Customer Satisfaction Survey Questions: 60+ Examples by Touchpoint

The best customer satisfaction survey questions are short, tied to a specific moment, and paired with one open-ended follow-up that asks why. A survey of five focused questions asked right after an experience will out-perform a twenty-question form every time, because people finish it and because each answer points at something you can act on. Below are more than sixty questions to choose from, grouped by touchpoint and by metric (CSAT, NPS and CES), plus the open-ended prompts that carry most of the value and the one rule that decides whether any of it is worth asking.

Pick the handful that match the decision you are trying to make. Do not ask all of them.

The three question types every satisfaction survey draws from

Almost every good customer satisfaction question is a version of one of three metrics. Knowing which is which keeps your survey focused.

Type	The core question	Scale	Ask it when
CSAT	How satisfied were you with [X]?	1 to 5	You want to know if one specific experience landed well
NPS	How likely are you to recommend us?	0 to 10	You want a relationship-level loyalty signal to trend
CES	How easy was it to get this done?	1 to 7	You are hunting friction in a support or self-serve flow

We cover how the three differ in detail in our guide to CSAT vs NPS vs CES. The short version: CSAT is a touchpoint metric, NPS is a relationship metric, CES is a friction metric. Most mature teams run all three, because they cost almost nothing to ask.

Customer satisfaction (CSAT) survey questions

Ask these right after the experience you want to measure, while it is fresh. The scale is usually 1 (very dissatisfied) to 5 (very satisfied), reported as the percentage who chose 4 or 5.

• Overall, how satisfied are you with [product/service]?
• How well did [product] meet your expectations?
• How satisfied are you with the quality of what you received?
• How satisfied are you with the value for the price you paid?
• How satisfied are you with your most recent experience?
• How would you rate [product] compared with what you expected before buying?
• How likely are you to buy from us again?
• What is the one thing we could do to make you more satisfied?

Customer service and support survey questions

Send these immediately after a support interaction closes. They are the highest-response-rate surveys you will run, because the experience is fresh and the customer already has your product on their mind.

• How satisfied were you with the support you received?
• Was your issue resolved?
• How easy was it to get your issue handled? (this is CES, a strong support question)
• How satisfied were you with the speed of the response?
• Did the person you dealt with understand your problem?
• How knowledgeable was the person who helped you?
• How many times did you have to reach out before your issue was solved?
• Is there anything about this interaction we could have done better?

A low support CSAT is rarely about the words of the survey. It usually points at slow or misrouted tickets, and making sure each request reaches the right person automatically fixes more of the score than any rewording will. Measure the experience, but be ready to fix the process behind it.

Net Promoter Score (NPS) questions

NPS is one core question and one follow-up. Do not add more; the follow-up is where the value is.

• How likely are you to recommend [company] to a friend or colleague? (0 to 10)
• What is the main reason for your score? (open-ended, and the most important question in the whole survey)
• What would it take to move your score closer to a 10?
• What do you value most about [product]?

Run NPS continuously on a sampled basis rather than as one big quarterly blast, and read every comment. Our NPS software page covers how to do that and what a good score actually looks like in context.

Customer effort (CES) questions

CES asks how hard a task was, usually as an agreement statement on a 1 to 7 scale. Use it on the flows where users drop out.

• [Company] made it easy for me to handle my issue. (strongly disagree to strongly agree)
• How much effort did you personally have to put in to get this done?
• How easy was it to complete [specific task]?
• What made this task harder than it needed to be? (open-ended)

Onboarding and first-experience questions

The first week is where most churn is decided, so it is worth its own short survey. Ask a few days in, once the customer has actually tried to get value.

• How easy was it to get started with [product]?
• Did you manage to do what you signed up to do?
• Was anything confusing or missing when you set up your account?
• How well did [product] deliver on what you expected when you signed up?
• What almost stopped you from getting started? (open-ended)

Product and feature satisfaction questions

• How satisfied are you with [specific feature]?
• How often do you use [feature]?
• How well does [feature] solve the problem you use it for?
• What is missing from [product] that would make it more useful to you?
• If you could change one thing about [product], what would it be? (open-ended)
• Which feature would you miss most if we removed it?

The open-ended questions that carry the value

Every rating question above is worth more when it is followed by an open-ended one. The number tells you something moved. The sentence underneath tells you what and why. These are the prompts worth including on almost any satisfaction survey:

• What is the main reason for your score?
• What is the one thing we could do better?
• What nearly stopped you from [buying / signing up / renewing]?
• What made this easy, or hard?
• Is there anything else you would like us to know?

Most teams add exactly one of these, get a few hundred thoughtful answers, and then never read them properly, because reading hundreds of free-text responses by hand is a job nobody has time for. That is the real bottleneck, and it is a solved problem now, without needing an enterprise suite to do it, as our Qualtrics alternative comparison sets out. Our guide on how to analyze open-ended survey responses covers the workflow, and survey analysis software can cluster those answers into themes automatically and trace each theme back to the exact responses behind it.

How many questions should a customer satisfaction survey have?

Fewer than you think. Response rates fall sharply with every extra question, and a form that takes more than a minute or two gets abandoned or answered carelessly. For a post-interaction survey, one rating question plus one open-ended follow-up is often the whole thing. For a periodic relationship survey, five to eight focused questions is a sensible ceiling.

The instinct to ask everything at once is what produces long, low-quality surveys. It is better to run several short, targeted surveys tied to specific moments than one giant questionnaire once a year. If you would rather start from something already assembled, our customer satisfaction survey template lays out a short form you can adapt, and the customer service survey questions collection covers the support-specific wording. You get higher response rates, fresher answers, and each result points at a decision instead of a vague overall mood.

When and where to ask

Timing changes the answer as much as the wording does. A few rules that hold up:

• Ask close to the moment. Post-support CSAT should fire when the ticket closes, not a week later. Onboarding questions should come once the user has tried to get value, not the second they sign up.
• Do not oversurvey. A single person should not see an NPS prompt, a CSAT prompt, and a feature survey in the same week. Sample and space them out.
• Match the channel to the moment. In-app surveys catch people while they are using the product; email works better for relationship surveys and post-purchase follow-ups.
• Always leave room to explain. The open-ended box is where the actionable content lives, so never run a satisfaction survey without one.

The one rule that decides whether any of this is worth it

You can pick the perfect questions, ask them at the perfect moment, and still get nothing out of it, because the answers pile up faster than anyone reads them. The single thing that separates a survey program that changes the product from one that produces slides is whether someone acts on the open-ended answers.

The rating tells you a score. Only the comments tell you why, and only the why is something you can fix.

That used to be a resourcing problem: coding a thousand free-text answers by hand took days, so it did not happen. It does not have to be manual anymore. Language models read open-ended responses, group the ones that mean the same thing, and quantify each theme, turning "CSAT is 72 percent" into "CSAT is 72 percent, dragged down by 214 people describing the same confusing billing step." The rule to insist on is traceability: any theme you are shown should link back to the exact answers behind it, so you can verify it before acting.

Choose a handful of the questions above, ask them at the right moment, and put your energy into reading what comes back. That is the part almost everyone skips, and it is the part that pays.