Customer satisfaction survey template

A customer satisfaction survey is useful when it tells you where the problem is, not only that there is one. This template pairs a single overall satisfaction question with a matrix that rates each part of the experience on the same five-point scale, so a low overall score arrives with its likely cause.

The survey beside this text is live. Fill it in and submit it — it validates exactly as it would for a real respondent, and saves nothing.

A matrix grid on one shared scale

Quality, value, support, speed and ease of use are rated in one table instead of five separate questions. Each row is answered on its own, and the Results tab draws one stacked bar per row so the weak area is visible at a glance.

A Likert question with an average

Overall satisfaction is a labelled five-point scale. Results show the count for every label and the average position on the scale, which is the number to track between rounds.

A follow-up that appears only when it applies

“What went wrong?” is shown only to someone who answers that their issue was not resolved. Everyone else never sees it, and a hidden question is never required.

No card required.

Customer Satisfaction Survey

Live preview

How to use this template

5 steps in the editor. Nothing here needs a paid plan.

  1. Press “Use this template” and rename the matrix rows to the parts of your own experience — keep it to five or fewer.
  2. Leave the five-point scale as it is. Changing the labels mid-collection makes earlier and later answers mean different things.
  3. Decide whether the improvement question is required. Optional gets more completed surveys; required gets more text.
  4. Under Settings → Appearance choose one question per page if the survey is mostly opened on phones — the matrix is easier to complete alone on a screen.
  5. Share the link, then compare the stacked bar for each matrix row on the Results tab: the row weighted furthest toward the low end of the scale is where to start.

Questions in this survey

6 questions, using 5 of the 24 question types available.

QuestionQuestion typeRequired
Overall, how satisfied are you with us?Likert scaleYes
How would you rate each of the following?Matrix gridOptional
If you contacted us, was your issue resolved?Radio buttonsOptional
What went wrong?Long textOptional
What is the one thing we should improve?Long textOptional
How long have you been a customer?DropdownOptional

What goes wrong with this survey

Specific to this template, not general advice about surveys.

Too many matrix rows

Past five or six rows people stop reading and tick the same column down the table. If an area matters, it deserves a row; if it does not, cut it rather than dilute the rest.

A matrix on a narrow phone

Five columns of labels need width. The table scrolls sideways rather than squashing, but a long scale is still work on a phone — keep the scale at five points and the labels short.

Averaging satisfaction across different customers

One average hides a happy majority and an unhappy segment. The customer-type question exists so you can split the CSV by it before drawing a conclusion.

Asking about things the customer never used

Someone who never contacted support cannot rate it and will pick the middle. If an area is optional for your customers, make the matrix not required so they can leave that row blank.

Questions about this survey

What is the difference between CSAT and NPS?

CSAT asks how satisfied someone is with a specific thing, usually on a five-point scale, and is read as an average or as the share who chose the top two answers. NPS asks how likely they are to recommend you, from 0 to 10, and is read as promoters minus detractors. CSAT is about an experience; NPS is about the relationship.

How is a matrix question exported?

As one cell per response, listing each row with the answer chosen for it — for example “Quality: Good; Support: Excellent”. Rows always appear in the order you set, so the column reads the same on every line.

Can respondents skip a row in the matrix?

Only if the question is not required. When it is required, every row needs an answer and the survey says “Please answer every row” rather than letting a half-filled table through.

Can I see results while the survey is still open?

Yes. The Results tab in your dashboard is live for you at all times. Whether respondents can see results is a separate setting under Voting & results, and it defaults to hidden until the survey closes.