15 customer feedback questions and how to use the answers

“How was everything?” can produce a score without telling you what to change. “How easy was it to arrange your delivery?” identifies a particular part of the experience. The right question depends on whether you need to fix a current problem, understand a recurring issue or check whether a recent change helped.

Choose the experience and the action first

Write down who will answer, which interaction they are rating and what you can change. For example: “Ask customers whose furniture arrived this week whether arranging the delivery was difficult, then review the booking instructions.” That gives you a more useful brief than “Measure happiness.”

Nielsen Norman Group’s feedback guidance  (opens in a new tab) recommends asking after the task, keeping the request short and leaving room for an explanation. Do not interrupt checkout to ask how checkout went. For a service with delayed results, choose a moment when the customer can reasonably judge what you are asking about.

If your goal is choosing a collection platform, start with customer feedback tools. This guide focuses on the questions and the review process. Our broader survey question examples cover events, training and teams as well.

Questions after a purchase or delivery

These help distinguish the product from buying and receiving it. Use the bracketed context to name the actual order or experience, without asking customers to enter information you already have.

1. How easy or difficult was it to find the information you needed before buying [the item]?

   Use five choices from Very difficult to Very easy, with Neither easy nor difficult in the middle. Include Not applicable if they did not look for information.

2. Was the total price clear before you placed the order?

   Use Yes / Partly / No / Don’t remember. This asks about price clarity, not whether the item was affordable.

3. Did the delivery arrive within the window we gave you?

   Use Yes / No / No window was given / Don’t remember. If delivery time is already recorded, use this to explore the promise the customer understood.

4. What, if anything, was different from what you expected?

   Optional open text. This can surface a mismatch in the item, packaging or communication without assuming there was a problem.

5. What is one thing we should change about ordering from us?

   Optional open text. Keep it about the ordering experience rather than asking customers to redesign the entire business.

Questions after a support conversation

Resolution, effort and clarity are different measures. A friendly conversation can leave a problem unresolved; a solved problem may still have taken too much work.

6. Was the issue you contacted us about resolved?

   Use Yes / Partly / No / Still waiting. Include the last option when the process genuinely takes more than one interaction.

7. How easy or difficult was it to explain your issue?

   Use the same five-choice ease scale as question 1. This can help identify an intake problem before judging the support team’s response.

8. Did you have to repeat information you had already given us?

   Use Yes / No / Not sure. If you need to know where the repetition happened, add an optional follow-up rather than assuming it.

9. How clear is what happens next?

   Use Not at all clear / Slightly clear / Moderately clear / Very clear / Extremely clear, plus Nothing else needs to happen.

10. What would help you finish resolving this issue?

    Optional open text. If everyone receives the same form, add “If it is not resolved” to the question or let resolved customers skip it. Conditional routing is not available in whatayarn yet.

Questions after an appointment or service

Use these for a completed consultation, lesson, installation or other named service. Avoid asking for an outcome the customer cannot assess yet.

11. How satisfied or dissatisfied were you with [today’s appointment]?

    Use Very dissatisfied / Somewhat dissatisfied / Neither satisfied nor dissatisfied / Somewhat satisfied / Very satisfied.

12. Did the service cover what you understood was included?

    Use Yes / Partly / No / Not sure. A mismatch is a reason to investigate expectations, not proof that either party is wrong.

13. Which part of the service was most useful to you?

    Optional open text, or a single-choice list of real service elements with None and Other available.

14. What should we do differently before your next appointment?

    Optional open text. Ask this when a next appointment is relevant; otherwise name a future service experience more generally.

15. Would you like someone to follow up about your experience?

    Use Yes / No, with an optional email field if you need contact details. Explain how that address will be used. An optional email does not by itself make the rest of a response anonymous.

Pew Research Center’s question-writing guidance  (opens in a new tab) explains why wording and response options matter. Keep a question about ease separate from a question about satisfaction, and keep wording stable when comparing repeated checks.

A short feedback survey you can actually send

For a small service business, start with these three fields:

• How satisfied or dissatisfied were you with your appointment today? Use the five labelled satisfaction choices above.
• What would you most like us to know about that experience? Optional text.
• Email, if you would like a reply. Optional email, with a clear follow-up explanation.

That is a constructed example, not a standardised customer-satisfaction instrument. Add a service category only if it helps you interpret the answers. If you need a repeatable CSAT measure, use the CSAT starter and document which answers you count as satisfied.

The customer feedback template below is another starting point: it uses a five-star experience rating, written comments and an optional reply address. It is not identical to the labelled satisfaction example above. Edit the rating wording and format to fit the measure you want.

Invite feedback without asking for praise

Here is an original request you can adapt:

Thanks for visiting us today. We’d like to know what worked and what we should change. There are three questions, and the comment and email fields are optional. We read the answers when planning next week’s service. [Share your feedback]

Replace the bracketed text with the form’s live link. Name a review schedule only if you can keep it. Do not promise a reply to every person if you do not have a process for doing that.

Offer the same neutral invitation to the customers whose experiences you want to understand, rather than only the ones you expect to be happy. A voluntary sample can still miss people: AAPOR’s best practices  (opens in a new tab) treat sampling and transparent reporting as part of survey quality. Record who was invited and who responded before treating a score as representative of all customers.

Turn the answers into a review queue

First distinguish a customer needing help from an idea for later improvement. The following comments are constructed examples:

In whatayarn, read ratings and comments together in Responses. Standard AI columns can help organise selected written answers with Sentiment, Urgency, Spam and Needs reply. They do not listen to recordings, inspect attachments or automatically send a reply. Review the underlying answer and correct a mistaken classification before using it to prioritise work.

For each issue worth acting on, record an owner, the next step and when you will review it. A task list or customer-support tool can hold that work; whatayarn’s integrations are coming soon, so moving answers into another system is a manual process today. CSV export is available for that handoff.

Close the loop and check the change

Tell a customer what you did when their comment prompted a specific fix. An original example reply is: “You mentioned that the parking instructions were missing. We’ve added the entrance photo and parking details to the appointment email. Thank you for pointing it out.” Only say this after the change happened.

Keep internal feedback separate from a public testimonial. If you want to publish a customer’s words, use a separate process that asks for permission; the testimonial template includes a sharing question.

After a change, repeat the relevant question with the same wording and scale. Look at both the score distribution and comments, along with who answered. A change in score alone does not establish that your update caused it. If you need to investigate an experience in depth, a conversation or usability test may answer more than another survey; NN/G’s survey guidance  (opens in a new tab) explains the limits of asking people to predict or reconstruct behaviour.