How to create a customer satisfaction survey: sample questions and tips

Asking your customers what they think rather than guessing is the most direct way to improve your product and service. A well-designed customer satisfaction survey helps you prevent churn, use your strengths in marketing and set priorities correctly. A poorly designed survey brings low response rates and misleading results. In this guide we cover the core satisfaction measures, sample questions, when and how to send the survey and how to use the results.

Define the survey's purpose

Every survey should seek the answer to one question:

• What is overall satisfaction?
• Would customers recommend us?
• How was a specific experience (purchase, support, delivery)?
• Why are customers leaving?
• Which new feature or service is wanted?

Writing questions before the purpose is clear leads to long, unfocused surveys.

Three core measures

1. CSAT (Customer Satisfaction Score)

Question: "How satisfied were you with [experience]?" (1–5) Calculation: Share of 4 and 5 ratings. Use: Right after a specific interaction (support, delivery, purchase).

2. NPS (Net Promoter Score)

Question: "How likely are you to recommend us to a friend or colleague?" (0–10) Calculation: Percentage of 9–10 (promoters) minus percentage of 0–6 (detractors). Use: Overall loyalty and brand relationship; see the NPS guide.

3. CES (Customer Effort Score)

Question: "How easy was it to [solve your problem / complete the process]?" (1–7) Use: Understanding how effortful processes are, especially in support and purchase flows.

Sample questions

Overall satisfaction

1. Overall, how satisfied are you with our product/service? (1–5)
2. How likely are you to recommend us? (0–10)
3. What is the main reason for your score? (open-ended)

Product and value

1. Which feature of our product is most useful to you?
2. What would you like us to improve in our product?
3. If our product did not exist, what would you use instead?
4. How would you rate the value you get for the price you pay? (1–5)

Service and support

1. How quickly did our support team solve your problem? (1–5)
2. How easy was it to solve your problem? (1–7)
3. How would you rate your communication with our team? (1–5)

Purchase and delivery

1. How easy was the purchase process? (1–5)
2. Did the product arrive when and in the condition you expected? (Yes/No)

Exit (churn) survey

1. What is the main reason you cancelled your subscription? (options + other)
2. What change would bring you back? (open-ended)

Rules for writing good questions

• Short and clear: One topic per question.
• Do not lead: Instead of "How did you find our great service?", "How would you rate our service?"
• Consistent scale: The same rating range within a survey.
• Limit open-ended questions: 1–2 are enough; the most valuable insights come from them.
• Do not ask unnecessary questions: Do not re-ask what you already know (email, purchase date).
• Total time: Ideally 2–3 minutes, 5–10 questions.

When to send?

• Post-transaction (CSAT, CES): Right after a support ticket closes, delivery or purchase.
• Relationship survey (NPS): Quarterly or every six months.
• After onboarding: When the customer first sees value.
• At cancellation: A short survey to learn why they leave.

Do not drown customers in surveys; surveying the same person often lowers response rates.

Raising the response rate

1. Keep it short and state the estimated time ("2 minutes").
2. Write a personal, friendly invitation; see email templates.
3. Use a mobile-friendly survey tool.
4. Embed the first question in the email (one-click answer).
5. Share how you will use the results and what you changed based on previous feedback.
6. If you use incentives, design them so they do not bias answers.

Analysing results

1. Track scores: Compare CSAT, NPS and CES over time and by segment (customer type, product, channel).
2. Categorise open-ended answers: Themes such as price, quality, support, ease of use. You can quickly group themes with AI, but read samples yourself. See market research with AI.
3. Prioritise: By frequency and impact; see collecting user feedback.
4. Take action: Each quarter, pick 2–3 concrete improvements and add them to the product roadmap.

Close the loop

• Contact low-scoring customers quickly; solving their problem can prevent churn. See reducing churn.
• Ask high scorers for reviews and referrals; see customer loyalty and referral.
• Announce improvements to customers.

Privacy and data

Survey answers may contain personal data. Provide a privacy notice, do not collect unnecessary personal data and store data securely; see the data protection checklist.

Common mistakes

• Long, tiring surveys
• Leading questions
• Collecting results and doing nothing
• Looking only at the average and missing segment differences
• Not following up with low scorers
• Surveying too often

FAQ

How many responses are enough? It depends on your customer count; even in small businesses a few dozen responses can show meaningful trends.

CSAT or NPS? They measure different things: CSAT the immediate experience, NPS overall loyalty. Use both.

Should it be anonymous? Anonymity increases honesty, but offer an optional contact field so you can follow up with low scorers.

Conclusion

A customer satisfaction survey is valuable when it has a clear purpose, short and neutral questions, the right timing and, above all, action on the results. Start small, measure regularly and show customers you listen. To measure product-market fit, see also the product-market fit guide.