Survey questions: examples and how to write good ones
Most bad survey data comes from the wording, not the respondents. Here are the traps with before-and-after rewrites, example questions by purpose, and a checker for your own questions.
Check a survey question for bias
Quick checks run as you type. Press the button for a fuller review and a neutral rewrite.
- Contains “and/or” — check it isn’t asking two things at once.
- Loaded word (enjoy, great, terrible…) — it assumes the answer.
The question you enter is sent to an AI model via Vercel AI Gateway to generate the review. Nothing is stored.
Bad vs good: seven wording traps
| Trap | Biased | Better |
|---|---|---|
| Leading | How much did you enjoy the workshop? | How would you rate the workshop? |
| Double-barrelled | Was the room clean and quiet? | How clean was the room? / How quiet was the room? |
| Loaded (assumes) | What do you like most about our new app? | Have you used the new app? → If yes: What, if anything, do you like about it? |
| Vague time frame | How often do you eat out? | In the past 7 days, how many times did you eat a meal from a restaurant or takeaway? |
| Overlapping options | Age: 18–25, 25–35, 35–45 | Age: 18–24, 25–34, 35–44, 45 or older, Prefer not to say |
| Jargon | How satisfied are you with our omnichannel CX? | How satisfied are you with the help you got, by phone, chat or email? |
| Negative framing | Don’t you think the fee is too high? | Is the fee too low, about right, or too high? |
Survey question examples by purpose
Satisfaction
- Overall, how satisfied are you with [experience]? (Very dissatisfied → Very satisfied)
- How well did [product] meet your needs? (Not at all well → Extremely well)
- What is the main reason for your rating? (open)
Loyalty
- How likely are you to recommend us to a friend or colleague? (0–10, see NPS)
- How likely are you to buy from us again in the next 6 months?
Behaviour
- In the past 30 days, how many times did you use [feature]?
- Which of these did you use this week? (checkboxes, with “None of these”)
Events
- Which session was most useful to you? (dropdown)
- How would you rate the venue? (1–5 stars)
- What one thing should we change next time? (open)
Demographics (ask last, keep optional)
- Which age group are you in? (non-overlapping bands + “Prefer not to say”)
- How did you hear about us?
Rules for answer options
- Mutually exclusive: no value fits two options.
- Exhaustive: everyone has an option. Add “Other” with a text box, “None” or “Don’t know” where they apply.
- Balanced: as many positive options as negative ones.
- Ordered naturally: low to high, and in the same direction throughout the survey.
How to make a survey, step by step
- Write down the decision the survey will inform.
- List only the questions that affect that decision.
- Choose the answer format for each: Likert, rating, choice or open.
- Order them: easy first, sensitive last, one topic per step.
- Build it in the survey maker and preview on a phone.
- Pilot with three people, fix what confused them, then share the link.
Questions people ask
What makes a good survey question?
It asks about one thing, uses neutral wording, has a clear time frame, offers answer options that cover every case without overlapping, and can be answered from the respondent’s own experience.
What is a leading question?
One whose wording nudges toward a particular answer, such as “How much did you enjoy…?” (assumes enjoyment) or “Don’t you agree…?”. Rewrite neutrally: “How would you rate…?”
What is a double-barrelled question?
A question that asks two things at once, like “Was the staff friendly and helpful?”. Split it into two questions.
How many questions should a survey have?
As few as will answer your decision, typically 5 to 10 for customer surveys. Completion rates fall as surveys get longer, and later answers get less careful.
Open or closed questions?
Closed questions (choices, scales) are quick to answer and easy to count. Open questions uncover reasons you didn’t anticipate. Most good surveys are mainly closed, with one or two open ones.