50 Customer Survey Questions That Earn Their Answers (2026)

TL;DR: Good survey questions share three habits: one idea per question, neutral phrasing, and a reason to exist. The bank below holds 50 of them, grouped by goal rather than by template, so you can lift the three or four that match your moment. Ask fewer than you think; surveys with 1 to 3 questions see completion above 80%. For the full method (metrics, timing, response benchmarks), our customer feedback survey guide is the long version.

How to use this bank

Pick the group that matches your decision, take the scored question plus one or two follow-ups, and always keep one open text field, optional. Nothing here is double-barreled, leading, or rhetorical. Every question survives mobile.

Measuring loyalty: NPS questions (1 to 8)

1. How likely are you to recommend us to a colleague, 0 to 10?
2. What is the single biggest reason for your score?
3. What would move your score up by one point?
4. What do you tell people when you describe us?
5. Which feature would you fight to keep?
6. What almost made you choose something else?
7. How does your team feel about us, in one sentence?
8. If we disappeared next month, what would you do instead?

Measuring a moment: satisfaction questions (9 to 18)

1. How satisfied were you with today's experience, 1 to 5?
2. Did we solve the thing you contacted us about?
3. Did you have to contact us more than once for this?
4. How would you rate the person who helped you?
5. How fair did the wait time feel?
6. Was anything about the process confusing?
7. How well did we understand your actual problem?
8. What should we have done differently today?
9. Was the resolution explained in plain language?
10. Is anything still unresolved we should pick back up?

Measuring friction: effort questions (19 to 26)

1. How easy was it to get what you needed, 1 to 7?
2. Where did you get stuck, even briefly?
3. How long did setup actually take you?
4. Did you need help from a person to finish?
5. What did you expect to be easier?
6. Which step would you delete entirely?
7. Did our docs answer your question before support did?
8. If a friend started tomorrow, what would you warn them about?

Measuring fit: product questions (27 to 36)

1. How disappointed would you be if you could no longer use the product?
2. What were you using before us?
3. What job did you hire us to do?
4. Which feature surprised you most, in either direction?
5. What do you use weekly? What have you never opened?
6. What would make this a must-have instead of a nice-to-have?
7. Who else at your company should be using this, and why are they not?
8. How does the price feel against the value, honestly?
9. What is the closest competitor we should worry about?
10. If you could add one integration tomorrow, which one?

Understanding goodbyes: churn questions (37 to 43)

1. What is the main reason you are leaving?
2. What did you switch to, if anything?
3. When did you start considering leaving?
4. What could we have done three months ago to keep you?
5. Which expectation did we fail to meet?
6. Would you come back if we fixed one thing? Which thing?
7. May we follow up once we have acted on this?

The openers that explain everything (44 to 50)

1. What almost stopped you from finishing this today?
2. What is the one thing you would improve first?
3. What do we do better than anyone else, in your view?
4. What question should we have asked but did not?
5. What is the most annoying small thing about the product?
6. Describe us to a stranger in one sentence.
7. Anything else you want the founders to read?

Three rules that keep the answers honest

Never lead. "How much did you love the new dashboard?" is a request wearing a question's clothes. Never double-barrel. "Was it fast and accurate?" cannot be answered by anyone for whom it was one but not the other. Always close the loop. People answer the second survey because something visibly happened after the first one.

When you are ready to ship one, the survey templates gallery has these pre-built, and TinyForms collects unlimited responses on every plan, free included, with branching so respondents only see questions meant for them. The timing and channel math (in-app roughly triples email response rates) lives in the main guide.

Survey question FAQ

How many questions should a customer survey have?

One to three for transactional surveys, five to eight maximum for relationship surveys with branching. Completion above 80% is realistic at the short end; it collapses past the two-minute mark.

What is the best first question for a survey?

A one-tap scored question matched to your goal: 0 to 10 likelihood for NPS, 1 to 5 satisfaction for CSAT, 1 to 7 ease for CES. Momentum from a single tap carries people into the open follow-up.

Should survey questions be required?

The scored question, yes. Open text, no. Required essays turn a 30-second survey into an abandoned one; the people with something to say will say it anyway.

What is the difference between NPS, CSAT, and CES questions?

NPS asks about recommending you (relationship). CSAT asks about satisfaction with one interaction (moment). CES asks how easy something was (friction). Match the metric to the decision you plan to make with it.