CSAT Survey: Questions, Formula, Timing and How to Read the Score
How to run a CSAT survey: the formula, questions by touchpoint, when to send, CSAT versus NPS and CES, response-rate bias and a worked score with margins.
In shortA CSAT survey asks about one interaction right after it happens, usually with one question on a 5-point scale. The common score is the share of 4 and 5 answers. Use it for touchpoints, NPS for the relationship and customer effort for friction. Read it with the number of responses beside it: 120 of 200 satisfied is 60%, with a margin of about ±6.8 points. Hercules Works can survey category users and competitor customers on its panel.
Contents
One Interaction, One Question, One Score
Customer satisfaction (CSAT) is a snapshot. It asks how satisfied a customer was with a specific purchase, delivery, support call or product. That makes it good at finding the touchpoint that is failing, and weak at telling you whether the customer will stay.
This guide gives the formula, a question bank by touchpoint, timing rules and a comparison with NPS and customer effort. The score in the worked example is illustrative. We do not publish a benchmark for good CSAT, because it depends on the touchpoint and the scale.
The Formula and a Worked Score
On a 5-point scale, a common convention counts answers of 4 and 5 as satisfied. Different tools use different conventions, so write yours down and keep it fixed.
CSAT % = (number of 4 and 5 answers ÷ number of all answers) × 100
Worked example, illustrative: 200 customers answer a delivery question and 120 choose 4 or 5. The score is 120 ÷ 200 × 100 = 60%. For a share near 60% with 200 responses, the margin of error is about ±6.8 points in a simple random sample. A move from 60% to 64% next month is inside that margin, so it is not yet a trend.
CSAT, NPS and Customer Effort: Which to Use
| Measure | Asks | Best for | Weak at |
|---|---|---|---|
| CSAT | How satisfied were you with this interaction? | Finding a failing touchpoint | Predicting loyalty |
| NPS | How likely are you to recommend us? | The overall relationship; see Reichheld, HBR 2003 | Explaining why |
| Customer effort | How easy was it to get this done? | Support, returns, onboarding | Product quality |
Use CSAT after the event, NPS on a regular rhythm and effort where customers must do work. Do not average them into one number. See the NPS methodology guide and customer experience surveys.
Questions by Touchpoint
Illustrative. Ask one rating, then one open question.
| Touchpoint | Rating question | Follow-up |
|---|---|---|
| Purchase | How satisfied are you with the buying experience? | What would have made it better? |
| Delivery | How satisfied are you with the delivery? | What went wrong, if anything? |
| Support | How satisfied are you with how your issue was handled? | Was your issue solved? If not, why? |
| Product | How satisfied are you with the product so far? | What do you like least? |
| Onboarding | How satisfied are you with getting started? | Which step was hardest? |
When to Send and How to Avoid Bias
- Send soon after the event, while the customer remembers it, and once per event.
- Keep it to one rating and one open question. Long surveys cut the response rate.
- Offer it in the customer's language.
- Watch who answers. Very happy and very unhappy people reply most, so a low response rate can distort the score. Report the number of responses with every score.
- Close the loop. Reply to unhappy customers, and record what you changed.
Your Customers Versus Category Users
A CSAT survey of your own customers must go to your own list, through your order, email or app flow. A consumer panel does not hold your customers. A panel suits a different job: measuring satisfaction among category users, comparing yourself with competitors' customers, or testing a service idea. Use the right tool for each. The platform checklist covers the buying decision.
Running This Study on Hercules
If your job is to measure satisfaction among category users or competitors' customers in India, this is the study at list price. For your own customers, use a tool that sends to your list.
| Step | Plan | Responses | Price per response | Cost |
|---|---|---|---|---|
| Test the questionnaire | Free | 30 | ₹0 | ₹0 |
| Small pilot | Plus | 100 | ₹150 | ₹15,000 |
| Main study | Pro | 400 | ₹120 | ₹48,000 |
Pro is sold in bundles of 400 to 1,000 responses and costs 20% less per response than Plus. Enterprise plans come with a dedicated research team and are priced on request.
The SuperJ network has a 25M+ consumer footprint. The survey-eligible panel is a subset, and availability depends on your targeting. Members complete Aadhaar-based KYC once, and each study receives a zero-knowledge proof of verification. Surveys run in as many Indian languages as the study needs. Quality checks flag speeders, straight-lining and inconsistent answers. Read /panel-methodology and pricing.
The limit: the app reaches smartphone users only, so offline households need face-to-face fieldwork.
Frequently asked questions
How do you calculate a CSAT score?
On a 5-point scale, count the 4 and 5 answers, divide by all answers and multiply by 100. Tools differ in convention, so state yours. With 120 of 200 satisfied, CSAT is 60%.
What is the difference between CSAT and NPS?
CSAT measures satisfaction with one interaction. NPS measures how likely a customer is to recommend you, so it reads the overall relationship. Use both for different jobs.
When should I send a CSAT survey?
Soon after the interaction, once per event, with one rating question and one open follow-up.
What is a good CSAT score?
It depends on the touchpoint, the scale and who answers, so we publish no benchmark. Track your own score over time and report the number of responses.
How many responses do I need for a CSAT score?
For a share near 60%, 100 responses give about ±9.6 points and 400 give about ±4.8 points in a simple random sample. Small changes need larger samples.
Ready to get real consumer insights?
25M+ consumer footprint. Free to start.