Short answer: a post-event survey should stay under eight questions, go out the day after the event, and open with a single rating question. That rating question is usually NPS. You ask attendees how likely they are to recommend the event to a colleague on a scale of 0 to 10, then subtract the share who answered 0 to 6 from the share who answered 9 or 10. The result is a score between -100 and +100, and every other question exists to explain it. Average satisfaction is not the thing worth chasing. Freeman's 2025 Experience Trends Report found that 85% of attendees who had a peak moment at an event are likely to return, while only 40% said they had one. What follows: when to send the survey, which questions earn their place, how the NPS maths works, and how to turn the answers into next year's brief.
When should a post-event survey go out?
One rule covers the timing: ask while the attendee can still picture the room. In practice that means three separate moments. The first is inside the hall, where a QR code on screen at the end of a session collects a one-question rating for the speaker. The second is the morning after, which is when the real survey goes out by email, after the dinner and the trip home but while the impression is fresh. The third is a single reminder three days later. Close the survey after a week and put the closing date in the invitation; a survey left open indefinitely delays the report for the sake of a handful of late replies. One detail that gets skipped: send a survey to people who registered and did not show up. They get one question, and it is a different one. Why could you not make it? Next year's date, start time and city usually come out of those answers.
How many questions is too many?
Six to eight closed questions, plus one open question. The limit is not arbitrary. As a survey gets longer, respondents stop thinking and start picking the middle option, so the data degrades with every extra screen. Say on the first screen how many questions there are and how long it takes. Avoid required fields; even one mandatory question raises the drop-off rate, with the opening rating as the single exception. Decide the scale up front and leave it alone for the rest of the year. The five-point agree-disagree scale that entered the literature with Rensis Likert's 1932 study works for satisfaction items, while the NPS question is 0 to 10 by construction. Test the form on a phone rather than a desktop, because most people fill it in standing up, one-handed, on the way home.
How is NPS calculated? A worked example
Net Promoter Score spread through Fred Reichheld's article "The One Number You Need to Grow", published in Harvard Business Review in December 2003, and it is now a standard line in event reports. There is one question: how likely are you to recommend this event to a colleague, on a scale of 0 to 10? Answers split into three groups. Scores of 9 and 10 are promoters, 7 and 8 are passives, and anything from 0 to 6 is a detractor. The calculation subtracts the percentage of detractors from the percentage of promoters; passives stay in the denominator but drop out of the difference. Say 200 people respond: 120 give a 9 or 10, 50 give a 7 or 8, and 30 give a 6 or below. Promoters are 60%, detractors are 15%, so NPS is 45. The score runs from -100 to +100, and it is not a percentage. One caveat: NPS on its own never tells you what went wrong, so put a "why did you give that score?" field directly under it.
Which questions actually earn their place?
A set that works covers seven topics: likelihood to recommend, whether the content matched expectations, the length and pace of the programme, venue and travel, audio and visual quality, catering, and registration and welcome. The eighth is open-ended, and it produces the most useful answers: what is the one moment from this event that stayed with you? In Freeman's 2025 Experience Trends Report, 85% of attendees who had a peak moment say they are likely to return, yet only 40% report having had one. The differentiator is not average satisfaction, it is a single memorable moment, and an open question is the only way to find out what it was. Add one more: would you attend the same event next year? Intent is a more honest signal than a satisfaction score, because this time the respondent is committing their own time.
Do not build the survey around the stage alone
Most event surveys ask about sessions, speakers and the main hall, and skip the places where attendees actually spend their time. Part two of Freeman's Unpacking XLNC report, published in July 2026 and based on a survey of more than 3,300 attendees and exhibitors, found that attendees would ideally spend 54% of their learning time outside traditional session rooms, and that the figure rises to 64% at trade shows. Part one of the same series, drawing on more than 4,700 attendees and 185 event organisers, found that 70% of attendees rank in-person events as their top source of professional learning. The practical reading is straightforward. Put questions about the coffee breaks, the networking area, booth visits and the catering layout into the survey as well, or you are reporting on half the attendee experience.
Five habits that corrupt the results
First, leading questions. A survey that asks how successful you found our organisation of the event is never going to return a low score, so keep the wording neutral. Second, rewriting the question set every year; if you want a comparable series, the core questions have to keep their exact wording and new ones should be added alongside. Third, making names mandatory, which softens criticism, so keep the name field optional. Fourth, picking a channel only happy attendees will reach, such as sending the survey exclusively to people who downloaded the event app. Fifth, and the most common, reducing everything to one average. A 4.2 can come from a room where everyone gave a 4, or from a room split between fives and threes. Read the distribution, not the average.
Collecting responses on site, and the data protection side
Email is not the only channel. A QR code on screen at the end of a session, a code printed on the back of the badge, a tablet at the exit and an in-app notification can all feed the same survey. Inside the hall, ask one question only, a rating for the speaker or the session, and collect the rest by email the next day. Two legal points apply in Turkey. If you link answers to individuals by matching them to an email address or a registration number, you are processing personal data, which triggers the disclosure obligation under article 10 of Law No. 6698, so put a short privacy notice and a link at the start of the survey. If you plan to send commercial messages to respondents afterwards, that is a separate consent; a survey form does not replace an IYS approval.
After the survey: the report, the supplier scorecard, the next brief
A survey is worth exactly as much as the document it ends up in. Produce three outputs. The first is a one-page management summary: NPS, number of responses, the three highest and three lowest scoring topics, and three quotes from the open-ended answers. The second is a supplier scorecard with separate ratings for audio, lighting, LED screens, catering, transport and the venue, because that scorecard is the only objective data you will have in the next round of quotes. The third is the action list that goes into the brief for the next event, and here apply the rule of three: pick the three most complained about issues, assign an owner and a date to each, and leave the rest for next year. A list that tries to fix everything fixes nothing. Tales Event works out of Istanbul and runs dealer meetings, product launches, conferences and gala nights with one team covering stage, audio, lighting, LED screens and registration, and we build the measurement plan alongside the production plan. Tell us your headcount and event type and we will draft the question set with you.
Let's talk about your event
Tell us your idea and we will prepare the concept and quote.