How to Create a Customer Satisfaction Survey

To create a customer satisfaction survey, start with the decision you need to make. Then choose a small set of questions, decide when to send them, and pair a score with one open field.

If you want to review the basics first, open the feedback and survey glossary.

Define the goal

Write down what you want to learn and what could change because of the answers. “Find out whether support resolved the issue” is more useful than “measure satisfaction” because it guides the questions and analysis.

Choose one goal per survey. When product, support, delivery, and the broader relationship compete for attention, the questionnaire grows and the answers lose context.

Choose the question type

Scores help track change. Open-ended questions reveal reasons, language, and situations the answer options did not anticipate. Multiple choice works when the alternatives are known and do not overlap.

Combine formats around the decision. A score followed by “What was the main reason for your answer?” usually gives both direction and context without demanding a long response. Compare NPS, CSAT, and CES when you are choosing a metric.

Keep the survey short

Keep only questions you plan to analyze. Give the survey a clear title, explain why the response matters, and mark nonessential fields as optional.

The customer satisfaction survey template is a starting point you can trim. For a recent interaction, use the customer service survey questions. The question generator helps adapt the flow to a specific goal.

Choose the moment and channel

Send when the person can evaluate the experience, but before the details fade. The right moment depends on what happened: a support case closed, an order delivered, or a step completed.

Choose a channel that fits the existing relationship and make the invitation acknowledge the context. Keep the same rule in future rounds so equivalent surveys stay comparable.

Analyze themes and scores

Look at the score distribution, not just the average. Then group comments by theme and return to the original wording before creating a category.

Track the response rate as well. People who answer may have a different experience from people who ignore the invite, so the result has a limit.

Close the loop

Share the learning with the people who can act and record what will be investigated. When appropriate and authorized, follow up with the person who raised a problem or suggestion.

Closing the loop does not mean promising every requested change. It means showing that the feedback was read, explaining the next possible step, and setting honest expectations.

Common mistakes

• Starting with questions before defining the goal.
• Combining several experiences in one score.
• Writing answer choices that push people toward one response.
• Making every field required.
• Comparing different audiences, channels, or moments without context.
• Collecting responses without assigning an owner for analysis.

Next step

Write the goal, choose one score and one open question, and check that every field supports a real decision.

Build my customer satisfaction survey