Building a rebooking conversion funnel from appointment held to deposit received

The sales director says rebooking is down four points and that the offer was wrong. The pricing lead says the offer was fine and the reps did not work the list. Both have been arguing for an hour, and neither has a number that separates the two claims, because the only number anyone measured was the one at the end.

A single conversion rate cannot tell you where a sale was lost. A rebooking conversion funnel can, and every organiser I have worked with already has four of its five numbers sitting in a CRM with nobody counting them.

Five stages, each with a timestamp

The stages that matter for onsite rebooking are the ones where a specific human action either happened or did not, and where the system records when.

A target list, fixed before the show opens and frozen. An appointment booked, meaning a slot in a rep's diary with an account against it. An appointment held, meaning somebody from that account turned up. A contract signed, for named space at an agreed rate. A deposit received, meaning cash cleared against the invoice.

The freeze on the target list matters more than any of the others. If the list can grow during show week, your top-of-funnel base moves while you are measuring conversion through it, and every stage rate below becomes uninterpretable. Fix the list at close of business the day before doors open, store it, and treat anything added later as a separate walk-up cohort with its own rates. Walk-ups convert differently from targets and mixing them hides both.

One edition, run through properly

Same industrial show, 640 exhibiting companies on the floorplan, all 640 on the target list.

End to end, 388 deposits over 640 targets is 60.6 per cent. The headline onsite rebooking rate for the same show, 402 signings over 640 companies, is 62.8 per cent, and it is the only figure here that cannot tell anybody what to fix.

Now look at where the volume actually goes. 120 accounts never booked an appointment at all. Another 90 booked one and did not turn up. Between them those two stages account for 210 of the 238 accounts that did not sign. Everything from the moment somebody sat down in the rebooking booth converted at 93.5 per cent, and the deposit stage held 96.5.

The reps did not lose these sales in the conversation. Most of them never got the conversation.

Where is a point of conversion cheapest to buy?

Take the two obvious interventions and price them in signings.

Improve the pitch. Suppose coaching, a better floorplan display and a sharper offer lift held-to-signed from 93.5 to 96.0 per cent. Applied to 430 held appointments, that is 430 times 0.025, or about 11 extra contracts.

Improve the diary. Suppose earlier outreach, a booking link in the exhibitor portal and two more people on pre-show calls lift the booking rate from 81.3 to 89.1 per cent, so 570 appointments instead of 520. Hold the downstream rates where they are, 82.7 per cent held and 93.5 per cent signed, and 570 times 0.827 times 0.935 gives about 441 contracts. That is 39 more than 402.

Thirty-nine against eleven, from two interventions of roughly comparable cost. At an illustrative average first-year contract of 6,500 in whatever currency you sell in, the diary work is worth about 253,500 and the pitch work about 71,500. I have watched teams spend an entire pre-show cycle on the pitch because the pitch is the part sales enjoys discussing, while the booking rate sat unmeasured.

Held-to-signed is also the stage with the least headroom, which is a general property of funnels worth internalising. A stage already converting at 93.5 per cent has 6.5 points available in total. A stage at 81.3 has 18.7.

Why is the leak worth money right now?

The margin structure of exhibitions makes conversion leakage expensive in a way that is easy to underrate. CEIR's Performance Benchmark Playbook, second edition, published in 2026, reports that large B2B exhibitions of 200,000 net square feet or more post an average net profit margin of 55 per cent, with 80 per cent of large-show organisers reporting profitability and median gross revenue of 12.5 million dollars. When your venue, marketing and staffing are largely committed before the floorplan fills, a marginal booth sale drops most of its revenue through to profit, and a marginal booth left empty takes most of it away.

That is happening while organisers are getting more cautious about earnings. UFI's 37th Global Exhibition Barometer, concluded in June 2026 across 466 companies in 59 countries and regions, found 19 per cent of companies expecting operating profit to rise by more than 10 per cent, with a majority at 54 per cent expecting stable profits. The 36th edition of the same survey, published in January 2026 from 378 companies in 57 countries and regions, had 33 per cent forecasting a rise above 10 per cent for that year. The share expecting a strong year went from 33 to 19 inside six months.

In a flat-profit year, 39 contracts recovered from a diary problem is a real contribution against a fixed cost base, and it does not require anyone to raise a rate.

Defining held, because this is where the data goes soft

Booked and signed are easy. Held is the stage that decides whether the funnel is trustworthy, and it is the one nobody has a clean field for.

The weak version is a rep marking an appointment complete in the CRM, which they do at the end of a long day for the ones they remember. The version I would build is an arrival event: a badge scan at the rebooking suite, a check-in on a tablet, or a stand visit logged against the account. Any of those produces a timestamp created by the meeting happening, which is the property you need.

If you only have the rep's flag, say so on the slide and treat the held rate as an estimate with a known bias, because reps under-record no-shows more often than they under-record meetings. You can size the bias cheaply. Take one show day, have a person on the door count arrivals against the diary, and compare that count with what the CRM says for the same day. If the CRM claims 94 held and the door counted 81, your held rate is overstated by about 16 per cent and every downstream rate is understated by the same mechanism.

Do that once and you will know whether the funnel is telling you the truth for the rest of the year.

Reading the funnel across editions

One edition's funnel tells you where this show leaked. A series tells you whether anything you did worked.

Store the five counts per edition per show, with the target list definition and the held definition recorded next to them. That series is the spine of any renewal intelligence reporting worth keeping. Then read the stage rates as a series and watch for the pattern where an intervention moves one stage and quietly damages the next. Pushing hard on booking volume with a broad email will lift the booked rate and drop the held rate, because appointments made by people with no authority get cancelled. If booked goes from 81.3 to 89.1 and held falls from 82.7 to 74.0, then 570 times 0.740 times 0.935 is about 394 contracts. You bought 50 appointments and gave back 8 contracts, which is the opposite sign to what the arithmetic above assumed.

The compound number is the one to govern with. Booked times held times signed times deposit, computed per edition, is the only figure that cannot be improved by moving a loss from one stage to another.

Where this stops

A funnel measures a process, and demand sits entirely outside it. The measurement will happily tell you that your sales team is performing while your show shrinks underneath them.

Every rate above is conditional on the target list. If 90 of your 640 exhibitors had already decided in the spring that this category no longer justifies a stand, those 90 will not book an appointment, and your booking rate will fall, and the funnel will point at the diary. The diary is not the problem in that case. The funnel cannot tell the difference between an appointment nobody chased and an appointment nobody wanted, because both produce the same absence of a record.

The partial fix is to work the non-bookers as their own cohort after the show and code why each one declined, using a fixed set of reasons rather than free text. If the codes cluster on scheduling, the diary reading holds. If they cluster on budget or on category, you have a demand problem that no amount of appointment-setting will touch, and the funnel has done its job by ruling itself out.

The other limit is honest attribution across the two halves of the cycle. A share of the accounts who did not sign in the hall will sign in the weeks after the hall empties, and counting them as funnel losses overstates the leak. Keep the onsite funnel and the post-show conversion as separate measurements, and reconcile them at the end of the cycle against the final book. That reconciliation also has to net off the signings that do not survive to the first payment deadline, which is a rate of its own and is not what the deposit stage above measures.

Take your last edition's CRM export this week and count four numbers: accounts on the frozen target list, accounts with an appointment record, accounts with an arrival record of any kind, and accounts with a signed contract dated during show week. If you cannot produce the third of those four from data that was created by the meeting rather than typed afterwards, that is the instrumentation to fix before the next show opens.