Customer satisfaction survey: questions and examples

A customer satisfaction survey works best when the customer knows which experience you mean. Ask about the order they just received or the support conversation that just ended. "How satisfied are you with our company?" leaves the customer to choose the subject for you.

Keep one overall satisfaction question, then ask for the detail your team needs to investigate a low rating. A delivery survey might ask about the condition of the parcel. A support survey could ask whether the issue was resolved. They need different follow-up questions even when the rating scale is the same.

Start with a clearly labeled satisfaction scale

For a support interaction, try "How satisfied were you with the help you received for this request?" Offer Very dissatisfied, Dissatisfied, Neither satisfied nor dissatisfied, Satisfied, and Very satisfied. Keep the order consistent wherever you use the question.

Qualtrics describes a common CSAT calculation as the percentage of respondents who select the two satisfied choices on a five-point scale. Its CSAT measurement guide also distinguishes satisfaction with a specific experience from a broader customer relationship. Define your measure before collecting the responses so everybody reports it the same way.

Example questions after a purchase

1. How satisfied are you with the order you received?
2. Did the item match the description you saw before buying?
3. Did it arrive in the condition you expected?
4. What should we improve about this order experience?

Use a satisfaction scale for the first question, a short choice list for the next two, and an optional text area for the last one. Offer an appropriate answer for somebody who has not received or used the item. Otherwise a customer with an unresolved delivery problem may be forced to rate a product they have never seen.

Adjust the timing to what you ask. A survey about checkout can follow the purchase. A survey about the product needs to reach people after they have had a chance to try it. Avoid combining those stages into one score if different teams own the experience. When the questions are about the product itself rather than the purchase, the product survey question guide separates usage, testing, and concept research.

Example questions after a support request

• How satisfied were you with the help you received?
• Was your issue resolved: fully, partly, or not yet?
• What could we have done differently?

The resolution question helps separate a courteous conversation from a completed fix. A customer might like the agent and still need the original problem solved. Decide who follows up on unresolved cases, and make sure they can find the request without asking the customer to write their whole history again.

If you need a reference number or contact address, explain why. If you promise feedback without identity details, use a separate route for people who want a reply. The anonymous feedback guide explains how to describe that choice accurately.

Calculate CSAT with a defined denominator

For the five-point scale above, use:

CSAT = satisfied and very satisfied answers / valid satisfaction answers × 100

For example, suppose forty customers answer the satisfaction question and thirty-two choose Satisfied or Very satisfied. The score is 80 percent. That is an illustrative calculation, not a benchmark for your business. Record the forty responses alongside the percentage.

Exclude unanswered and Not applicable entries from this denominator, and state that rule in your report. An average of values from one to five is a different measure. Likewise, a recommendation question needs its own interpretation. Do not relabel an overall experience slider as CSAT or NPS merely because it returns a number.

Set up the questions in Thunder Forms

Open the Customer Feedback template and change the satisfaction label to the interaction you are reviewing. The template also has a recommendation question and an experience slider. Remove those if your survey only needs satisfaction and an explanation. The template anatomy article covers its actual fields and defaults.

Review responses in the dashboard or sync them to Google Sheets for your calculation. Thunder Forms collects the answers. Calculate and label the CSAT percentage in your reporting workflow. Submit both a positive and a negative test response to check that the saved labels match the categories your calculation counts.

Read the comments before choosing a change

Compare surveys of the same experience using the same question and scale. A delivery rating and a support rating may move for different reasons. Record when the wording or invitation changes so you can interpret later results with that context.

Assign comments about damaged items to the team that can investigate packing or delivery. Assign unresolved support cases to somebody who can contact the customer when contact details are available. The report should show the score, response count, and actions accepted for review, with responsibility for each action.

Write down the survey plan before sending it

A simple plan prevents the score from becoming detached from the experience. Record the event being measured, who receives the survey, when it is sent, the exact satisfaction question and choices, how unanswered and not-applicable entries are handled, and who reviews low scores. Keep those details with each reporting period if you compare results over time.

For example, a support team could send the survey after a ticket is closed, ask about that interaction, and route unresolved answers to the team lead for review. A different team measuring a purchase should wait until the customer can assess delivery. Mixing those answers into one trend hides whether a change came from support or fulfillment.

Use the response count when you present the score

Report the number of valid satisfaction answers next to the percentage, and keep the exact question wording. If 8 of 10 valid responses are in the satisfied group, report 80 percent from 10 answers. Do not compare that directly with a later score if the scale, audience, or experience changed. A small set of answers is a signal to investigate comments, not proof that every customer feels the same way.

Keep negative comments with the score they explain. A low rating about delivery timing calls for a different follow-up from a low rating about product instructions. Assign each issue to a team that can inspect the cause and note whether the team changed anything after the review.

For question banks by goal, from product testing to employee pulses, see the survey questions guide.