Exhibitor pipeline reporting lag and how to report before the deals close

Rebooking on plenty of shows opens the Tuesday after breakdown and closes about six weeks later. The exhibitor pipeline report, the one with opportunity counts on it, becomes worth reading somewhere around month seven.

That five month gap is exhibitor pipeline reporting lag, and it puts the renewal conversation roughly 150 days before the evidence that would settle it exists. Your sales lead quotes a figure computed at day 30. The exhibitor quotes a different figure, also computed at day 30, and neither describes what the show produced, because at day 30 almost nothing has finished happening.

This is a scheduling problem wearing an attribution problem's clothes. The fix costs one extra report template and a naming convention.

Two clocks that were never going to agree

The organiser's clock is set by the floor plan. Space has to be sold, the best positions go first, and every week an account sits undecided is a week the neighbouring stand could have taken it. Six weeks after breakdown is not an arbitrary deadline. It is when the inventory decision has to be made.

The exhibitor's clock is set by their buyers. A badge scanned on the Wednesday becomes a call the following week, a qualification conversation the week after, an opportunity record once procurement gets involved, and a closed deal after whatever their cycle happens to be. None of those steps care about your floor plan.

On a B2B floor the two clocks sit about five months apart, and further on capital equipment. No change to the attribution method closes that gap. How long you count opportunities for is a separate decision, and choosing the window against the exhibitor's own sales cycle belongs to E15. What follows is about when you publish.

What is a single post-show figure actually claiming?

A post-show report carrying one pipeline number makes an implicit claim, which is that the number is finished. Readers take it that way, because that is how a report reads.

It is a partial count of a process still running, taken on whatever date your reporting cycle happened to end, and that date is usually set by when the marketing team wanted the deck rather than by anything about the exhibitor's business.

Statistical agencies gave up pretending about this decades ago. The US Bureau of Economic Analysis publishes an advance estimate of quarterly GDP about a month after the quarter ends, a second estimate a month after that, and revises again as fuller source data arrives. Its own explanation of the practice says the advance estimate rests on source data available for only two months, with assumptions filling the rest. BEA's worked illustration is the fourth quarter of 2008, where the advance estimate showed a 3.8 per cent decrease and the second estimate showed 6.2 per cent. The first figure was not an error. It was the honest reading of what had been collected by then, published with the word advance attached.

An exhibitor pipeline count at day 30 is an advance estimate of exactly that kind. The difference is that nobody labels it.

Three legs, each answering its own question

The version I would ship is the same report three times, on a published calendar, with each edition answering a question the previous one could not.

Day 7, lead delivery. Did the exhibitor receive a complete, usable file, and when did they open it? This comes entirely out of systems the organiser owns: the capture platform, the export log, the portal access log. No exhibitor cooperation required, no CRM involved, and it is available for every stand on the floor rather than for the minority who answer emails.

Day 30, follow-up. What share of delivered leads carry a logged follow-up activity, and how many opportunity records exist so far? This needs the exhibitor's CRM, so coverage drops. It measures the exhibitor's sales operation, which is the useful reading, and it says close to nothing about the show.

Day 180, opportunity count. How many opportunities now exist against scanned contacts, and at what stage? Getting there at all depends on the join from badge scans through CRM contacts to opportunities already working, which is E14's subject. This is the first leg with any claim to describe the show's contribution, and it lands long after the rebooking decision that people wanted it for.

None of the three is a revenue claim. Attaching money to a scan needs its own written rule, and the verbs an organiser is entitled to use about revenue are E17's argument.

The day 180 leg being commercially late is the point. You are not going to have it in time, so stop building the renewal argument on a bad early proxy for it and start building the renewal argument on the two legs you can actually stand up.

Staging also reads better to the person holding the budget. CEIR's 2026 Marketing Spend Decision Report, whose research ran in the latter part of 2025, describes exhibitors going into 2026 with broadly flat participation plans, with 47 per cent expecting to take part in the same number of exhibitions and 28 per cent expecting to add shows. A buyer in that position is comparing your show against alternatives on the strength of the evidence you hand them. A report that arrives three times, says what it can support each time, and revises in the open is a stronger artefact in that comparison than one confident figure delivered once and quietly wrong.

Why does the day 7 leg earn its place?

A latency measure sounds like housekeeping until you look at what the delay costs.

Oldroyd, McElheran and Elkington reported in Harvard Business Review in 2011, under the title The Short Life of Online Sales Leads, that most companies were not responding to online customer queries anywhere near fast enough.

Web queries are not badge scans, and no multiple published about one transfers to the other. What transfers is the shape. The value of a lead decays, the decay starts on the show floor, and every day between close and file delivery is spent. A day 7 report showing that four of your exhibitors had not downloaded their file yet is a finding you can act on while it still matters, and it costs an export timestamp and a portal log to produce.

Working the revision on 310 leads

Take one exhibitor with 310 unique leads from a single edition.

Day 7: file generated on the Monday after close, exhibitor first downloaded it on the Tuesday. Of the 310 rows, 286 carry a usable email address, which is 286 divided by 310, or 92.3 per cent. That is the lead delivery report, and it is complete.

Day 30: 214 of the 310 contacts carry at least one logged follow-up activity in the exhibitor's CRM. That is 214 over 310, or 69.0 per cent. Opportunities created against a scanned contact: 22.

Day 180: opportunities created against a scanned contact, cumulative: 58.

Now the figure that should govern how you talk about the day 30 number. Twenty-two is 22 divided by 58, or 37.9 per cent of what the same edition eventually showed. Put the other way, the count multiplied by 58 over 22, which is 2.64 times, between the two reports.

An organiser who published 22 as the show's pipeline contribution published a figure low by a factor of nearly three, and did it in the document the exhibitor's finance director read before deciding about next year. The arithmetic is not subtle and neither is the consequence.

Projecting the later count from the earlier one

The revision ratio is not a nuisance. It is a measurement, and once you have two editions of it you can use it.

Suppose that for the prior edition you have both a day 30 and a day 180 opportunity count from 26 exhibitors. Compute the day 30 count as a share of the day 180 count for each, and read the quartiles off those 26 values. Say they come out at a lower quartile of 0.24, a median of 0.35, and an upper quartile of 0.47.

Apply that to this exhibitor's 22. The central projection is 22 divided by 0.35, which is 62.9, so about 63. The band runs from 22 divided by 0.47, which is 46.8, up to 22 divided by 0.24, which is 91.7. So 47 to 92.

The line I would publish at day 30 reads: 22 opportunities recorded at day 30, projecting 47 to 92 by day 180 on this show's own ratios from the prior edition, with a central estimate near 63. The eventual 58 sits inside that band and below the central estimate, which is what an honest interval is supposed to do a fair share of the time.

Two things make this defensible. The ratio came off your own floor rather than a published benchmark, and the interval is published rather than collapsed into a midpoint. An exhibitor arguing with a range has to argue with a method. An exhibitor arguing with a single number just picks a different single number.

Where this stops

The day 180 leg depends on exhibitors handing back CRM state half a year after a show most of them have stopped thinking about, and a large share will not.

That makes the response set self-selecting in a direction you can predict. Exhibitors whose pipeline went well are more willing to send the extract than exhibitors whose pipeline went nowhere, so the day 180 population skews good, and the median revision ratio you compute from it belongs to exhibitors with functioning CRM hygiene and a reason to talk to you. The day 7 and day 30 legs do not carry that bias in the same way, because the first comes wholly from your own systems and the second at least covers everyone who agreed to connect a CRM at all. Report the number of exhibitors behind each leg on the leg itself, every time.

The projection carries a second assumption worth saying out loud. Dividing by last edition's median share assumes the conversion path is behaving the same way this year, and it will not be if the show moved date, changed hall layout in a way that shifted the visitor mix, or if the economy turned between editions. Two editions of history is thin. Four is workable. Below two, publish the day 30 count with its date attached and skip the projection entirely.

There is also a temptation, once the staged report exists, to quietly overwrite the earlier legs so the report always shows the latest figure. Do not. Keep every leg addressable with its own date, print the prior value next to the revised one, and let people see the revision. A number that changed and shows its old value is credible. A number that changed silently gets found out once, and then none of your exhibitor reporting is trusted.

Pull last edition's lead export and whatever CRM extracts you already hold, and for the handful of exhibitors with data at two different dates, compute the day 30 opportunity count as a share of the latest one. The median of those few values is your first revision ratio, and it will tell you within an afternoon how wrong your current single-figure report is.