The Pre-Event Survey Nobody Fills Out — And What to Send Instead

I work DC association and policy events. Membership organizations, advocacy groups, think tanks, government-adjacent conferences. My attendees are professionals with demanding schedules, 400-email inboxes, and the kind of time pressure that makes completing a 22-question pre-event survey feel roughly as appealing as doing their taxes.

And yet, every year, someone on the organizing committee says: “We should send a pre-event survey to understand what attendees want.” And every year, we send a pre-event survey. And every year, 11-15% of attendees complete it.

Let me explain why this is a problem beyond the obvious. An 11-15% completion rate is not a representative sample. The people who complete pre-event surveys are the people who had time to complete them — which, in my world, skews toward a specific subset of members: the most organizationally committed, the most digitally engaged, the semi-retired, and occasionally the people who are registered but haven’t confirmed their own attendance and have time precisely because they’re not that busy.

The people who don’t complete the survey — the 85-89% — are typically: senior staff who are calendar-constrained, external speakers who treat registration as the full extent of their pre-event commitment, private-sector attendees who are fitting the event around a full-time job, and first-time attendees who don’t feel the organizational loyalty that drives survey completion.

When you build your agenda from a 13% response rate, you are building your agenda for 13% of your attendees. The other 87% will arrive and encounter a program shaped around different priorities than their own. This is not the outcome the pre-event survey was designed to produce.

Why the 22-question format fails

The 22-question survey fails because it tries to do too many things simultaneously:

• Understand content preferences
• Identify session format preferences
• Collect logistical information (dietary restrictions, accessibility needs, roommate matching)
• Measure pre-event excitement and awareness
• Collect speaker feedback from previous years
• Add value by including “resources” as attachments
• Provide an opportunity for open-ended comments

These are five different surveys disguised as one. Each additional question reduces completion probability by measurable amounts — research on survey design consistently shows completion rates drop sharply after 5-7 questions and again after 10-12. At 22 questions, you’ve lost the majority of your audience before question 10.

The result is a long document that does each of its five jobs poorly because the completion sample is too small and too self-selected to be representative.

The two-question alternative

This is not original to me — I learned it from a facilitator I hired for a policy summit three years ago, who had gotten it from a researcher in organizational behavior. But I’ve used it at every event since and the results consistently exceed the 22-question approach.

You send a two-question email. The subject line is specific and human, not survey-language: “Quick question before the summit” or “One thing I want to know before we finalize the agenda.”

Question 1: “What’s the one thing you most hope to take away from [event name] this year?”

Question 2: “What’s the one thing you’re most worried about, or most want to avoid, at this year’s event?”

That’s it. Two text fields. No dropdowns, no radio buttons, no rating scales. Just two open-ended questions.

In my experience, this format generates:

• 70-82% response rates (versus 11-15% for the long survey)
• Responses that are specific and usable (versus the checkbox aggregations from the long survey that tell you “65% want more networking” without telling you what networking means to them)
• Qualitative signal that is genuinely useful for program design (the “one thing I’m worried about” answers in particular surface concerns that a fixed-option survey would never surface)

The last time I ran this format, for a 340-person policy conference, the “worried about” responses included:

• “That the panels will be the same people having the same conversations as last year” (said by 23 different respondents in different phrasing — this is not an answer any checkbox would have captured)
• “I’m new to this space and I don’t want to feel like I’m the only one who doesn’t know everyone” (said by 11 first-time attendees — we addressed this directly with a first-timer orientation)
• “The Wi-Fi situation at the venue last year was unusable” (we confirmed with the venue that the infrastructure had been upgraded and included that in the confirmation email)

None of these responses would exist in a 22-question survey. All of them were actionable.

What to send instead for the logistics stuff

The legitimate logistical data that gets buried in the 22-question survey — dietary restrictions, accessibility needs, roommate matching — needs its own dedicated communication, separate from the program-design survey.

A two-part strategy:

Registration form. Dietary restrictions and accessibility needs should be collected at registration, as required fields (with appropriate framing that makes clear they’re confidential and used only for event logistics). If you can’t modify your registration form, send a targeted email immediately after registration confirmation that asks only: “To help us plan your meals and ensure the event is accessible for you, please take 30 seconds to share any dietary restrictions or accessibility requirements.” One purpose, one email, two fields.

Roommate matching (if applicable). A separate form, sent to room-block guests only, with the specific questions relevant to roommate matching. Not buried in a general survey.

Separating these logistics collections from the program-design survey has two benefits: it keeps the program-design survey short (improving completion) and it makes the logistics responses more accurate (people answer logistics questions more carefully when they’re not fatigued from 20 preceding questions).

On the post-event survey, while we’re here

The post-event survey has the same structural problem as the pre-event survey but is often even worse — sent two days after an event when attendees have returned to their full-time jobs and the event feels like something that happened to another version of themselves.

Post-event survey completion rates in my sector: typically 8-14%. The people who complete it are again self-selected: the most satisfied (who want to express appreciation), the least satisfied (who want to complain), and the most organizationally committed. The modal attendee — adequately satisfied, time-pressed, not moved to effort — does not complete it.

The replacement I’ve moved to: a single-question text to registered phone numbers within four hours of the event ending. “On a scale of 1-10, how useful was today? One word for why.” Response rate: 45-60%, because it’s low-effort, immediate, and the format signals that you actually want a quick answer rather than a 15-minute survey.

The 1-10 score is comparable across events and is the only number I track year-over-year. The “one word for why” responses, aggregated, give me the qualitative signal I need to understand what moved the number. The combination takes attendees 20 seconds and gives me more usable data than a 22-question post-event survey.

The argument for doing it anyway

I want to acknowledge the counterargument: the pre-event survey, even at 11-15% completion, gives the organizing committee something to point to. It demonstrates process. It creates the appearance of having solicited input from attendees. In politically complex organizations — member associations, advocacy coalitions, anything with a board — appearing to have solicited input can matter more than the quality of the input you received.

I understand this. I’ve been in those organizations. Sometimes the survey exists not because the data will be used but because the process of soliciting data is itself a political necessity.

If that’s your situation, I’m not here to tell you to stop sending surveys. I’m saying: send the two-question email first, get the 75% response rate, use the responses to actually design your program — and then send the 22-question survey as your documented “attendee input process,” understanding that the 12% who complete it are not driving your decisions.

This is somewhat cynical. Events are somewhat political. The framework accommodates the reality.

For DC policy and association events, conference centers in Washington DC that have both plenary and breakout configurations are the right starting point — because once you’re running a two-question pre-event survey and actually hearing what attendees want, you’ll often find they want smaller, structured breakout sessions more than the plenary schedule you’ve been defaulting to. The DC corporate event venue directory covers a wide range of formats.

Worth reading alongside this: the networking piece on concentric circles addresses another format that emerges from listening to what attendees actually want, and your CEO hates these speeches applies the same “listen first” logic to the leadership program.

Send me the attendee list demographics and what the event is trying to accomplish. I’ll write you a two-question survey that will tell you what you need to know.