Registration abandonment does not end when someone leaves a form unfinished. It also appears when registrants cancel, fail to check in, or sign up too late to attend reliably. The 2026 PheedLoop Event Data Lab reports show median no-show rates ranging from about 17% to 34% across registration-completion and ticket-choice groups. A Clutch benchmark covering executive events from February 2025 through March 2026 found registration-to-attendance rates from 19% to 86%, depending on format.
Contents
- Registration completion and abandonment benchmarks
- Payment and ticket choice
- Registration timing and late sign-up risk
- Event format and attendance fallout
- Audience, season, and geography
- Capacity planning benchmarks
Registration completion and abandonment benchmarks
PheedLoop defines registration completion as completed registrations divided by registration attempts. Its 2026 anonymized live-event aggregates show that a higher completion rate does not automatically mean lower no-show risk. The median no-show rate was about 18% for events below 70% completion, 17% for events at 70–80%, 20% at 80–90%, 29% at 90–95%, and 21% at 95–100%. These are event-level medians, not a forecast for every event.
The relationship is therefore not a straight line. Events in the 90–95% completion band had the highest median no-show rate in this comparison, at about 29%, while the 70–80% band had the lowest at about 17%. Completion is useful for identifying friction in the registration journey, but it should be measured alongside attendance behavior.
The same PheedLoop report found a substantial difference between free and paid registration completion. Free events had a median completion rate of about 96%, compared with about 86% for paid events. A free form may attract more completed attempts, but completed registration is not the same as a firm attendance commitment.
One particularly exposed group was registrants with no ticket selected. Their median no-show rate was about 34%. That figure supports tracking ticket selection as a separate step rather than treating every completed form as an equally committed attendee.
Source: PheedLoop, Event Data Lab Report #06, published in 2026.
Payment and ticket choice
Payment status was associated with attendance reliability in the PheedLoop aggregates. Free events had a median no-show rate of about 28%, compared with about 17% for paid events. In the September 2026 addendum, paying attendees showed up 2–4 percentage points more reliably than non-paying attendees at the median event. Paying attendees were more reliable at 57–65% of events.
Among paid events, the no-show pattern across completion bands was narrower than the all-event comparison:
| Paid-event registration completion | No-show rate |
|---|---|
| Below 70% | About 18% |
| 70–80% | About 16% |
| 80–90% | About 18% |
| 90–95% | About 19% |
| 95–100% | About 16% |
Ticket quantity also mattered within paid events. Registrations for 0–1 tickets had about an 18% no-show rate; 2–3 tickets, about 21%; 4–5 tickets, about 17%; and 6 or more tickets, about 17%. The 2–3-ticket group was the highest in this set, but the differences are benchmarks rather than proof that ticket quantity causes attendance loss.
For event planners, these figures suggest separating at least four funnel measures: attempts, completed registrations, selected tickets, and checked-in attendees. A registration can be complete while still lacking the commitment signaled by payment or a selected ticket.
Source: PheedLoop, Event Data Lab Report #06. The payment interpretation was qualified in its September 2026 addendum.
Registration timing and late sign-up risk
PheedLoop Report #09, published in September 2026, found a clear pooled relationship between lead time and no-shows. Same-week registrants had a 42.4% no-show rate. The rate was 40.0% for people registering 1–2 weeks before the event, 36.2% at 2–4 weeks, 28.6% at 1–2 months, 26.3% at 2–3 months, 24.2% at 3–6 months, and 18.4% for people registering at least six months ahead.
| Registration lead time | No-show rate |
|---|---|
| Same week | 42.4% |
| 1–2 weeks | 40.0% |
| 2–4 weeks | 36.2% |
| 1–2 months | 28.6% |
| 2–3 months | 26.3% |
| 3–6 months | 24.2% |
| At least 6 months | 18.4% |
The report also cautions against treating timing as a universal attendee trait. Within the same event, registrants joining at least 30 days out had 84.1% median attendance, versus 80.8% for those joining inside 30 days. However, the median within-event difference was only 0.6 percentage points, and late registrants attended more reliably than early registrants at 44% of events.
Free and paid events had different timing profiles. Free events had a median registration lead time of 16 days, versus 41 days for paid events. Sixty-nine percent of free-event registrations arrived within 30 days, compared with 38% of paid-event registrations.
In pooled free-event data, same-week registrants attended at 28.2%, while those joining 2–3 months out attended at 59.3%. In pooled paid-event data, same-week registrants attended at 72.8%, while registrants joining at least six months out attended at 81.6%. These pooled figures combine events; they should not be read as a guarantee for a particular event.
Source: PheedLoop, Event Data Lab Report #09.
Event format and attendance fallout
The Clutch benchmark covers anonymized free-to-attend executive events in ANZ, APAC, and North America from February 2025 through March 2026. It measures attendance as attended divided by registered, and dropout as pre-event cancellations plus day-of no-shows divided by registrations.
Registration-to-attendance rates differed sharply by format. Sport and VIP experiences reached 86%, followed by executive lunches and dinners at 70%, customer and partner events at 62%, practitioner-led conferences at 38%, and vendor-owned conferences at 19%.
Private roundtables had a 61% attendance rate, with the middle half of events between 54% and 67%. At practitioner-led conferences, 22% of registrants cancelled in advance and 40% were day-of no-shows. At private roundtables, 11 of every 12 no-shows cancelled before the event, indicating a very different operational problem from last-minute nonattendance.
Topic orientation also correlated with attendance. Operator-leaning conference topics averaged 44% attendance, compared with 33% for builder-leaning topics. In Sydney, conferences had 34% attendance, while Sydney roundtables had 61%.
These comparisons show why a single “event no-show rate” can be misleading. Format, audience expectation, event size, and commitment level can all change the number of registrations needed to fill a room.
Source: Clutch, The State of Executive B2B Event Attendance, covering February 2025–March 2026.
Audience, season, and geography
In the Clutch data, directors and above attended at 62%, compared with 56% for technical doers. Technical doers had a no-show rate three times higher than directors and above. This audience difference matters when planners use historical attendance from one invite list to estimate another.
Seasonal differences were also reported. Roundtable cancellations were about 5 percentage points higher in the second half of the year than in the first half. Conference cancellations were about 9 percentage points higher in H2 than H1. These are cancellation differences, not total attendance differences.
Geography produced another large contrast. North American guests no-showed roughly 10 times more often than ANZ guests. In the city-level comparison, San Francisco roundtables had a 14.5% no-show rate, versus 0.6% in Sydney roundtables. The Clutch benchmark includes ANZ, APAC, and North America, so these results should be treated as regional benchmarks rather than universal geographic rules.
For planning, the practical implication is to segment historical data by audience, city or region, format, and season before applying it to a new event. Combining unlike groups can hide the actual source of registration fallout.
Source: Clutch, The State of Executive B2B Event Attendance, February 2025–March 2026.
Capacity planning benchmarks
The Clutch planner translated attendance benchmarks into registration targets. These outputs are benchmark planning targets, not observed attendance counts. Examples include 24 registrations to seat 20 people for sport or VIP experiences, and 29 registrations to seat 20 for executive lunches or dinners.
Larger or lower-attendance formats required much more over-registration. Customer or partner events required about 161 registrations to seat 100 people. Vendor-owned conferences required about 527 registrations to seat 100. Networking dinners required about 48 registrations to seat 25, while morning briefings required about 89 registrations to seat 40.
An operator-audience roundtable targeting 30 seats called for 49 registrations in the Clutch planner. For a 20-person engineering roundtable in October, the benchmark was closer to 38 registrations, compared with 33 in a stronger seasonal setup.
| Target seats | Format or setup | Benchmark registrations |
|---|---|---|
| 20 | Sport or VIP experience | About 24 |
| 20 | Executive lunch or dinner | About 29 |
| 25 | Networking dinner | About 48 |
| 30 | Operator-audience roundtable | 49 |
| 40 | Morning briefing | About 89 |
| 100 | Customer or partner event | About 161 |
| 100 | Vendor-owned conference | About 527 |
The figures are best used as starting points for a capacity model. A planner can compare the new event with the closest benchmark by format, audience, location, and season, then track actual cancellations and check-ins separately. Because the underlying Clutch figures describe free-to-attend executive events, they are not directly interchangeable with paid consumer conferences or any event outside that scope.
Source: Clutch, The State of Executive B2B Event Attendance, February 2025–March 2026.