Count satisfied responses
Count ratings that match your survey definition of satisfied—commonly 4 and 5 on a five-point scale.
Free CSAT calculator
Turn customer survey responses into a clear satisfaction percentage. The calculation runs only in your browser and needs no sign-up.
CSAT = satisfied responses ÷ total responses × 100
For a standard five-point satisfaction survey, responses rated 4 or 5 are usually counted as satisfied. Keep the question and scale consistent when comparing periods.
Review the CSAT method from QualtricsMore free customer-support tools
Turn customer conversations into organized support work
Try SupoviaHow the CSAT calculation works
Enter all responses
Enter the total number of valid responses from the same survey and reporting period.
Calculate the percentage
The calculator divides satisfied responses by total responses and multiplies by 100.
What CSAT measures
Customer satisfaction score summarizes the share of respondents who selected a satisfied answer. It is useful after a support interaction, purchase or onboarding step.
CSAT reflects the survey responses you collected, not every customer. Record the question, scale, channel and survey timing so comparisons remain meaningful.
Use the result responsibly
Compare like with like: the same survey question, satisfaction threshold and response window. A small sample can move sharply when one response changes.
Pair the percentage with response volume and qualitative feedback inside your own reporting system. This page deliberately does not upload or retain either.
CSAT calculator questions
Which responses count as satisfied?
On a five-point satisfaction scale, ratings of 4 and 5 are commonly counted as satisfied. If your survey uses another scale, define the positive choices before comparing results.
Does this calculator store survey data?
No. The calculation runs in your browser. Supovia analytics receive only the tool name and page language, never response counts or the calculated score.
Can I compare CSAT across channels?
You can, but only when the question, scale, threshold and measurement window are consistent. Channel and response-rate differences can otherwise distort the comparison.