and it is also the easiest place in the whole calculation to cheat, because the denominator is paid once while the numerator keeps accruing. Anyone can produce an impressive ROI by counting cost for four days and profit for four years.
The honest treatment is two reported numbers with one clearly stated measurement date. Continue the illustrative case: the digital edition produced 18 qualified leads during the live period, and the profile produced 11 more over the following eleven months. At the same 22% close rate the 29 total qualified leads give 6 closed-won customers and 8 orders — €144,000 of attributed revenue and €60,480 of gross profit against the same €21,046 of cost. Trailing-twelve-month ROI is 187%. Edition ROI at the close of the live period, on the 18 leads it actually generated, was 79.6%. Both are true, they answer different questions, and reporting only the larger one without its measurement date is where credibility is lost.
This is also the point where format comparison becomes fair. A physical stand and a year-round profile are not the same product, and a single ROI percentage flattens that. Report edition ROI for both, then report cost per qualified lead per month for the year-round one, and the comparison stops depending on how generous you were with the calendar. The mechanics of that year-round layer are set out in what a digital trade fair actually is.
Every metric below is in active use, appears in post-show reports, and is wrong in a specific and repairable way. The replacement is usually not harder to collect — it is just less flattering, which is why the original survived.
Instead of badge scans or cards collected, report qualified leads under a written test — a scan measures footfall past a reader, not interest, and the scan-to-qualified ratio is often five to one or worse. Instead of revenue attributed to the fair, report gross profit attributed to the fair, because revenue overstates return by the inverse of the margin and a company cannot spend revenue. Instead of cost taken from invoices only, report fully loaded cost including internal person-days, since preparation, travel and follow-up days are real cost that nobody bills you for. Instead of pipeline generated reported as ROI, report weighted pipeline as a leading indicator and keep ROI for closed-won, because unweighted pipeline assumes a 100% close rate. Instead of asking which channel touched the deal last, ask the buyer directly where the relationship started and store that answer as the fair-sourced flag; last-touch attribution systematically credits whichever channel is closest to the signature. Instead of counting meetings held, count meetings that produced a scheduled next step, since a meeting with no follow-on is a cost, not an outcome. Instead of comparing this edition to last edition on raw lead count, compare cost per qualified lead over identical window lengths, because a longer window always produces a better number. Instead of counting orders from existing customers who visited the stand, count only new relationships and report existing customer meetings as retention activity, otherwise the fair takes credit for revenue that was arriving anyway. Instead of reporting visitors to your online profile, report enquiries started from the profile, because traffic is not a lead in either format.
One number cannot carry a decision that spans a year. Report four, on their own schedules, and the argument about whether to exhibit again becomes short.
Cost per qualified lead is the workhorse. It compares a physical edition with a digital one, this year with last year, and a fair with paid search, as long as the qualification test is genuinely identical across everything being compared. If it is not identical, you are comparing definitions. Once you have the ratio, the editions currently open is where you find the fairs to apply it to, and platform pricing explains how participation is structured before you look at any individual fair's fee.
A negative ROI is a finding, not a verdict. Four different failures produce it and each has a different response; three of them are cheap to fix and one of them means stop. Find yours by walking the funnel backwards.
Türkçe