Good customer satisfaction survey questions are short, ask about one thing at a time, and use the same wording and scale every time you send them. The first two habits get you accurate feedback. The third is what lets you measure consistency, because a score only means something when you can compare it with last month’s.
One good rating tells you about one conversation. The same question after every ticket for six months tells you whether customers get that experience, no matter who answers or how busy the queue is.
Below are 22 customer satisfaction survey questions you can use as written, grouped by what they measure. You’ll also find two copy-ready survey templates, the common mistakes that skew results, and a way to save your questions as Snippets so every agent sends the same version.
Key takeaways
- Pick the metric before you write the question. CSAT measures satisfaction with an interaction, CES measures how easy it was to get help, and NPS measures how likely customers are to recommend you.
- Ask about one thing per question, on a scale you can track. Yes-or-no and two-part questions are hard to compare over time.
- Keep post-ticket surveys to five questions or fewer, and send them soon after the ticket closes.
- Consistency comes from repetition. Use the same wording, scale, and order every time so a change in the score reflects a change in service.
- Save survey requests and follow-up replies as shared Snippets so every agent sends the current version.
First, a quick look at the three metrics most customer satisfaction surveys are built on.
How CSAT, CES, and NPS surveys differ
Most customer service surveys use one of three standard customer satisfaction metrics. Each one answers a different question about the customer experience, so the metric you choose decides what you can learn.
| Metric | What it measures | Standard question | Scale | How it’s calculated | When to send it |
|---|---|---|---|---|---|
| Customer satisfaction score (CSAT) | Satisfaction with a specific interaction, product, or service | How satisfied were you with [interaction]? | 1 to 5, Very dissatisfied to Very satisfied | Share of 4 and 5 responses out of all responses | Soon after a ticket closes or a purchase |
| Customer effort score (CES) | How easy it was to get an issue handled | [Company] made it easy for me to handle my issue. | 1 to 7, Strongly disagree to Strongly agree | Average score, or share of 5 to 7 responses | After a support interaction or a self-service visit |
| Net Promoter Score® (NPS®) | Likelihood to recommend is a signal of customer loyalty | How likely are you to recommend [company] to a friend or colleague? | 0 to 10 | Percentage of promoters (9 to 10) minus percentage of detractors (0 to 6) | Quarterly or twice a year |
Customer effort deserves more attention than it usually gets. In Zendesk’s CX Trends 2026 survey of more than 6,000 consumers in 22 countries, 86% said responsiveness and accurate resolution strongly affect their purchase decisions. That echoes the 2010 Harvard Business Review research that introduced CES, which found effort was a better predictor of loyalty than satisfaction or NPS. Most customers want a quick, simple fix more than a memorable one.
CSAT and NPS still earn their place. CSAT tells you how a specific interaction went, and NPS tracks the broader relationship across the customer journey. A common setup is CSAT or CES after each ticket and NPS once or twice a year.
5 Customer Satisfaction (CSAT) survey questions
CSAT questions are the most direct way to measure customer satisfaction. They ask how satisfied customers are with a specific interaction, customer service team member, product, or service. Use a 5-point scale from “Very dissatisfied” to “Very satisfied,” and keep the labels consistent across all surveys.
1. How satisfied were you with your recent customer service experience?
This is the core CSAT question. Send it soon after the ticket closes, while the conversation is still fresh, and keep the wording the same from survey to survey. If you change it, this quarter’s CSAT score won’t compare cleanly with last quarter’s.
2. Did we resolve your issue?
Offer Yes, No, and Partly as answers. A customer can be happy with a friendly reply and still not have a fix. Pairing this question with question 1 separates how the conversation felt from whether the customer got their issue resolved, and “Partly” captures cases where you solved one part but missed another.
3. How would you rate the knowledge of the customer service representative who helped you?
Use the same 1 to 5 scale. This question separates the person from the policy. Customers sometimes rate an interaction low because they didn’t like the answer, such as a refund they couldn’t get. If knowledge scores stay high while CSAT drops, look at the policy before you look at coaching.
4. How satisfied were you with how quickly we responded?
Your help desk already tracks response time. This question tells you whether that speed met customer expectations, and the answer often depends on the channel. A two-hour reply may be fine for email and slow for live chat.
5. How satisfied are you with [product or service] overall?
This is a relationship question, so ask it less often, such as once a quarter. It measures overall satisfaction with what you sell and provides a baseline for comparing against the ticket-level scores from questions 1 to 4.
3 Customer Effort Score (CES) questions
CES questions measure how much effort it took customers to get what they needed. High effort shows up as repeat contacts and handoffs, and customers notice it even when every agent they talk to is helpful.
6. “[Company] made it easy for me to handle my issue.”
This is the standard customer effort score question. Replace [Company] with your company name and keep the rest of the wording as is. It’s a statement, so customers answer on a 7-point Likert scale from “Strongly disagree” to “Strongly agree.” Report CES scores as either the average or the share of customers who chose 5, 6, or 7, and stick with one method to keep the numbers comparable.
7. How many times did you contact us about this issue?
Offer 1, 2, 3, and 4 or more as multiple-choice options. Your help desk counts contacts within a single ticket, but customers who reach out across multiple channels often start a new ticket each time. Their answer shows effort that your data can miss.
8. Did you have to repeat any information to us?
Use Yes or No, with an optional text box that asks what they had to repeat. Customers notice when they have to start over. In Zendesk’s CX Trends 2026 research, 74% of consumers surveyed in 22 countries said they’re frustrated when they have to repeat information. A pattern of “Yes” answers usually points to handoffs between agents or customer support channels that lose context.
3 Net Promoter Score (NPS) questions
NPS measures how likely customers are to recommend you, which makes it a signal of customer loyalty more than a read on a single interaction. Bain & Company groups answers of 9 and 10 as promoters, 7 and 8 as passives, and 0 to 6 as detractors. Your Net Promoter Score is the percentage of promoters minus the percentage of detractors.
9. How likely are you to recommend [company] to a friend or colleague?
Use a 0 to 10 scale labeled “Not at all likely” at 0 and “Extremely likely” at 10. Keep this wording as written. Changing it makes it hard to compare your score with your past results or industry benchmarks.
10. What’s the primary reason for your score?
The number tells you where a customer stands, and this open-ended question tells you why. Make it optional and read the detractors’ answers first. They describe the pain points most likely to cost you customers.
No team makes every customer happy. Act on the reasons that come up repeatedly instead of on every individual complaint.
11. What’s one thing we could do to make you more likely to recommend us?
Asking for one thing keeps answers short and turns a score into meaningful, actionable feedback. This question is especially useful for passives, who scored 7 or 8. They’re close to recommending you, and their answer tells you what’s in the way.
4 questions that measure consistency over time
Satisfaction scores can appear steady even as customer relationships weaken. In the first quarter of 2026, the American Customer Satisfaction Index fell 0.3% to 76.7 out of 100 while customer retention rose. ACSI warned that “pent-up customer defection is brewing,” meaning some customers are staying for now despite lower satisfaction. That kind of gap can show up later as customer churn.
These questions show whether the experience holds up from one contact to the next, which a single CSAT score can’t tell you. For more on why that matters, see our guide to consistency in customer service.
12. Was this experience better or worse than your previous experiences with us?
Offer five answers from “Much better” to “Much worse,” plus “This was my first time.” A scale shows the direction of change, and the first-time option keeps new customers from guessing.
13. How consistent is the help you get from our team, no matter who answers?
Use a 1 to 5 scale from “Very inconsistent” to “Very consistent.” This question asks customers to judge your team as a whole. If scores dip while you’re hiring, it can be a sign that newer agents don’t yet have the same answers as experienced ones.
14. How many times have you contacted our support team in the past six months?
Offer ranges such as 1, 2 to 3, 4 to 5, and 6 or more. If your help desk already tracks contact history, skip this question and match responses to that data instead. That’s one less question for the customer to answer.
15. How does our service compare with other companies you’ve worked with?
Offer five answers from “Much better” to “Much worse.” Customers judge your service against every other company they deal with, and this question shows where your service quality lands in that comparison.
3 questions about a specific interaction
When a score drops, these questions help you find which part of the interaction caused it.
16. Were you kept informed about the status of your request?
Offer three answer options, “Yes, the whole time,” “Sometimes,” and “No.” A long wait is easier to accept when customers know what’s happening. This question shows whether your status updates are reaching people.
17. How many people on our team did you talk to about this issue?
Offer 1, 2, and 3 or more. Handoffs are often where context gets lost. Compare the answers with question 6 to see whether escalations and transfers make things harder for customers.
18. If you could change one thing about this experience, what would it be?
Leave a text box so customers can answer in their own words. Open text responses take longer to review than ratings, and they surface problems you didn’t think to ask about, which often makes them the most valuable feedback in the survey. Asking for one thing keeps the answer focused.
4 questions about your product or service
Support interactions and the product experience are closely linked, so many customer satisfaction surveys include a few questions about the product or service itself. Ask these less often than ticket questions, such as in a quarterly survey.
19. How well does [product or service] meet your needs?
Offer five answers from “Extremely well” to “Not at all well.” This is a broad product quality question about how well you meet customer needs. Add follow-up questions for low scores, such as what’s missing.
20. How would you rate the value for money of [product or service]?
Use a 1 to 5 scale from “Very poor value” to “Excellent value.” This question shows how customers see your pricing, separately from how they feel about the product.
21. Did [product or service] have the features you expected?
Offer Yes, Mostly, and No, with an optional text box. An earlier version of this article also asked whether those features were as easy to use as expected. That’s two questions in one, so ask about ease of use separately if you need it.
22. How likely are you to buy from us again?
Use a 0 to 10 scale, or 1 to 5 if you want it to match your CSAT questions. Repeat purchase intent is a direct measure of customer loyalty and is useful alongside NPS. Satisfied customers aren’t always loyal, and loyal customers aren’t always the ones who recommend you.
Customer service survey template you can copy
Here are two customer service surveys built from the questions above. The first goes out after a ticket closes. The second is a quarterly relationship survey for existing customers. Whatever tool you use to create surveys, save these as templates so the wording never changes.
Post-ticket survey (5 questions)
| Invitation (save as ;csat) Subject: How did we do with [ticket topic]? Hi [customer name], Thanks for contacting us about [ticket topic]. Would you answer a few quick questions about how it went? It takes about a minute to complete.[survey link] Thanks, [your name] |
Quarterly relationship survey (6 questions)
| Invitation (save as ;nps) Subject: A quick question about your experience with [company] Hi [customer name], You’ve been working with us for a while, and we’d like to hear how it’s going. This short survey has six questions and takes about two minutes.[survey link] Thank you, [your name] |
Send post-ticket CSAT surveys by email or in-app within a day of closing the ticket. Skip the quarterly survey for anyone who answered a post-ticket survey in the past few weeks, since fewer, better-timed surveys help protect your response rates.
Your survey tool sends the questions, but agents still type the invitations, the follow-ups, and the replies to unhappy customers. Those messages are worth saving once and reusing.
Ask the same questions, every time
Save your survey invitations and follow-up replies as Snippets. Type a short abbreviation and the full invitation appears in your help desk, email, or chat, with Fill-ins for the customer’s name and ticket number. Share them so everyone on the team asks the same way.
Get TextExpander for Free. No credit card required.
How to save survey questions as Snippets
Survey requests drift. One agent pastes an outdated link, and another rewrites the first question to make it sound friendlier. When the wording changes, the scores stop being comparable, and you lose the consistency you set out to measure.
Saving your survey requests as Snippets fixes the problem where it starts. Here’s how to set it up in TextExpander.
- Create a Snippet group called something like “Customer surveys” to keep survey requests and follow-up replies together.
- Save each survey invitation as its own Snippet, with an abbreviation such as ;csat for the post-ticket survey and ;nps for the quarterly survey. Copy the question wording directly from your survey tool to keep it identical.
- Add Fill-ins for the relevant information that changes, such as the customer’s name, ticket number, and topic. TextExpander prompts you for them when the Snippet expands, and the rest of the message stays the same.
- Share the group with your team. When the survey link or wording changes, update the Snippet once, and everyone sends the new version from then on.
- Save your follow-up replies, too, including a thank-you for promoters and a personal reply for detractors.
Snippets save time on the parts of each message that don’t change, and they expand in the apps your team already uses, so the same survey request works in your help desk, email, and live chat. The same approach works for live chat scripts and other replies your team sends every day.
Michael Purdin, Customer Success Team Lead at Netwrix, used TextExpander himself for years before rolling it out across his department. He says it “allows our engineers to use pre-approved templated responses to common issues.”
Common survey mistakes that skew your results
Most problems with customer satisfaction surveys come from the questions, not the customers.
Asking the right questions starts with avoiding these mistakes, so check for them before you send anything.
Asking two things in one question
A double-barreled question asks about two things but allows one answer. Pew Research Center calls these questions “difficult for respondents to answer” and notes that the answers are hard to interpret.
The original version of this article asked, “Did our product offer the features that you expected, and were they as easy to use as you expected?” A customer who liked the features but found them confusing has no right answer. Split it into two questions.
Using yes or no when you need a trend
Yes-or-no and other nominal questions, which sort answers into categories with no order, work for facts, like whether an issue was resolved. For opinions, they hide the difference between “slightly” and “very.” Our earlier question, “Was your last experience with our support team similar to this experience with our team?” could indicate that something changed, but not whether it improved or worsened.
Leading the customer
Leading questions tell customers what you want to hear, like “How great was your experience with our friendly team?” Ask “How would you describe your experience with our team?” and let the answer come from them.
Leave adjectives out of your questions, and balance your scale labels so the positive and negative sides carry equal weight.
Building the whole survey from agree-or-disagree statements
Likert-scale questions based on agree-or-disagree statements have a known weakness. Some respondents tend to agree with a statement no matter what it says, a pattern Pew calls acquiescence bias.
The CES statement is the exception worth keeping, because its standard wording lets you compare results. For everything else, ask a direct question with labeled choices.
Making the survey too long
Long surveys add to survey fatigue, and every extra question gives customers another reason to stop. Keep post-ticket surveys to five questions or fewer, and show a progress bar on anything longer.
Add a “Doesn’t apply” answer option so no one is forced to guess, since forced answers often lead to inaccurate responses.
Changing the wording or order between surveys
Pew found that earlier questions can shape how a respondent answers later ones. If you reorder or reword questions between surveys, a change in your score might be due to the survey instead of your service.
Lock the wording and order once you start tracking trends.
Asking about hypotheticals
Questions like “If we added new features, how would you feel?” ask customers to predict their own reactions, and the answers are hard to act on.
Ask about what they’ve already experienced, such as which features they use and what’s missing.
Asking for details you already have
Every demographic question, such as plan type or company size, is one fewer question about the experience. If that data is already in your help desk or CRM, connect it to responses instead of asking again.
Ask for details like role, job title, or income level only when you’ll use them to compare customer segments.
What to do with your survey results
Collecting responses is the easy part. The value comes from how your customer support team uses them to make informed decisions in the weeks after.
Track trends by segment
Review scores monthly and compare customer interactions by channel and issue type. A steady overall CSAT can hide a drop in a single channel or among a subset of your customer base. Look for lasting changes, not single bad weeks, and use what you find to identify areas for improvement and to coach the whole team instead of singling out one agent.
Read the open-ended answers
Ratings give you quantitative data, while open-ended survey responses provide qualitative feedback that explains the numbers and reflects customer sentiment. Tag them by theme, such as billing, wait times, or product bugs, and count how often each theme appears to identify patterns. Teams that leverage AI to sort open-ended responses still benefit from reading a sample themselves, because a tag can’t capture tone.
Follow up with unhappy customers
Reply to detractors, and to anyone who answered “No” to question 2, within a few days. Negative feedback is easier to recover from when you respond quickly.
Acknowledge the problem, say what you’ll do about it, and offer to talk. A personal reply makes customers feel heard, even when you can’t give them what they asked for. Save the reply as a Snippet with Fill-ins so it’s quick to send and still specific to the customer. These phrases to show empathy in customer service can help you get the tone right.
Share what you learn
Send product feedback to the product team with examples in customers’ own words. When the team ships a fix, tell the customers who asked for it so they know providing feedback made a difference.
Consistency in what agents send is part of keeping scores steady as volume grows. Fullscript’s support team maintained a 95.2% CSAT score while handling roughly 62,000 support interactions a month, and Director of Knowledge Management Rachel Montgomery credits that in part to how the team governs its messaging.
Even though our platform continues to get more complex, our CSAT really hasn’t dropped, and I think that is through that brand governance, messaging governance that we have, but it’s through tools like TextExpander that continue to make that easy.
Rachel Montgomery, Director of Knowledge Management, Fullscript
If your team is growing, see how TextExpander for customer support helps agents keep shared replies up to date as more people answer the same kinds of questions.
Start with three questions
Before you send anything, decide what you want to learn and what you’ll do with the answers. A survey without a clear objective collects scores nobody acts on. If you don’t have a customer service survey yet, start small. Send questions 1, 2, and 6 after every closed ticket for a month. That gives you satisfaction, resolution, and effort in less than a minute of the customer’s time.
Once you have a baseline, add NPS to a quarterly survey and read the open-ended answers every week. Keep the wording the same, and you’ll be able to see whether customers get the same experience every time they reach out.
Frequently asked questions
How many questions should a customer satisfaction survey have?
After a support interaction, one to five questions are enough. A short survey asks less of the customer, and a rating question plus an optional comment box gives you a score with context. Quarterly relationship surveys can run longer, around 6 to 10 questions.
How long should a customer satisfaction survey take?
Keep a post-ticket survey to about a minute and a relationship survey to a few minutes. In a SurveyMonkey analysis of roughly 100,000 surveys, abandonment rose for surveys that took more than 7 to 8 minutes, with completion rates dropping by 5% to 20%. Respondents also spent less time on each question as they moved through a survey, so questions near the end got the least attention.
What does it mean to close the feedback loop?
Closing the feedback loop means acting on what customers tell you and letting them know you did. For an individual customer, that's a personal follow-up after a low score or a complaint. For the business, it's fixing the recurring problems behind the scores and telling customers what changed. When customers see their feedback lead somewhere, they have a reason to keep answering your surveys.
How often should you survey the same customer?
Set a limit to avoid survey fatigue, so customers who contact you often don't get a survey after every ticket. A common approach is to send relationship surveys no more than once every 30 to 90 days and to skip post-ticket surveys for anyone who answered one recently. Many help desk and survey tools let you set these rules.
What's a good CSAT score?
Start by comparing against your own trend. Industry benchmarks vary by industry, channel, and how the score is calculated, so a number that's strong in one context may be average in another. A steady or rising score from the same questions, asked the same way, is the clearest sign your service is improving.
What's the difference between a customer satisfaction survey and a customer feedback survey?
Customer satisfaction surveys measure how happy customers are with the service provided, typically using a rating scale such as CSAT. A customer feedback survey is broader and can cover feature requests, usability, or market research. Many teams combine the two by adding one or two open-ended questions to a CSAT survey.
