30 Customer Satisfaction Survey Questions That Get Honest Answers

Good customer satisfaction survey questions share three traits: they’re specific, they’re answerable from memory, and they never lead the witness. “How amazing was our fast support?” fails all three. “How long did you wait for a first reply?” passes.

Most surveys fail before the first response comes in, because the questions were written to fish for compliments instead of information.

Below are 30 customer satisfaction survey questions I’ve used across client sites, grouped by what you’re trying to learn: overall satisfaction, product quality, support, loyalty, open-ended feedback, and post-purchase timing. Treat the 30 as a bank, not a survey. You’ll pick 5–8 per survey — more on that near the end.

Scales measure, open text explains. A good survey uses both.

The 3 rules of questions that get honest answers

Honest answers come from questions that are specific, single-topic, and matched to the right format. Every question in this list follows three rules:

Specific beats general. “How was your experience?” produces “fine.” “How easy was checkout, from finding the product to payment confirmation?” produces something you can act on. The narrower the question, the more useful the answer.

One idea per question. “Was the product high quality and delivered on time?” is two questions wearing one label. A customer who got a great product late has no honest way to answer it. Split it.

Scales for measuring, open text for learning. A 1–5 scale tells you that satisfaction dropped. Only an open text box tells you why. You need both, in that order.

Overall customer satisfaction survey questions

These five measure the headline number — how satisfied customers are right now:

1. Overall, how satisfied are you with [product/company]? (1–5 scale) — The baseline CSAT question. Ask it the same way every time so the trend line means something.
2. How well does [product] meet your needs? (1–5 scale)
3. How would you rate the value for the price you paid? (1–5 scale) — Separates “I like it” from “I’d pay for it again.” Those are different answers more often than you’d think.
4. Compared to what you expected, was [product] better, worse, or about the same?
5. How satisfied are you with your most recent interaction with us? (1–5 scale)

Product or service quality questions

These five isolate the product itself from everything around it — support, shipping, billing:

6. How would you rate the quality of [product]? (1–5 scale)
7. How easy is [product] to use? (1–5 scale) — Low scores here predict churn earlier than almost any other signal I’ve seen.
8. Which feature do you use most?
9. Which feature could you not live without? — The answer tells you what to protect in your roadmap. Cutting or complicating this feature is how products quietly lose their best customers.
10. Did [product] solve the problem you bought it for? (Yes / Partially / No)

Support experience questions

Send these five only to people who actually contacted support — surveying everyone else adds noise:

11. How satisfied are you with the support you received? (1–5 scale)
12. How long did you wait for a first response? (Under 1 hour / Same day / 1–2 days / Longer) — Asks for a fact, not a feeling. Facts from memory are more reliable than vibes.
13. Was your issue resolved on the first contact? (Yes / No)
14. How knowledgeable was the person who helped you? (1–5 scale)
15. What could we have done to resolve your issue faster?

Loyalty and repeat business questions

These five predict future behavior instead of measuring past experience:

16. How likely are you to recommend [product] to a friend or colleague? (0–10 scale) — The NPS question, introduced in Frederick Reichheld’s Harvard Business Review article. Its real power is the follow-up: ask “why?” and read every answer from anyone who scored 6 or below.
17. How likely are you to purchase from us again? (1–5 scale)
18. If [product] disappeared tomorrow, how disappointed would you be? (Very / Somewhat / Not at all) — The Sean Ellis question. “Very” from 40%+ of users means you’ve built something sticky.
19. Have you recommended us to anyone in the past three months? (Yes / No)
20. What would make you switch to a competitor?

Open-ended questions

Five questions with no scale at all. Fewer people answer these, but the answers are where the surprises live:

21. What almost stopped you from buying? — My single favorite survey question. The answers are your objection list, written by the people who overcame the objections.
22. What’s one thing we could do to improve?
23. What do you wish [product] did that it doesn’t?
24. Describe [product] in one sentence to someone who’s never heard of it.
25. Is there anything else you’d like us to know?

Post-purchase timing questions

Ask these five within a week of delivery, while the memory is fresh:

26. How easy was the checkout process? (1–5 scale)
27. Did your order arrive when you expected? (Yes / Earlier / Later)
28. How accurate was the product description compared to what you received? (1–5 scale) — Low scores here are a returns forecast. Fix the listing before you fix anything else.
29. How easy was it to get started after delivery? (1–5 scale)
30. Would you buy this specific product again? (Yes / No / Not sure)

Scale vs multiple choice vs open text

Match the format to the job. Scales (1–5, 0–10) are for anything you want to trend over time — satisfaction, ease of use, likelihood to recommend. Multiple choice is for facts with known options — wait times, yes/no resolution, delivery timing. Open text is for the why.

The ratio I use: mostly scales and multiple choice, one or two open text boxes, never more. Open text is high effort for the customer, so spend it where it counts — usually right after your NPS question.

If you only want one number and nothing else, a single-question widget does the job better than a survey. The polls feature covers that case: one question, one click, running tally.

Keep it short: 5–8 questions per survey

Completion drops as length grows — every added question costs you finished responses, and the people who quit mid-survey skew toward your least patient (and often least happy) customers. The pattern I’ve seen across client sites: surveys under 8 questions get finished, surveys over 15 mostly don’t.

So don’t send the 30. Pick 5–8 that match one goal. A post-purchase survey pulls from groups one and six. A support follow-up pulls from group three plus the NPS question. Next quarter, rotate.

One trick that keeps surveys short without losing depth: only show the “why?” box when the score is low. With conditional logic, a 1–3 rating reveals the open text field and a 4–5 doesn’t. Happy customers finish in 40 seconds; unhappy ones get room to vent.

Build it in WordPress

A satisfaction survey is a form: radio buttons for scales, a textarea for open text, a submit button. Core Forms stores every submission in your database, so you’re not reconstructing results from notification emails.

Start from the form templates so the field markup is done, swap in your questions from the bank above, and add the conditional follow-up. For a working example of surveys driving real decisions, the course creators piece shows the same pattern applied to student feedback.

Ship it to a small segment first. If question 3 confuses people, better to learn that from 50 recipients than 5,000.

FAQ

How many questions should a customer satisfaction survey have?

Five to eight. Completion rates fall as surveys get longer, and drop-offs skew toward unhappy customers — the ones you most need to hear from. Keep one goal per survey, pull questions from a larger bank, and rotate topics across quarters instead of cramming everything into one send.

What’s the difference between CSAT and NPS?

CSAT measures satisfaction with a specific experience (“How satisfied are you with your recent purchase?”) on a 1–5 scale. NPS measures loyalty (“How likely are you to recommend us?”) on a 0–10 scale. CSAT tells you how a transaction went; NPS predicts whether the relationship lasts.

When should I send a customer satisfaction survey?

Within a week of the moment you’re asking about — after delivery, after a support ticket closes, after onboarding ends. Memory decays fast, and answers about last month’s experience are mostly reconstruction. Relationship-level surveys (NPS) work on a schedule instead: every quarter or twice a year.

Should survey responses be anonymous?

For honest criticism, yes — anonymity raises candor, especially on support questions. The tradeoff is you can’t follow up with unhappy respondents. My compromise: make the email field optional, and say why you’re asking. Roughly half will fill it in voluntarily.