The exhibitor scorecard metrics worth putting in front of a stand manager

The post-show exhibitor report had thirty-one tiles on it. A stand manager screenshotted one of them, cropped it, and sent it to her commercial director with two lines of text.

The tile she chose was her lead count against the same figure from the previous edition. Everything else on that page, the sentiment scores and the dwell estimates and the engagement index, went unread by the only person whose behaviour the report was supposed to change.

That is the real design brief for exhibitor scorecard metrics, and most scorecards fail it by being too long. A stand manager has about forty seconds and a screenshot. What survives the crop is the whole product.

Four tests a line has to pass

Before arguing about which measures belong, it helps to have a rule for admitting them, because otherwise every internal stakeholder adds one and you end up at thirty-one tiles.

Measured, not asserted. The value has to come out of a system the organiser operates: the capture platform, the registration database, the contract or floor plan system. If it comes from a survey, an exhibitor's own claim, or a vendor's black box, it belongs somewhere other than a per-stand scorecard.

Available for everyone. A measure present for 40 per cent of the floor cannot sit on a template sent to 100 per cent of it. Blank tiles get read as failure, and a measure that only exists for the exhibitors who cooperated is a measure of cooperation.

Actionable before the next edition. The stand manager should be able to name what she would do differently. Scanner coverage passes this test easily. A sentiment index does not.

Stable across editions. If the definition moves, the year-on-year comparison is a comparison of definitions. This test quietly kills most clever measures, because clever measures get improved.

What does UFI actually audit, and why is that list so short?

There is a working precedent for this discipline at show level, and it is worth reading before designing anything at stand level.

UFI's auditing rules for approved events require independent verification of exactly three statistics: the number of visitors split into international and domestic, the net exhibition space, and the number of exhibitors split the same way. Visits, meaning visitors plus repeat entries, may also be counted, and where they are, UFI's published rules are explicit that this must be clearly indicated on the standard audit certificate. Audits have to be carried out by a UFI member auditor, a third party trained in the UFI rules, or a qualified financial auditor, and the organiser's internal accountant is ruled out on independence grounds.

Three numbers, one flagged optional extra, and a rule about who is allowed to compute them. An industry body that spent decades on this problem landed on a very short audited core with everything else labelled as something else. An exhibitor scorecard built to the same shape is defensible in a way that a thirty-one tile dashboard never is.

Which six measures belong on the card?

Here is the set I would ship, and the reason each one earns a place.

Unique leads. Distinct badges captured by that exhibitor across the edition. The count the exhibitor came for, the count they will screenshot, and the one measure that has to be right before anything else matters.

Total scans. The raw capture event count, printed beside unique leads rather than instead of it. Two exhibitors with the same unique count and very different scan counts ran very different stands, and the ratio between the two has its own reading, which is E12's.

Grade mix. The distribution of lead grades the exhibitor recorded, using the show's own scale. Always printed with the grading completion rate beside it, for reasons E6 sets out properly. Without that companion figure the mix describes whichever slice of the stand's captures somebody bothered to grade.

Scanning users active. How many of the licensed scanner users actually recorded a capture. This is the one line on the card that a stand manager can fix with a briefing and no budget, which makes it the most useful thing on the page even though it is the least interesting.

Category percentile. Where the stand's unique lead count sits inside its own product category at this show. Percentiles turn a number nobody can interpret into a position anybody can. How the category cell gets built, and what happens when it goes thin, is E24's subject and it matters more than it sounds.

Prior edition comparison. The same stand's unique lead count last time, and the change. This is the line most likely to be misread, and the next section is about why.

Six lines, all computed from systems the organiser runs, all available for every exhibitor who scanned anything. Nothing here needs the exhibitor to answer an email.

What the prior edition comparison hides

Take a stand with 812 unique leads this edition against 690 last time. The change is 122, and 122 divided by 690 is 17.7 per cent. On the card that reads as a good year.

Now put the show underneath it. Verified attendance went from 8,400 to 10,100, a rise of 1,700, which is 1,700 divided by 8,400, or 20.2 per cent.

Compute the stand's share of the audience in each edition. Last time, 690 divided by 8,400 is 8.21 per cent. This time, 812 divided by 10,100 is 8.04 per cent. Index the second against the first: 8.04 divided by 8.21 is 0.979, so 98 on a base of 100.

The stand captured 17.7 per cent more people from an audience that grew 20.2 per cent. Both readings are arithmetically correct and they point in opposite directions. The raw change says the stand improved. The share of audience says it slipped by about 2 per cent while the show did the work.

So the prior edition line needs two figures on it, the raw change and the share-of-audience index, or it will flatter every exhibitor in a growing year and punish every exhibitor in a flat one. Index against verified attendance rather than registrations, since registrations move with campaign spend and the stand never met the people who did not turn up.

What to keep off the card

Revenue. An organiser does not hold it, cannot audit it, and asserting it on the exhibitor's behalf is how a rebooking conversation goes wrong, which E17 covers.

Survey-derived measures, on a per-stand card. The 2025 Channel Insights report, produced by Explori for UFI with support from SISO and drawing on 3,245 B2B events since 2017, tracks four exhibitor measures: overall satisfaction, net promoter score, likelihood of return, and event importance. Those are the right measures for the organiser's own reporting and they belong at show level. On an individual stand's card they fail the second test above, because they exist only for exhibitors who answered, and they fail the third, because a stand manager cannot act on her own satisfaction score.

Anything modelled without the inputs shown. Engagement indices and propensity scores are fine internal tools and they read as magic on a page you cannot audit. If a modelled figure has to appear, print the components next to it, which is the argument E22 makes about combining several percentiles into one score.

The same Explori and UFI report gives organisers a takeaway worth holding against your own template: there is an expectation that organisers improve return on investment measurement beyond simple lead counts. That is an argument for the other five lines existing, and not an argument for thirty-one.

The page I would actually send

One page. Six lines in a single column, each with the raw value, its comparator, and the date the figure was computed. The category percentile carries the cell size next to it, so a stand manager can see whether she is 62nd percentile of 38 companies or of 6. The scanner line carries both numbers, so 4 of 9 licensed users active reads as what it is rather than as 44 per cent of something unspecified.

Underneath, one paragraph in plain sentences naming the two things that would move the card most for this stand, generated from the lines themselves rather than written by hand for 600 exhibitors. For the stand above, that paragraph is about the five dormant scanner licences and the share-of-audience index, and it would be the same paragraph for every stand with the same shape of card.

Then a methodology note, on the same page, saying which system each figure came from and what the deduplication rule was. Nobody reads it until somebody disputes a number, and on that day it is the only part of your exhibitor analytics that matters.

Where this stops

Every line on this card measures capture, and capture is only one of the reasons companies exhibit.

The Explori and UFI 2025 data makes that concrete. Comparing post-pandemic events against pre-2020 ones, exhibitors reported better achievement on 8 of the 14 objectives tracked, and the single objective that went backwards was finding new distributors, agents or partners. A stand exhibiting to sign two distributors will meet its objective with four conversations, record four scans, sit near the bottom of every percentile on your card, and rebook happily. Treat that card as a performance verdict in a renewal call and you will have an argument with a satisfied customer.

The second limit is coverage of the capture system itself. Exhibitors running their own universal capture kit across every show they attend appear on your floor with an empty scorecard, and they are frequently the most sophisticated exhibitors you have. Give that case its own flag on the card, reading no organiser capture, because an empty scorecard and a genuine zero look identical in a database and mean opposite things.

Print your current exhibitor report and count the tiles. For each one, write down the table and column it came from and the share of exhibitors it is populated for. Anything you cannot source in a minute, or that is blank for more than a fifth of the floor, comes off before the next edition.