It’s Thursday morning, three days after the event closed. The survey goes out. Subject line: “We’d love your feedback!” Fourteen questions. The first one asks attendees to rate their overall experience on a scale of one to seven.
By Thursday your attendee has been back at their desk for two days with a full inbox, and genuinely cannot remember whether the Tuesday breakout was the good one or the one where the microphone cut out twice. They archive it.
Then we look at the response rate and conclude that attendees don’t like giving feedback.
Your number is probably fine, which is the problem
Explori, which runs one of the largest post-event survey datasets in the events industry, puts a good response rate at 10 to 20 percent for conferences and corporate events, and around 12 to 15 percent for live trade shows.
Sit with that for a second. At the top of the range, four out of five people didn’t tell you what they thought. In the middle of it, six out of seven didn’t. And that’s the number we’ve agreed to call healthy.
Somebody out there is getting 80 percent
Nicola Kastner runs ELX, an invite-only community for heads of events at large corporations. Her post-event response rates sit consistently in the eighties and nineties. A bad result, by her definition, is anything below 75 percent. That has happened once.
Her members are SVPs and heads of events running their own portfolios while attending hers. They are precisely the time-poor senior audience that usually gets blamed for low response rates. And she uses no incentives at all: no vouchers, no prize draws, no charitable donation in your name.
What she does instead is tell attendees at the start of every event that their feedback determines what the next one looks like, then prove it during the event by naming specific changes and attributing them to what members said last year. By the time the survey lands, completing it is self-interest.
Explori’s read on the industry-wide decline is blunt: it’s usually a symptom of a broken feedback loop, not a technology problem and not an audience problem.
Where your survey actually loses people
It arrives after the memory is gone
Specific recall decays fast. The thing that irritated someone at 2pm on Tuesday, the room that was too cold, the Q&A that ran four minutes and got cut, is gone by Thursday. What’s left is a vague overall impression, which is the least actionable feedback there is and precisely what a delayed survey is built to collect. Kastner deploys hers as attendees are leaving, so people open it on the journey home while the event is still in front of them.
It asks for more than you’ve earned
Fourteen questions is a significant ask from someone who will never see the report it feeds. Every question past the point of genuine interest costs you completions from people who started in good faith and bailed somewhere in the middle.
Nobody knows what the numbers mean
A one-to-five scale with no labels means one thing to the attendee who treats 3 as “fine” and something completely different to the attendee who treats 3 as “disappointing.” You average across that noise, get 4.1, compare it to last year’s 4.3, and have a conversation about a decline that may not exist. Unlabeled scales produce precision without accuracy, which is the most dangerous kind of number to put in a board deck.
It asks about the event when you need to know about the session
One survey covering three days, asking how attendees felt about “the sessions” collectively, produces an average that cannot tell you which specific breakout to cut. You wanted session-level intelligence and built an instrument that can only return event-level mush.
The fix is less about the form than you’d like it to be
Worth being straight about this, because it cuts against the obvious advice: Explori’s position is that the corrective action is not a new distribution channel, a shorter survey, or a gift card. Shortening your survey and fixing your timing will help at the margins. Neither will take you from 14 percent to 80.
What moves the number is showing people the loop closing. Tell attendees at the start that feedback shapes next year. During the event, point at the thing you changed and say who asked for it. After the event, report back on what you heard and what you’re doing about it. That’s a programming and communications commitment more than a survey tactic, and it’s the part most organizations skip because it’s the part that requires actually acting on the results.
Then do the mechanical things too. Ask while the event is still live in people’s minds. Cut the question count to what you’ll genuinely act on, chosen against a decision you’re about to make: if you’re deciding whether to rebook a speaker, ask about the speaker. And where you can, replace the rating scale with something unambiguous. “Would you attend another session by this speaker?” answered yes, no, or maybe beats a seven-point scale nobody calibrates the same way.
What attendance data tells you that no survey can
Even at Kastner’s numbers, a survey has a structural blind spot, and it’s the one that matters most for programming decisions.
The people who complete a session evaluation are, by definition, people who stayed in the room long enough to have an opinion worth submitting. The attendee who walked out at minute twelve because the speaker was reading their slides aloud did not fill out your form. They’d already written the session off and gone to find coffee. So your 4.2 out of 5 is the average opinion of people who liked it enough to sit through it.
A session that holds 90 percent of its room for the full 45 minutes and a session that loses half the room by minute 15 can produce identical survey scores, because in both cases only the people who stayed filled out the survey.
Attendance tracking gives you a census instead of a sample. Every person who badged into that room is in the dataset, including everyone who would never answer a survey. That hands you several things no survey can:
- Who actually came, as opposed to who registered for the session or said they planned to attend
- When people left, which is the single most honest signal a session produces and one no post-event survey will ever capture
- Whether the room was full because the content drew people or because you put it in a small room, which are very different findings that look identical in a photo
- Which sessions pulled your first-time attendees versus the ones that only drew people who were coming back anyway
- Verified time in session, which is what accreditation bodies want when someone claims a credit, and what gives an attendee real proof of participation when their licensing board asks
Attendance tells you what happened in that room. Surveys tell you why, from the people willing to explain.
Run the two together and they answer different halves of the same question. Most organizations are trying to run the whole operation on the “why,” gathered late, from a small and self-selected group of people who already liked it.
What to do before your next event
Pick one change you made because of last year’s feedback and say so from the stage, naming what prompted it. That single act does more for your next response rate than any redesign of the form.
Then pull your attendance data for last year’s lowest-rated session and look at when people left. If they stayed to the end, you have a content problem worth solving. If half the room emptied by minute fifteen, your rating was never measuring the session at all. It was measuring the people who didn’t mind it.
Stop planning next year’s agenda on a 14 percent sample.
A2Z Events by Momentive records session attendance through badge scanning on phones, tablets, or handheld scanners, and in/out scans capture how long people stayed. Real-time scan summaries show who is in which room as it happens, and conditional filters in reporting let you compare session popularity and participation across attendee groups. Every person who walked through the door is in that data, which is a different foundation to plan from than a survey.
