Customer Satisfaction Survey Template: 8 Examples and Samples You Can Copy

A customer satisfaction survey template is a complete, ready-to-send survey built for one specific moment, not a general list of questions. The strongest templates share a shape: one rating question about a single recent experience, one open-ended follow-up that asks why, and nothing else. Everything below is written out in full so you can copy it, change the product noun, and send it this week.

Last updated: August 2026.

Most survey advice hands you a question bank and wishes you luck. That is useful when you are designing a program from scratch, and we have a full bank of customer satisfaction survey questions for exactly that. But it is not what you need at four o'clock on a Thursday when someone has asked you to start measuring support quality by Monday. Then you want a survey, finished, that you can paste into your help desk.

So here are eight. Each is tied to a trigger, because a survey without a trigger is a survey nobody answers. Each is short, because length is the single biggest predictor of whether people finish. Pick the one that matches the moment you care about and ignore the rest.

How to choose which survey template to use

The choice is not really about the metric. It is about what just happened to the customer, because that determines what they can accurately tell you. Somebody who closed a support ticket two minutes ago can tell you about that ticket. They cannot tell you whether they would recommend your company, and if you ask, you will get a number shaped by the last five minutes rather than the last five months.

If you can only run one, run the first. Post-ticket CSAT is the cheapest to set up, the easiest to act on, and the only one that tells you which individual interaction went wrong while the details are still recoverable.

Customer satisfaction survey examples: the eight templates

1. Post-ticket CSAT survey template

Send it automatically when a ticket moves to solved, from the same address the conversation happened on, so it reads as a continuation rather than a marketing email.

Subject: How did we do with your request?

Hi [First name], we have marked your request about [ticket subject] as resolved.

How satisfied were you with the support you received?
Very satisfied / Satisfied / Neutral / Dissatisfied / Very dissatisfied

What is the main reason for your score? (optional, one box)

Thanks, that is the whole survey.

Two questions, one of them optional. The reason box is where the value is: the score tells you there is a problem, the sentence tells you what it was. Referencing the actual ticket subject lifts response rates noticeably, because the customer does not have to work out which of their four open issues you mean.

2. Post-chat survey template

This is the one most teams get wrong, because they reuse the email survey inside a chat widget. A post-chat survey has to be answerable in the two seconds before the customer closes the tab, so it is one tap, then an optional box that appears only after they tap.

Shown in the widget the moment the chat ends:

How was this chat? 👍 / 👎

If thumbs down: Sorry about that. What went wrong? [one text box, optional]

If thumbs up: Thanks! Anything we could have done better? [one text box, optional]

Binary beats a five-point scale in chat. The customer is one click from leaving and a scale asks them to make a finer judgment than they want to make. If you need CSAT to be comparable with your email survey, run the same five-point scale, but expect a lower completion rate, and never put the comment box on the first screen. Chat surveys generally outperform email because they arrive while the customer is still present, which is also why the timing has to be immediate. A post-chat survey sent as an email an hour later is just a slow email survey. If you are also writing the conversations themselves, our live chat support scripts cover the closing lines that set up this survey properly.

3. Post-purchase survey template

Trigger this on delivery confirmation, not on order confirmation. Asking about an order the customer has not received yet measures your checkout, which is rarely what people mean when they say they want post-purchase feedback.

Subject: Your [product] arrived. How is it?

Hi [First name], your order [number] was delivered on [date].

How satisfied are you with your purchase? 1 to 5

How satisfied are you with the delivery experience? 1 to 5

Anything you want us to know? (optional)

Two ratings is the exception to the one-question rule, and it earns its place because product and delivery fail independently. A customer who loves the product and hated the courier gives you a middling single score that hides both facts. Splitting them tells you which team has the problem.

4. Onboarding milestone survey template

Onboarding is a process, not a moment, so measure effort rather than satisfaction. Send it when the customer completes setup, not when your internal checklist says onboarding is done, because those are usually different weeks.

Subject: Two questions about your setup

Hi [First name], you finished setting up [product] this week. Two questions, then we will leave you alone.

Getting set up with [product] was easy.
Strongly disagree / Disagree / Somewhat disagree / Neutral / Somewhat agree / Agree / Strongly agree

Which step took the longest? (optional)

The second question is the one that pays. Aggregate a quarter of answers and you get a ranked list of the friction in your own onboarding, sorted by how often customers noticed it. That is a far more actionable artifact than a satisfaction average, and it maps directly onto the stages in your customer onboarding process.

5. B2B relationship survey template (sample client satisfaction survey)

This is the annual or semi-annual survey for accounts with a named contact, the closest thing to a traditional client satisfaction questionnaire. It is longer than the others because the person answering has a relationship with you and will spend two minutes on it.

Subject: Your view on how [our company] is doing

Hi [First name], we are doing our [half-year] check with the accounts we work with most closely, and yours is one of them. Five questions, about two minutes.

1. Overall, how satisfied are you with working with us? 1 to 5

2. How satisfied are you with the responsiveness of your main contact? 1 to 5

3. How satisfied are you with the accuracy of our billing and paperwork? 1 to 5

4. What is the one thing we could do that would make the biggest difference to you?

5. Is there anything we do that you would not want us to change?

Question three is unusual and it is deliberate. In B2B, the relationship is very often damaged by back-office friction (a wrong invoice, a contract that took three weeks to countersign) rather than by anything the account manager did, and nobody ever asks about it. Question five stops you optimizing away the thing your best clients quietly value. If the survey feeds a scheduled account conversation, the answers slot straight into a quarterly business review agenda.

6. NPS relationship survey template

Subject: One question, [First name]

How likely are you to recommend [company] to a friend or colleague?
0 (not at all likely) to 10 (extremely likely)

What is the main reason for your score?

Two questions and nothing else. The temptation to bolt on four more is strong and it is what turns a survey with a good response rate into one people ignore. Send it at relationship checkpoints, roughly twice a year, and never right after a support interaction. The NPS survey guide covers scoring and what the number is and is not good for.

7. Customer effort survey template

[Company] made it easy for me to handle my issue.
Strongly disagree (1) to Strongly agree (7)

What made it easy or hard? (optional)

Use this after anything multi-step: a return, a claim, a plan change, a migration. Effort predicts whether people come back better than satisfaction does, largely because a customer can be satisfied with the outcome and exhausted by the process, and only one of those makes them renew. The customer effort score guide has the scoring and the benchmarks.

8. Cancellation survey template

Before you go, why are you cancelling? (pick one)

• Too expensive for the value I got
• Missing a feature I need
• Too hard to use
• Had a bad support or billing experience
• My situation changed, no longer need it
• Switching to another provider
• Something else

Anything you would want us to know? (optional)

Fixed options, not a free text box, because you need to count these. One reason per customer, shown on the cancellation screen before the cancellation completes, never as an email afterwards. Include the "situation changed" option honestly: without it, customers who were never going to stay get miscoded as pricing objections and you spend a quarter discounting for no reason. The reasons feed directly into how you read your churn rate.

What response rate should you expect?

Published benchmarks for 2026 vary widely enough that you should treat any single figure with suspicion. Across vendor research the commonly cited ranges are roughly 10 to 30 percent for email surveys, higher for in-app and in-widget surveys because they catch the customer mid-session, and higher again for SMS in the narrow cases where it is appropriate. Different vendors report in-app rates anywhere from the low teens to sixty percent, which mostly tells you they are measuring different things: some count a survey shown, others a survey opened.

The one finding that repeats consistently across sources is the length effect: surveys of two or three questions get roughly double the completion of surveys with seven or more. That is a bigger lever than anything else on this page, and it is free. Compare your rate to your own last quarter rather than to a benchmark, because audience and channel move the number far more than survey quality does.

Survey design rules worth following

• One moment per survey. Ask about the thing that just happened, not about your company in general.
• Rating first, comment second. Putting the open box first collapses response rates, because it looks like work.
• Never make the comment mandatory. Forcing it produces "n/a" and "fine" at scale, which is worse than an empty field.
• Avoid double-barrelled questions. "How satisfied are you with the speed and quality of our service?" cannot be answered by anyone who felt differently about the two.
• Do not lead. "How would you rate our excellent support?" and "Was our agent helpful?" both nudge upward. Neutral wording costs you nothing and buys you honest data.
• Cap the frequency. One survey per customer per month, whatever the trigger, or your most active customers get punished for being active.
• Close the loop on detractors. A survey that generates no follow-up trains customers not to answer the next one.

What to do with the answers

Route them, do not report them. A low score with a comment should create a task for a named person within a day, not appear in a dashboard someone reviews monthly. The half-life of a recoverable complaint is short: a customer who told you on Tuesday that their invoice was wrong is reachable on Wednesday and gone by the following month.

Tag every comment so the free text becomes countable. Six or eight tags is plenty (billing, onboarding, product gap, speed, agent knowledge, pricing), and the ranking those produce after a quarter is what turns a survey program into a change program. That is the point of the whole exercise, and it is where most programs stall: the scores get collected, nothing operational changes, and the score stays flat. The fixes almost always sit in the back-office processes customers actually feel rather than in how the survey was worded.

The high scores are worth something too, and almost nobody uses them. When a survey comes back at five out of five with a specific sentence about what went well, that sentence is a testimonial the customer has already written for you. Ask permission and publish it as social proof while it is fresh; a request that quotes their own words back to them gets said yes to far more often than a cold ask months later.

Frequently asked questions

What questions should be on a customer satisfaction survey? One rating question about a specific recent experience, and one optional open-ended question asking why they gave that score. Everything beyond that trades response rate for detail. If you need more, add a second rating only when two things can fail independently, such as product quality and delivery.

How long should a customer satisfaction survey be? Two to three questions for transactional surveys and no more than five or six for an annual relationship survey. Reported data consistently shows short surveys getting roughly double the completion rate of surveys with seven or more questions, so every extra question has to earn its place.

What is a good customer satisfaction survey response rate? There is no single number worth chasing. Email surveys are commonly reported in the 10 to 30 percent range and in-widget surveys higher, but channel and audience move the figure more than quality does. The useful comparison is your own rate last quarter on the same trigger and channel.

When should you send a customer satisfaction survey? Immediately after the moment you are asking about, while the details are still fresh: when a ticket closes, when a chat ends, when an order is delivered, when a setup milestone completes. Relationship surveys such as NPS are the exception and belong on a calendar, roughly twice a year, well away from any support interaction.

What is the difference between CSAT, NPS, and CES surveys? CSAT asks how satisfied someone was with a specific interaction. NPS asks how likely they are to recommend you, measured at the relationship level. CES asks how much effort a process took them. They answer different questions and are not interchangeable; running one where another belongs is the most common survey design error.

Pick one template, wire it to its trigger, and send it this week. A short survey running on a real trigger will teach you more in a month than a perfect instrument that never ships, and every one of these can be adapted in about ten minutes. For the wording variations behind each question, and the full bank organized by what you are trying to learn, work from our customer satisfaction survey questions and examples, and see where these surveys sit among the wider customer experience metrics worth tracking.