Lead capture coverage rate shows how many exhibitors captured nothing at all

The benchmarking slide in your post-show deck says median leads per exhibitor: 84. Somebody on the commercial team will ask where that came from, and the honest answer is that it came from the exhibitors who produced a lead file.

There were 640 exhibiting companies on that floor. The scan table contains rows for 470 of them. The median is a median of 470 stands, and the 170 companies that appear nowhere in it have not been counted as zero. They have been dropped. The size of that hole is what a lead capture coverage rate measures.

Those 170 stands will still fill in the post-show survey. They will still take a renewal call in six weeks. What they will not have is a single piece of evidence about their own show, which means the conversation about next year runs entirely on how the four days felt to whoever was standing there.

Defining the rate so it survives an argument

Lead capture coverage rate is the number of exhibiting companies with at least one recorded capture, divided by the number of exhibiting companies on the floor. It is the first number I would put in front of any exhibitor analytics discussion, because every other figure in the pack is conditional on it.

Both halves of that need pinning down before anyone reports it.

The numerator is companies. An exhibitor with four licences across two stands who scanned once is one covered exhibitor. Counting licences instead inflates the rate for exactly the large exhibitors who least need help.

The denominator is exhibiting companies with contracted space at the edition, taken from the contract system rather than the floorplan. Co-exhibitors, pavilion participants and shared stands are where this gets decided, and the decision matters more than which way you decide it. A national pavilion with 22 companies under one contract can be one exhibitor or 22, and the coverage rate moves by three points depending on the answer. Write the rule down, put it in the footnote, keep it for five years.

Withdrawals and no-shows come out of the denominator. A company that cancelled in September never had the chance to scan and including them makes your show look worse than it is.

On our floor: 470 divided by 640 is 73.4 per cent. One exhibitor in four generated no evidence of their own show.

Why are the zeros the interesting part?

Every instinct in a reporting team pulls toward the exhibitors with data. They have distributions, percentiles, category comparisons and year on year movement. The 170 with nothing have a blank row, and a blank row is hard to put on a slide.

CEIR's 2026 Marketing Spend Decision Report puts B2B exhibiting at 40.8 per cent of exhibitor marketing budgets, the largest single channel in the mix, and reports that management evaluates exhibition return primarily on sales metrics, with lead volume and post-show closed deals ranking highest. Hold those two findings together. The stand that captured nothing is spending its employer's largest marketing line and will be assessed on a number it cannot produce.

What happens next is predictable. The marketing manager at that company has to write something in a wrap report. In the absence of a lead file they write an impression, and impressions written under budget pressure are rarely generous.

The Explori and UFI 2025 Channel Insights report, drawing on more than 3,000 events surveyed since 2017, tracks overall satisfaction, likelihood of return, net promoter score and event importance as the four exhibitor measures organisers already collect. Coverage rate sits underneath all four as a supply problem. You cannot influence how an exhibitor rates a show they have no record of.

Splitting the 170

An aggregate coverage rate tells you the size of the gap. Splitting it tells you what to do, and the splits that pay are boring ones.

By whether a licence was ordered. This is the first cut and it separates two entirely different problems. Suppose 553 of the 640 exhibitors ordered at least one licence and 470 of those scanned. That means 87 companies never ordered a scanner and 83 ordered one and never used it. The first group needs a sales and communication fix at order time. The second is an onboarding failure, and the gap between ordering and scanning has its own measure and its own owner.

By stand size. Compute coverage separately for stands under 18 square metres and stands above it. If coverage is 61 per cent among small stands and 88 per cent among large ones, your silent quarter is concentrated in the part of the floor with the least internal marketing resource and the highest churn risk.

By tenure. First edition exhibitors against returning ones. New stands have nobody who remembers that the scanner has to be collected from the service desk.

By category. If one product category has coverage 20 points below the floor, check whether the category is clustered in a hall with a connectivity problem before you conclude anything about the exhibitors.

By country. International exhibitors dealing with roaming, app store restrictions and a manual in a second language drop out of digital capture at a visibly higher rate.

Run those five cuts on 170 companies and you usually find that three quarters of them fall into two buckets, and those two buckets have different owners inside your business.

Then track the rate as a series rather than a snapshot. Coverage moving from 68 per cent to 73.4 per cent across two editions is about 35 more exhibitors holding evidence, and that is a sentence a show director can take to a board without any claim about what those exhibitors did with it. Coverage moving the other way while lead volume rises is the pattern to watch for, because it means your total is being carried by fewer, larger stands while the spread of lead volume across the floor gets more skewed than it already is.

The number to act on during the show

Coverage rate computed in December is a diagnosis. Computed at 14:00 on day one it is a task list.

The query is trivial: exhibiting companies with contracted space, left joined to scans recorded so far, filtered to those with no rows. On a 640 exhibitor show that list will be around 300 names at lunchtime on day one and it will fall through the afternoon.

What matters is what you do with it. The stands that are still on that list at close of day one are the visit list for the first hour of day two, and the visit takes about four minutes. Half of them have a device sitting in a box under the counter. A few have an app that nobody has entered the access code into. One or two are genuinely running a meetings only stand and want to be left alone, which is a legitimate answer and worth recording so nobody visits them again.

I would rather have a floor team clearing that list on day two than a perfect scorecard in January. The scorecard describes a show that has already happened. The day two list changes the show while it is running, and it improves the same scorecard as a side effect.

What does coverage do to every other number you publish?

Once you know coverage, every benchmark in the pack needs a decision about the missing 170, and the decision must be stated.

Two defensible options. Report benchmarks over scanning exhibitors only, labelled as such, so the median of 84 becomes "median among the 470 exhibitors with a lead file". Or report over all exhibitors, treating non-scanners as zero, which drops the median hard: with 170 zeros added, the median of 640 values sits at the 320th value, and on most floors that lands somewhere near 41.

Both are honest. Neither is honest if unlabelled. I would publish the first as the headline and the second in the footnote, because an exhibitor comparing themselves to a peer group wants the peer group to be stands that were playing the same game, and the show director looking at portfolio health wants the number that includes the silence.

What you cannot do is publish 84 with no denominator note and let a board read it as a per exhibitor average. The same caution applies downstream, since a report only counts once somebody opens it, and whether exhibitors ever read the pack is measurable in its own right.

The same discipline applies to any total you quote. If the pack says the show generated 57,000 leads, that total came from 470 companies and the covering sentence should say so. Divided by the 640 exhibiting companies it reads as 89 leads each. Divided by the 470 that scanned it reads as 121. Both divisions are correct, they differ by a third, and the reader will pick whichever one you put nearest the headline.

Where this stops

Coverage rate measures capture, and capture is not participation.

A stand running a hosted buyer programme with 30 scheduled meetings and no scanner will show as a zero and may have had the best show on the floor. So will an exhibitor whose staff collected business cards, which is still common and, on the numbers, still works for some categories.

There is also a measurement asymmetry you should say out loud. A company using its own third party capture app rather than the show system produces no rows in your table at all. Cvent's exhibitor guide describes buying badge kit access as a separate purchase so third party lead solutions can read the badge, and any exhibitor going that route disappears from your coverage figure entirely while capturing perfectly well. Ask about it in the exhibitor manual with a single yes or no question and you can subtract those companies from the denominator instead of guessing.

The last limit is that coverage says nothing about quality. A show with 95 per cent coverage where every stand scanned twelve badges casually is in worse shape than a show at 73 per cent where the scanning stands worked properly.

Take your last edition, join the exhibitor contract list to the scan table, and count the companies with no rows. Then order that list by contracted value. The names at the top are exhibitors paying you the most and holding the least evidence, and they are the renewal conversations to prepare for first.