The ten-question survey, timing, response rates, and how to read the results honestly.
Read the transcript
Surveys and feedback.
The ten-question survey, timing, response rates, and how to read the results honestly.
Why forty-eight hours.
Memory of an evening fades fast and gets rewritten by whatever people said about it on the way home.
Send the survey within forty-eight hours of the last guest leaving and you capture what they experienced; send it a week later and you capture what their friends told them.
The ten questions.
Ten questions is the ceiling.
Every question past ten costs you responses, and you will not act on the eleventh anyway.
Response rates and what they mean.
In our experience, a guest survey sent within two days with a single reminder on day five returns roughly fifteen to thirty percent of invited guests for a social or fundraising event, and somewhat higher for a corporate event where attendance was expected.
Below ten percent, the sample is too small to read as anything but anecdote.
Reading the results honestly.
Average scores hide the shape.
Count how many people gave a 1 or a 2 on each question; three angry guests out of fifty tell you something an average of 4.1 does not.
Sharing findings with vendors.
Vendors almost never hear what guests thought, which is why the ones who do remember you.
Send each vendor lead the two or three lines that concern them, good and bad, within a week.
That is the lesson. The worksheet and the quiz are just below this video.
What you will be able to do
- Send the survey within 48 hours
- Analyse results
- Share findings with vendors
Why forty-eight hours
Memory of an evening fades fast and gets rewritten by whatever people said about it on the way home. Send the survey within forty-eight hours of the last guest leaving and you capture what they experienced; send it a week later and you capture what their friends told them. Forty-eight hours is also before most people have deleted the event from their inbox, which is where your response rate lives. Draft the survey before the event, load the guest emails, and the only thing left on the Friday morning is to press send.
The ten questions
Ten questions is the ceiling. Every question past ten costs you responses, and you will not act on the eleventh anyway. Use a rating scale for the first seven so you can compare events, and leave three open. The survey template has the wording ready to copy.
- Overall, how would you rate the event? (1 to 5)
- How would you rate the venue: location, room and comfort? (1 to 5)
- How would you rate the food and drink? (1 to 5)
- How would you rate the program: speakers, timing, length? (1 to 5)
- How would you rate the entertainment? (1 to 5)
- How easy was it to register, arrive and find your seat? (1 to 5)
- How likely are you to attend next year? (1 to 5)
- What was the best part of the evening?
- What would you change?
- Anything else you would like us to know?
Add one question specific to the event’s purpose if the client needs it: for a fundraiser, “Did you feel informed about where your donation goes?”; for a corporate event, “Did you make at least one useful new contact?” Swap it for question six rather than adding an eleventh.
Do not ask for a name. Anonymous surveys get honest answers. Do ask, optionally, for a contact if the person wants a reply, and treat that contact under PIPEDA: say why you are collecting it and delete it when the follow-up is done.
Response rates and what they mean
In our experience, a guest survey sent within two days with a single reminder on day five returns roughly fifteen to thirty percent of invited guests for a social or fundraising event, and somewhat higher for a corporate event where attendance was expected. Below ten percent, the sample is too small to read as anything but anecdote. Above thirty, you can start to trust the averages.
Who responds matters as much as how many. People with a strong opinion respond more than people with a mild one, so your scores will be more spread than the room really felt. Guests who left early rarely respond at all, which hides the one thing you most want to know. Pair the survey with the door count and the time the room emptied; those numbers do not have a response bias.
Reading the results honestly
| What you see | What it probably means | What to do |
|---|---|---|
| Overall 4.6, food 3.2 | The evening worked despite the food; the hotel kitchen or the menu choice is the weak point | Share with the venue in writing; change the menu or the venue next year |
| Program 3.4, open comments mention “long” | The speeches ran over, which you know from the run of show | Show the client the data beside the timings; propose a shorter program with a hard cap |
| Entertainment 4.8, four comments name the mentalist | The act landed and is worth rebooking | Tell the performer, with the quotes; use it in the project write-up |
| Arrival 2.9 | Coat check or registration backed up | Staffing fix, not a venue fix |
| Likely to return 4.5 but only 12 responses | Not enough to read | Report it as “early signal”, not as a result |
Average scores hide the shape. Count how many people gave a 1 or a 2 on each question; three angry guests out of fifty tell you something an average of 4.1 does not. Read every open comment, group them into themes, and count the themes. “Five comments about the sound level at the back of the room” is actionable. “Some people did not like the sound” is not.
Sharing findings with vendors
Vendors almost never hear what guests thought, which is why the ones who do remember you. Send each vendor lead the two or three lines that concern them, good and bad, within a week. Keep it factual and quote the guests rather than editorialising. For a problem, pair the quote with what you observed on the night and ask how they would handle it next time; that is a conversation most vendors welcome, and it tells you who to keep on the list.
Worked example: the Harbour Gala
The survey goes to 203 attendees on Saturday morning, under the foundation’s name, with a reminder on Wednesday. Fifty-one responses by the following Friday, which is twenty-five percent and enough to read. Overall 4.5. Venue 4.3. Food 3.6, with nine open comments mentioning cold mains at the back tables, which matches the captain’s warning that twelve servers was thin for twenty rounds. Program 4.0, with the chair’s speech named by six people as “a bit long”. Entertainment 4.8; the mentalist is mentioned by name eleven times and the strolling magicians by seven. Arrival 3.9, with the coat check named four times. Likely to return 4.4. The purpose question, “Did you feel informed about where your donation goes?”, scores 4.6, which the executive director puts in her board report.
The lead sends the hotel the food comments with the server count beside them and asks for fourteen servers next year at the same price. The mentalist gets the eleven mentions and three direct quotes, which he asks to use on his own site. The coat check line goes on the fix list for next year with a second station from 5:45. All of this is in the client’s hands within ten days of the event.
Numbers to remember
- Send within 48 hours; one reminder on day five.
- Ten questions maximum; seven scaled, three open.
- Roughly 15 to 30 percent response, in our experience; under 10 percent is anecdote.
- Count the 1s and 2s, not just the average.
- Vendors hear their two or three lines within a week.
Worksheet
Download the worksheet, complete it for your own event (or the running gala example), and keep it in your course binder. It takes about 10 minutes.
Check your understanding
Related guides, venues and vendors
Reading a floor plan and a BEO
Module 1: Foundations · 15 min
Two documents decide whether a room works. Learn to read scale, capacities, egress and the BEO line items that hide costs.
Read more
Vendors, sponsors and revenue
Specialist · Festivals & Public Events · 15 min
Vendor fees, sponsorship tiers, ticketing vs free, and the festival budget shape.
Read more
Favours, themes and budgets under $1,000
Specialist · Kids' Parties · 15 min
Where the money goes and what kids actually remember.
Read more