Every registration form loses some of the people who open it. Many teams respond by deleting questions, because that is the most common advice. It feels like progress and costs you the data you wanted. The evidence points somewhere else.
What good looks like
The largest public dataset on event registration we have found comes from PheedLoop, an event platform, which analysed registration flows across more than 3,600 live events. It measured completion: the share of people who finished registration out of everyone who started.1
Median completion was above 90% for conference-style events and about 84% for trade shows. The more useful finding is that variation within each event type was wider than the gap between types,1 so if your completion is well below 90%, the cause is probably in your own flow rather than in the kind of event you run.
Choices cost you, questions do not
Ticket types were the strongest predictor of completion. Median completion fell from 93% with one ticket type to 88% with two or three, and to 83% with eleven or more. The steepest fall came in the first few options.1
Custom questions showed no negative effect. Events with 21 or more questions completed at higher rates than events with one or two, which PheedLoop attributes to the kind of event that asks them rather than to the questions themselves.1 Either way, the long forms were not losing people.
Grouping tickets helps. Registration categories on their own made no difference. But for events with two or three ticket types, those with four or five categories completed at about 90%, against about 81% for single-category events with the same tickets.1 The likely mechanism is simple: a delegate who picks "doctor" first sees only the tickets for doctors.
In plain numbers: if 1,000 people start registering, 93% completion gives you 930 registrations and 83% gives you 830. That is a hundred people lost to the design of a ticket menu, before anyone has touched the marketing.
One worry worth putting to rest: making registration easier does not fill the room with people who never show up. Among paid events, no-show rates were flat whatever the completion rate.3 We covered that in event no-shows.
The payment step
PheedLoop could not isolate the payment step, so it is worth borrowing from the people who study checkouts for a living. Baymard Institute surveys US online shoppers about why they abandon orders. Leaving aside people who were only browsing, the leading reasons were extra costs appearing late (40%), being asked to create an account (18%), a long or complicated checkout (17%), and not being able to see the total cost up front (12%). Not enough payment methods came in at 9%.2
Shoppers are not delegates, so treat these as direction rather than benchmarks. But the direction is clear, and it lines up with the event data: people leave over surprises and decisions, not over typing. A platform fee or convenience charge that appears on the last screen is exactly the late extra cost shoppers say they leave over.
What nobody has measured yet
PheedLoop is candid about the gaps. Its dataset does not cover the required base fields in a form, the difference between mobile and desktop, the payment step itself, or how clearly ticket options are described.1 Baymard's work suggests the first and third matter in checkouts. So: keep the questions you will actually use, and be ruthless with everything that asks people to decide, create, or pay more than they expected.
What to fix first
Measure before you change anything. Completion is finishes divided by starts, so if your platform only reports finished registrations, you cannot see this problem at all. Ask for the number of people who started, then make one change at a time.
When the alternative wins: for an invitation-only event with a known guest list, the best form is no form. Import the list, send each guest a personal invitation with their details already filled in, and ask them only to confirm.
Where Aurentex sits
Registration is included in every Aurentex plan and runs on your own domain and branding. For known guest lists, bulk import and invitations let you skip the open form altogether. If you charge, payments go to your own gateway and we take 0% of ticket sales, so there is no platform fee to surprise anyone on the last screen. Your first event is free, which gives you a completion rate of your own to compare against the numbers above.
Common questions
What is a good completion rate?
Above 90% for conferences, about 84% for trade shows, at the median.1
Do long forms reduce sign-ups?
Custom questions did not, in data from 3,600+ events.1 Required base fields and mobile were not measured.
How many ticket types should I offer?
As few as you need. Completion fell from 93% with one to 83% with eleven or more.1
Why do people abandon at payment?
In checkout research, mainly late extra costs, forced accounts and long checkouts.2