# Customer Satisfaction Survey: 25 Ready-to-Use Questions

Customer satisfaction survey guide with 25 adaptable questions, eight rewrites, and three short templates for actionable customer feedback.

- Canonical URL: https://www.harmate.com/en/blog/customer-satisfaction-survey-25-ready-to-use-questions
- Author: Harmate Team
- Published: 2026-08-16
- Updated: 2026-08-16T03:10:01.494783+00:00
- Language: en

## Content

# Customer Satisfaction Survey: 25 Ready-to-Use Questions

A good customer satisfaction survey does not include every question you could ask. It includes the few questions that inform a specific decision: remove friction, improve a service, understand disappointment, or learn why a customer returns.

Below are **25 adaptable questions**, organized by decision rather than field type. Do not copy all of them. Select the questions whose answers already have a clear owner and next step.

## Before choosing questions, name the decision

Complete this sentence: “If the answers show that…, we will decide to…” If you cannot finish it, the question may produce an interesting number without supporting any action.

This prevents one form from becoming a customer satisfaction survey, market study, customer profile, and testimonial request at the same time. The guide to [actionable questionnaires](https://www.harmate.com/en/blog/actionable-questionnaires-start-with-the-decision-not-the-questions) explains how to work backward from a decision.

For each selected question, define:

- who will read the answer;
- what signal will trigger action;
- which segments are genuinely useful;
- how open-text answers will be reviewed.

## 25 customer satisfaction survey questions

Replace text in brackets with your context. The scales below are examples. Keep wording and labels stable when you compare results over time.

### Decision 1 — Measure the overall experience

These questions provide a summary signal. They do not explain the result by themselves, but they can track a comparable experience.

1. **Overall, how satisfied are you with [the experience]?**  
   Answer: a scale from “Not at all satisfied” to “Completely satisfied.”

2. **To what extent did [the product or service] meet your expectations?**  
   Answer: “Far below expectations” to “Far above expectations.”

3. **What is the main reason for your rating?**  
   Answer: a short text field immediately after the rating.

4. **Compared with your previous experience, was this experience better, the same, or worse?**  
   Answer: three options, shown only to returning customers.

5. **If you had to summarize this experience in one sentence, what would you say?**  
   Answer: open text.

A rating provides a trend; the follow-up explains where to look. Two customers can select the same score for entirely different reasons.

### Decision 2 — Locate journey friction

These questions identify where a problem occurred instead of merely recording dissatisfaction.

6. **Which step required the most effort?**  
   Answer: the real journey stages, plus “None” and “Other.”

7. **Was there a moment when you considered stopping?**  
   Answer: yes/no, followed by “What happened?” when yes.

8. **Was the information you needed easy to find?**  
   Answer: an ease scale with “Not applicable” where necessary.

9. **At what point did you need help?**  
   Answer: a journey-stage list or “I did not need help.”

10. **What would have made this step easier?**  
    Answer: open text focused on the selected stage.

Do not ask whether a journey was “simple and fast.” One answer cannot tell the team which of the two dimensions needs work.

### Decision 3 — Understand perceived value

Satisfaction depends on more than technical quality. It also reflects the job the customer wanted to accomplish and the expectations they brought to the experience.

11. **What goal did you want to achieve with [the product or service]?**  
    Answer: options identified through pretesting, plus “Other.”

12. **Were you able to achieve that goal?**  
    Answer: “Yes, completely,” “Partly,” “No,” and “Not yet.”

13. **Which part was most useful to you?**  
    Answer: one choice or short text, depending on how well you know the use cases.

14. **Which part was least useful to you?**  
    Answer: the same list, with “None.”

15. **What did you expect [the product or service] to do that it did not do?**  
    Answer: open text.

An open-ended question produces **far richer material** than a checkbox: it preserves the customer’s own language and surfaces unexpected causes, exceptions, and tensions that the researcher’s categories could hide. This is central to Harmate’s approach: let the experience emerge before compressing it into scores.

Closed questions remain useful for measurement, segmentation, and comparison. The formats are not interchangeable: use open questions to discover and understand, then close the question once the categories have evidence behind them and comparison becomes necessary. Pew Research Center shows that open and closed formats can surface very different answers ([Pew Research Center, *Writing Survey Questions*](https://www.pewresearch.org/writing-survey-questions/)).

### Decision 4 — Prioritize improvements

The goal here is to obtain a trade-off. An unranked wish list simply moves the prioritization problem to the team reading the survey.

16. **If we could improve only one thing, which should it be?**  
    Answer: one choice from tested themes, plus “Other.”

17. **Why is this point the priority for you?**  
    Answer: a conditional short-text field.

18. **What should we keep exactly as it is?**  
    Answer: open text or a short list.

19. **Which of these improvements would have the greatest effect on your experience?**  
    Answer: a limited ranking or up to two selections.

20. **In what situation would this improvement help you?**  
    Answer: open text focused on a concrete use case.

Do not offer improvements the team is not genuinely considering. The survey would create an expectation without offering a real decision path.

### Decision 5 — Prepare the next relationship step

These questions help distinguish a one-off issue from a lasting risk and organize follow-up with the customer’s consent.

21. **How likely are you to use [the product or service] again?**  
    Answer: a clearly labeled likelihood scale.

22. **How likely are you to recommend [the brand or service] to someone with a similar need?**  
    Answer: a 0-to-10 scale when tracking NPS, or another consistent scale for your purpose.

23. **What is the main reason for that answer?**  
    Answer: short text shown after the previous question.

24. **What could make you decide not to use [the product or service] again?**  
    Answer: open text or a list derived from earlier feedback.

25. **May we contact you to follow up on your answer?**  
    Answer: yes/no, with contact details collected separately when needed.

Stated intention is not a guaranteed prediction of future behavior. Treat it as a follow-up signal and compare it with observed use or renewal when possible.

## Eight survey questions to avoid—and how to rewrite them

A peer-reviewed review catalogued dozens of potential biases across individual questions, the questionnaire, and its administration ([Choi and Pak, 2005](https://pmc.ncbi.nlm.nih.gov/articles/PMC1323316/)). For a typical satisfaction survey, start with these eight visible problems.

| Avoid | Why it fails | Better version |
|---|---|---|
| “Are you satisfied with our excellent service?” | The adjective leads the answer | “How satisfied are you with the service?” |
| “Was the service fast and effective?” | It combines two concepts | Ask about speed and effectiveness separately |
| “Did you like your experience?” | “Like” is vague | Name the experience and use a defined scale |
| “Why was the price too high?” | It assumes the diagnosis | “How would you rate the value you received for the price?” |
| “Did you always receive an immediate answer?” | Absolutes make recall unreliable | Ask about the latest episode or a defined period |
| “Which features do you want?” | A wish list does not reveal the job | “What goal can you not accomplish today?” |
| “Provide your age, gender, city, and job” | The data may not support a decision | Ask only for a genuinely necessary segment |
| “Any other comments?” | The invitation is too broad | “What important point did we not ask about?” |

Closed questions should provide understandable, non-overlapping, sufficiently complete options. Each question should cover one concept. Pew Research Center identifies both principles as foundations of sound question wording.

## Three short satisfaction survey templates

These sequences are not universal templates. They show how to select questions from the bank without sending all 25.

### After a purchase or service

1. Overall satisfaction.
2. Goal achievement.
3. Highest-effort step.
4. Most useful element.
5. Priority improvement.
6. Main reason for the rating.

### After customer support

1. Issue resolved completely, partly, or not resolved.
2. Ease of getting help.
3. Clarity of the answer.
4. Most difficult step.
5. Change that would have simplified the experience.
6. Permission to follow up if the issue remains.

### For an ongoing relationship

1. Overall satisfaction during a defined period.
2. Change since the previous measurement.
3. Current main goal.
4. Most useful element.
5. Priority improvement.
6. Intention to continue.
7. Open explanation of that intention.

Training feedback requires a different question set and schedule. The dedicated guide to the [training satisfaction questionnaire](https://www.harmate.com/en/blog/training-satisfaction-questionnaire-why-your-results-are-often-unusable) covers that use case.

## How to make the survey usable

### Choose the right moment

Send the survey close enough to the experience for customers to recall concrete events. When comparing periods, keep timing and channel reasonably consistent. A survey link open to anyone, a customer email, and an in-product prompt do not necessarily reach the same people.

### Collect only what is necessary

Every question adds effort and may collect personal data. The CNIL’s guidance on data minimization says personal data should be adequate, relevant, and necessary for the defined purpose ([CNIL, *Minimiser les données collectées*](https://www.cnil.fr/fr/minimiser-les-donnees-collectees)). Do not request a full demographic profile “just in case.”

### Pretest with people similar to the audience

Ask a few people to explain what they think each question means, how they chose an answer, and which option was missing. AAPOR recommends pretesting, including cognitive interviews, before fielding a survey ([AAPOR, *Best Practices for Survey Research*](https://aapor.org/standards-and-ethics/best-practices/)).

### Plan the analysis before sending

Decide how you will handle ratings, missing responses, “Not applicable,” and open text. Do not keep only comments that support the team’s initial belief. The guide to [actionable open-ended questions](https://www.harmate.com/en/blog/open-ended-questions-get-actionable-data-without-bias) provides a focused method.

## What a template can accelerate—and what it cannot decide

A template removes the blank page and provides a starting structure. It does not know your decision, audience, or exact journey moment. Every question still needs adaptation, review, and pretesting.

Harmate’s [public template library](https://www.harmate.com/en/product/template-library) lets teams browse FR/EN questionnaire templates, preview their questions, and adapt them. The public catalog is a working base; it does not turn a generic template into a valid instrument for every context.

## Frequently asked questions

### How many questions should a satisfaction survey include?

There is no universal ideal. Ask only what supports a decision, then test completion time and comprehension with the actual audience. Six relevant questions can be more useful than an unfiltered list of 25.

### Which satisfaction scale should I use?

Choose a scale respondents understand and that you can keep stable over time. Label its endpoints, provide “Not applicable” when legitimate, and avoid reversing direction from one question to the next.

### Should the survey be anonymous?

It depends on the topic and expected follow-up. Anonymity may make sensitive feedback easier, while an identified survey allows a team to resolve an individual problem. Explain who will read the answers, why the data is collected, and whether follow-up is possible.

### When should I send the survey?

After a concrete experience: delivery, product use, support interaction, or completion of a journey stage. Avoid waiting so long that customers must reconstruct their memory. For trend comparisons, use a comparable schedule.

## Pre-send checklist

- [ ] Every question supports a named decision.
- [ ] The survey does not mix unrelated objectives without necessity.
- [ ] Each question covers one concept.
- [ ] Scales and options have clear labels.
- [ ] “Not applicable” or “I don’t know” appears where legitimate.
- [ ] Personal data is limited to what is necessary.
- [ ] Timing and delivery channel are defined.
- [ ] The survey has been pretested with people similar to the audience.
- [ ] The team has a plan for open-text answers.
- [ ] Someone owns the follow-up action.

## Key takeaway

Do not paste 25 questions into one form. Copy the questions that inform the next decision, adapt them to the real journey, and remove anything the team cannot use. A useful satisfaction survey is defined less by the amount of feedback collected than by the clarity of what happens next.