The CSAT formula is straightforward, but applying it consistently — with the right scale, timing, and segment logic — is what makes the score useful. This guide walks through the formula, a worked example, scale variations, and the most common calculation errors to avoid.
CSAT (%) = (Number of satisfied responses ÷ Total responses) × 100
What counts as “satisfied” depends on your scale:
| Scale | Satisfied responses |
|---|---|
| 1–5 | Scores of 4 and 5 |
| 1–10 | Scores of 9 and 10 |
| 3-point (bad / ok / good) | “Good” only |
You send a post-support CSAT survey and receive 200 responses:
Satisfied responses = 120 + 40 = 160 Total responses = 200
CSAT = (160 ÷ 200) × 100 = 80%
This means 80% of respondents were satisfied with the support interaction.
A single overall number is less useful than scores broken down by segment:
Apply the same formula to each segment subset. A segment with 40 responses and 30 satisfied gives a CSAT of 75% — even if the overall score is 85%.
Counting neutral responses as satisfied. A score of 3 on a 1–5 scale is not satisfied. Only 4 and 5 count.
Mixing response windows. If one week’s data uses a 48-hour survey window and the next uses 7 days, the scores are not comparable.
Ignoring low response rates. A CSAT of 95% from 8 responses is not a reliable signal. Aim for at least 30 responses per segment before drawing conclusions.
Comparing across different scales. A 1–5 scale and a 1–10 scale produce different numbers even with identical sentiment. Pick one and stick to it.
A single CSAT score tells you where you are. Trend direction tells you whether things are improving. Track weekly or monthly movement and connect score changes to specific events:
When scores drop, the open comment field tells you why. Review low-score comments weekly and group them into themes before deciding what to fix.
What is a good CSAT score? It depends on industry and touchpoint, but scores above 80% are generally considered acceptable, and above 85% is strong. See CSAT benchmarks by industry for context.
Should I use a 5-point or 10-point scale? A 1–5 scale is simpler for respondents and produces higher response rates. A 1–10 scale gives more granularity but is harder to benchmark consistently. Most small-to-mid teams use 1–5. Choose one and do not switch — consistency matters more than scale choice.
How many responses do I need for a reliable CSAT score? A minimum of 30 responses per segment is a common rule of thumb. Below 30, variance is too high to trust the number. For overall scores with broad surveys, aim for at least 100 responses per period.
Can I calculate CSAT in a spreadsheet? Yes. Sum the rows where score is 4 or 5, divide by the total row count, and multiply by 100. A tool like Wyapy does this automatically with trend charts and segment breakdowns.
For benchmark context, read CSAT benchmarks by industry. For survey wording ideas, see CSAT survey examples.