A Survey Is a Favour, Priced in Minutes

Nobody owes you feedback. An attendee who fills in your survey is doing you a favour, and they are paying for it in the only currency that matters at nine o'clock at night - minutes. The published data on where they stop paying is remarkably consistent, and almost every post-event survey is priced well above it.

Quick answer

An analysis of 267,564 survey responses found completion rates of 83.3 percent for surveys of one to three questions, 65.2 percent at four to eight, 56.3 percent at nine to fourteen, and 41.9 percent at fifteen or more.

Post-event surveys specifically average around a 38.8 percent response rate on SurveyMonkey's platform data - notably higher than pre-event surveys at 28.7 percent, because the experience is fresh and the obligation feels real.

The channel matters as much as the length: email averages 49.2 percent, a web link 30.0 percent, SMS 18.5 percent, and a social post 7.6 percent. A short survey sent badly still fails.

The price curve, measured

The most useful single dataset on survey length comes from Survicate's analysis of 267,564 responses, which sorted completion rates by the number of questions asked. The shape of it is not subtle.

Questions - Completion rate - Drop from previous band

1–3 - 83.34% - -

4–8 - 65.15% - −18.2 points

9–14 - 56.28% - −8.9 points

15+ - 41.94% - −14.3 points

The steepest single fall is the first one: moving from three questions to four costs more than eighteen points of completion. That is the most expensive question you will ever add, and most survey drafts add it without a thought.

An independent medical research study found the same shape from a different direction, with completion dropping from 63 percent on its shortest instrument to 37 percent on its longest - a 26-point decline attributable to length alone, holding the subject matter constant.

Question four is the most expensive one in the survey. It costs eighteen points of completion, and it is almost always the one you added because it seemed harmless.

What people will actually spend

SurveyMonkey's platform data puts a number on the tolerance directly: 48 percent of respondents say their maximum is one to five minutes. Not their preference - their maximum.

That figure is worth converting into questions rather than leaving as time, because organisers systematically underestimate how long their own survey takes. A considered answer to an open text question is not fifteen seconds; it is closer to sixty or ninety, and a survey with three open fields has quietly become a five-minute commitment on its own.

The practical translation: a post-event survey should be readable in a single screen, answerable in under two minutes, and should contain at most one open text field. Everything beyond that is being paid for with your response rate.

Post-event surveys start from a better position than most

There is genuinely good news in the benchmark data, and it argues for asking sooner rather than more.

SurveyMonkey's 2025 platform data breaks response rates down by survey type, and post-event surveys average 38.81 percent - well above general market research at 38.59 percent and customer feedback at 31.81 percent, and dramatically above pre-event surveys at 28.65 percent.

Survey type - Average response rate

Market / concept testing - 66.58%

Course feedback - 48.68%

Post-event survey - 38.81%

General market research - 38.59%

Academic research - 37.20%

Customer feedback - 31.81%

Pre-event survey - 28.65%

The gap between post-event and pre-event is the interesting one. The same people, asked about the same event, respond ten points more often afterwards. Recency and reciprocity are doing that work: they have just received something, and the memory is intact.

That decays fast. The advantage is largest immediately after the event and shrinks with every day you spend polishing the survey, which is a strong argument for sending a rough short survey on the night rather than a perfect one on Friday.

Channel is not a detail

The same platform data breaks response rates down by how the survey was delivered, and the spread across channels is far larger than most organisers assume.

Channel - Response rate - Best use

Audience panel - 91.61% - Paid respondents; not applicable to events

Offline / kiosk - 57.49% - In-room, before people leave

Email - 49.17% - The reliable default for events

Mobile SDK - 34.37% - In-app prompts

Web link - 29.95% - Shared link in a follow-up

SMS - 18.54% - Short, urgent asks only

Facebook - 7.61% - Broad reach, poor conversion

Website popup - 3.65% - Effectively noise

Two things stand out. Email at 49 percent is far stronger than a shared web link at 30 percent, despite both being a link that arrives on a screen - the difference is that an email is addressed to a person and a link is addressed to nobody.

And the offline or kiosk channel at 57 percent outperforms everything except a paid panel. That is a direct argument for the least fashionable option available: ask people while they are still in the room.

Ask in the room, then ask again by email

The two highest-performing channels available to an event organiser are an in-room prompt and an email, and they are not alternatives. They catch different people.

The in-room ask catches people at maximum warmth and minimum friction, but only reaches those who are still there at the end and not mid-conversation. The email catches people who left early, people who wanted to think, and people who will not fill in anything while standing up.

• In the room, one question. A single visible question with a QR code or a card. Not a survey - one question. This is where the 57 percent channel lives.
• By email, within 24 hours. Three to five questions maximum, one of them open. This is the 49 percent channel, and it decays quickly after the first day.
• Never a third ask. The people who did not answer twice are giving you information too, and a third prompt costs goodwill for a response you will not get.

The one-question in-room ask deserves emphasis because it feels too small to be worth doing. At 83 percent completion for one-to-three question instruments, on the highest-response channel available, it is very likely the single most efficient feedback you will ever collect.

The questions worth their price

If a post-event survey can only afford four or five questions before its completion rate collapses, the drafting problem is not what to include but what to cut. A useful discipline is to require every question to change a decision.

A question that will not alter what you do next time is not research; it is curiosity, and it is being paid for out of the same budget as the questions that matter.

Common question - What it costs - What to ask instead

Rate the venue 1–5 - One slot, rarely actionable - Would you come back to this venue? Yes / No

Rate each speaker 1–5 - One slot per speaker - Which session would you tell a colleague about?

How did you hear about us? - One slot, better answered at registration - - (move it to sign-up)

Any other comments? - 60–90 seconds of respondent time - Keep it - but make it the only open field

Demographic questions - Multiple slots - - (move to registration)

Two patterns come out of that. Anything you could have asked at registration should have been asked at registration, where it costs a different budget. And rating scales are expensive relative to what they tell you: a five-point venue rating averaging 3.8 does not tell you whether to rebook, while a yes-or-no on returning does.

If the answer would not change what you do next time, the question is curiosity, and your respondents are paying for it.

Three questions before you send

• How many questions is it, honestly, counting each sub-item? A matrix with six rows is six questions wearing one heading, and the completion curve counts it that way.
• Did you ask anyone in the room before they left? The highest-response channel available to you is the one that closes when the doors do.
• Would every question change a decision? Cut every one that would not, and watch the completion rate on the remainder rise.

What to take from this

Feedback is not owed to you and it is not free. It is bought with your attendees' minutes, at a price curve that is steep, well-measured, and largely unforgiving: three questions holds 83 percent, fifteen holds 42.

So ask fewer things, sooner, in the room if you can and by email if you cannot, and make every question earn its place by changing something you will actually do. A four-question survey answered by half your attendees is worth vastly more than a twenty-question one answered by a tenth of them - and it costs the people who came a great deal less.

Frequently asked questions

How many questions should a post-event survey have?

Ideally three to five. Completion data across 267,564 responses shows 83.3 percent completion at one to three questions, falling to 65.2 percent at four to eight and 41.9 percent at fifteen or more.

What is a good response rate for a post-event survey?

SurveyMonkey's platform data puts the average for post-event surveys at 38.81 percent, notably higher than pre-event surveys at 28.65 percent, because recency and reciprocity both work in your favour afterwards.

When should I send a post-event survey?

Within 24 hours. The advantage post-event surveys hold over pre-event ones comes from the experience being fresh, and that advantage decays with every day spent refining the questions.

What is the best way to deliver an event survey?

Email performs best among practical options at around 49 percent, but an in-room offline or kiosk ask performs better still at roughly 57 percent. Web links average 30 percent and social posts around 8 percent.

How long will people spend on a survey?

Around 48 percent of respondents say one to five minutes is their maximum, and open text fields consume that budget far faster than organisers expect - often 60 to 90 seconds each.

Should I ask demographic questions in a post-event survey?

Better to move them to registration. Anything that could have been captured at sign-up is spending your most constrained resource - post-event question slots - on information you could already have had.

Are rating scales worth including?

Often not, relative to their cost. A five-point venue score averaging 3.8 rarely tells you what to do, whereas a yes-or-no question about whether someone would return gives you a decision directly.

Should I send a reminder if people don't respond?

One follow-up is reasonable; a third ask is not. People who have declined twice are giving you information, and further prompts cost goodwill for responses that are unlikely to arrive.

Ontaym Editorial Team

Ontaym builds tools for organising real-world gatherings, so the team spends its days on the coordination problems this article describes. Articles are researched against primary sources, reviewed before publication, and revised when the underlying facts change rather than on a schedule.

Close the loop on your next event without another form.