Customer Satisfaction Score (CSAT): How to Calculate It, Survey Questions and Examples
What CSAT is, the formula with worked examples, 40 CSAT survey questions by touchpoint, when to ask, what a good score means, how to analyse the comments, and how CSAT compares with NPS and CES.
Quick answer
Customer Satisfaction Score (CSAT) measures how satisfied customers are with a specific interaction, product or experience. The usual question is "How satisfied were you with [experience]?" on a 1 to 5 scale. CSAT = (number of 4 and 5 answers ÷ total answers) × 100, giving a percentage of satisfied customers. Ask it right after the moment you want to measure, and always follow the score with "What is the main reason for your rating?"
CSAT is the most direct customer experience metric there is. It asks customers about one moment, while it is fresh: a support case, a delivery, an onboarding call, a checkout. That makes it ideal for finding which touchpoints work and which ones fail.
This guide covers what CSAT is, how to calculate it with worked examples, 40 survey questions you can use, when to send them, what a good score means, how to analyse the comments, and how CSAT compares with NPS and CES.
What is CSAT?
Customer Satisfaction Score is a metric based on a single question about a specific experience:
> "How satisfied were you with [your support experience / your delivery / your purchase today]?"
The most common scale has five points:
| Answer | Score |
|---|---|
| Very dissatisfied | 1 |
| Dissatisfied | 2 |
| Neutral | 3 |
| Satisfied | 4 |
| Very satisfied | 5 |
Some teams use 1 to 3 (often with faces or emojis), 1 to 7, or 1 to 10. Whatever you choose, keep the scale consistent so results are comparable over time.
CSAT is a transactional metric: it measures how a specific moment went, not the overall relationship. For loyalty to the brand as a whole, use Net Promoter Score.
How to calculate CSAT
The standard method counts only satisfied customers, those who answer 4 or 5 on a 5-point scale:
CSAT (%) = number of satisfied responses (4 and 5) ÷ total responses × 100
Worked example 1
After support cases closed last month, you received 250 responses:
| Answer | Responses |
|---|---|
| 5 – Very satisfied | 120 |
| 4 – Satisfied | 70 |
| 3 – Neutral | 25 |
| 2 – Dissatisfied | 20 |
| 1 – Very dissatisfied | 15 |
| Total | 250 |
Satisfied responses: 120 + 70 = 190. CSAT = 190 ÷ 250 × 100 = 76%.
Worked example 2: the same score can hide different stories
Split the same 250 responses by support channel:
| Channel | Responses | Satisfied (4–5) | CSAT |
|---|---|---|---|
| Live chat | 150 | 132 | 88% |
| 100 | 58 | 58% | |
| All | 250 | 190 | 76% |
The overall 76% looks reasonable. Email support is a problem. Always break CSAT down by channel, team, product, region or customer segment.
Average score method
Some teams report the average rating instead, for example 3.9 out of 5. It is easy to calculate but harder to interpret: an average of 3.9 can come from many satisfied customers and a few very unhappy ones, or from a lot of lukewarm answers. The percentage of satisfied customers is usually clearer. Whichever you use, state the method next to the number.
CSAT in a spreadsheet
| What | Formula (ratings in A2:A251) |
|---|---|
| Satisfied responses | =COUNTIF(A2:A251,">=4") |
| Total responses | =COUNT(A2:A251) |
| CSAT (%) | =COUNTIF(A2:A251,">=4")/COUNT(A2:A251)*100 |
| Average rating | =AVERAGE(A2:A251) |
What is a good CSAT score?
There is no single good number. CSAT depends heavily on:
- The touchpoint. A quick delivery confirmation and a complicated billing dispute will never score the same.
- Industry and expectations. Customers judge an airline, a bank and a fashion retailer against very different standards.
- Scale and method. "Percent satisfied" on a 5-point scale and an average on a 10-point scale are not comparable.
- Who answers and when. Surveying only some customers, or asking long after the experience, changes the result.
The most useful comparisons are with your own past results, between your own channels and teams, and with direct competitors if you have comparable data. A rising trend and a shrinking gap between your best and worst segments matter more than any absolute number.
When to send a CSAT survey
Ask right after the moment you want to measure, while the experience is fresh:
- After a support case closes (email, chat or call)
- After a purchase or checkout
- After delivery or installation
- After onboarding or a training session
- After using a new feature for the first time
- After an appointment, a stay or a visit
- After a return or refund is completed
Keep it short: the rating plus one open question. Send it in the channel the interaction happened in: an in-app prompt, a chat message, an email, an SMS or WhatsApp message, or a short call.
40 CSAT survey questions
Core questions (use one of these)
- How satisfied were you with your experience today?
- Overall, how satisfied are you with [product / service]?
- How satisfied were you with the help you received?
- How satisfied are you with your recent purchase?
Follow-up questions (open-ended)
- What is the main reason for your rating?
- What could we have done better?
- What did you like most about your experience?
- Was there anything that surprised you, good or bad?
- Is there anything else you would like to tell us?
Customer support
- How satisfied were you with how quickly we responded?
- How satisfied were you with the agent's knowledge?
- Was your issue fully resolved? (yes / no)
- How many times did you have to contact us about this issue?
- How satisfied were you with how clearly we explained the solution?
- How easy was it to reach the right person?
Purchase and checkout
- How satisfied were you with the checkout process?
- Did you find everything you were looking for? (yes / no)
- How satisfied are you with the payment options?
- What almost stopped you from completing your purchase?
- How clear was the pricing information?
Delivery and fulfilment
- How satisfied were you with the delivery time?
- Did your order arrive in good condition? (yes / no)
- How satisfied were you with the delivery updates you received?
- How satisfied were you with the packaging?
Product and features
- How satisfied are you with the quality of [product]?
- How well does [product] meet your needs?
- How satisfied are you with [feature]?
- Which feature do you use most, and why?
- What is one thing you would add or change?
Onboarding and service
- How satisfied were you with the onboarding process?
- How satisfied were you with the training or setup support?
- How confident do you feel using [product] after onboarding?
- How satisfied were you with the staff you dealt with?
Website and app
- How satisfied are you with our website / app?
- How easy was it to find what you were looking for?
- Did anything not work as expected?
Returns, billing and account
- How satisfied were you with the returns process?
- How satisfied are you with the clarity of your invoice?
- How satisfied were you with how we handled your refund?
- How satisfied were you with how we handled your account change?
Pick one core question and one open follow-up for most surveys. Add a touchpoint-specific question only when you will act on the answer.
How to analyse CSAT comments
The rating tells you whether a moment went well. The comments tell you why, and what to fix.
- Tag each comment by theme: response time, agent knowledge, delivery delay, product quality, billing, website and so on.
- Score sentiment per theme. A single comment can praise the agent and complain about the wait. Our sentiment analysis guide explains how.
- Compare satisfied and dissatisfied customers. What do 5s mention that 1s and 2s never do, and the other way round?
- Break it down by segment: channel, team, product, region, customer tier.
- Find the root cause of the biggest negative themes. "Slow response" might mean understaffed weekend shifts, one broken routing rule or a single product line generating most tickets. See our guide to root cause analysis in customer feedback.
- Track themes over time and alert on sudden changes, not just on the monthly score.
How to improve CSAT
- Fix the most common reasons for low ratings first. They usually show up clearly once comments are tagged.
- Follow up with dissatisfied customers quickly; a resolved problem often turns a 2 into a 5 next time.
- Coach with real comments, not only scores. Specific feedback is easier to act on.
- Remove effort: fewer handoffs, clearer instructions, self-service for simple tasks.
- Set expectations: many low ratings come from a gap between what customers expected and what happened.
- Close the loop: tell customers what you changed because of their feedback.
For a wider look at collecting feedback beyond surveys, see how to collect customer feedback.
CSAT vs NPS vs CES
| CSAT | NPS | CES | |
|---|---|---|---|
| Question | How satisfied were you with [interaction]? | How likely are you to recommend us? | How easy was it to [get something done]? |
| Scale | Usually 1 to 5 | 0 to 10 | Usually 1 to 7 (or 1 to 5) |
| Measures | Satisfaction with a moment | Loyalty and advocacy | Effort to complete a task |
| Time frame | Short-term, right after an interaction | Long-term relationship | Short-term, right after a task |
| Best for | Support, delivery, onboarding, checkout | Brand health, churn risk, board reporting | Self-service, returns, account changes |
| Weakness | Says little about the overall relationship | Hard to act on without comments and segments | Narrow: only effort |
They work best together: CSAT after key moments, CES on high-effort processes and NPS for the relationship, each with an open "why" question. Our NPS guide covers NPS in depth.
Limitations of CSAT
- Short-term view. A customer can be satisfied with every interaction and still leave for a cheaper competitor.
- Response bias. Very happy and very unhappy customers answer more often than everyone in between.
- Scale and culture. People in some countries rarely give top scores, which makes cross-market comparisons tricky.
- Survey fatigue. Asking after every small interaction lowers response rates and quality.
- A number without reasons. Without the comments, a falling CSAT tells you that something broke, not what.
Where Pivony fits
- Sonic Survey lets you ask a CSAT question and an open follow-up in writing, or by voice through an AI phone call or an on-site voice assistant, triggered after moments such as a purchase or a CRM event. AI analyses every answer, and there is a free tier of 25 responses a month.
- The VoC platform brings CSAT comments together with support tickets, app reviews and other feedback, analyses every comment by theme and sentiment, breaks it down by customer segment, and shows the root cause behind each change.
Next step
Want to see what is behind your CSAT? Request a free 24h root cause scan on a sample of your survey comments, reviews or tickets, or book a demo.
Frequently asked questions
What is CSAT?
CSAT (Customer Satisfaction Score) measures how satisfied customers are with a specific interaction, product or experience. Customers answer a question such as "How satisfied were you with your experience today?", usually on a 1 to 5 scale, right after the moment you want to measure.
How do you calculate CSAT?
Divide the number of satisfied responses (4 and 5 on a 5-point scale) by the total number of responses and multiply by 100. With 250 responses of which 190 are 4s and 5s, CSAT is 190 ÷ 250 × 100 = 76%. Some teams report the average rating instead; state which method you use.
What is a good CSAT score?
It depends on the touchpoint, industry, scale and survey method, so there is no universal number. The most useful comparisons are with your own trend, between your channels and teams, and with direct competitors where you have comparable data.
What questions should a CSAT survey include?
One core rating question, such as "How satisfied were you with your experience today?", and one open follow-up, such as "What is the main reason for your rating?". Add a touchpoint-specific question, for example about response time or delivery, only if you will act on it.
When should you send a CSAT survey?
Right after the moment you want to measure: when a support case closes, after a purchase, delivery, onboarding session, appointment or return. Send it in the channel the interaction happened in and keep it short.
What is the difference between CSAT and NPS?
CSAT measures satisfaction with a specific interaction, usually on a 1–5 scale, right after it happens. NPS measures long-term loyalty with "How likely are you to recommend us?" on a 0–10 scale. CSAT tells you which touchpoints work; NPS tells you how the overall relationship is doing.
What is the difference between CSAT and CES?
CSAT asks how satisfied customers were; CES (Customer Effort Score) asks how easy it was to get something done. CES is most useful for processes such as self-service, returns and account changes, where effort drives frustration.
How can I improve my CSAT score?
Analyse the comments behind low ratings, fix the most common causes first, follow up quickly with dissatisfied customers, coach teams with real comments, reduce effort and handoffs, set clear expectations and tell customers what you changed.
Compare it on your data
A short demo shows the difference. If you would rather skip the call, start with a free RCA.