Customer Satisfaction Score (CSAT)
CSAT measures how satisfied customers are with a specific interaction, feature, or experience — usually immediately after it happens. Unlike NPS (which measures overall loyalty), CSAT is transactional: it captures sentiment at a precise moment. It's the go-to metric for support teams, onboarding flows, and feature releases where you need fast, contextual feedback.
Note: Satisfied = scores of 4 or 5 on a 1–5 scale (or "Good"/"Very Good" on emoji/text scales). Neutral and dissatisfied responses are excluded from the numerator but counted in the denominator.
CSAT > 80% is considered good; > 90% is excellent for support interactions
CSAT < 70% on support interactions suggests a systemic quality or expectation-setting problem
Benchmarks by segment
How to improve CSAT
Survey at the right moment — CSAT is most accurate immediately post-interaction (within 1 hour), not days later
Close the loop: contact every respondent who gives a score below 3 within 24 hours to understand and resolve their issue
Train support on first-contact resolution (FCR) — the highest CSAT scores come from resolving issues in one interaction
Use CSAT verbatims to identify recurring themes; convert top themes into product or documentation improvements
Common measurement mistakes
Frequently Asked Questions
CSAT is transactional: use it to measure specific touchpoints (support, onboarding, a feature). NPS is relational: use it to measure overall loyalty and likelihood to recommend. Best practice is to use both — NPS quarterly for trend, CSAT continuously for operational quality.
For 95% confidence with ±5% margin of error, you need about 385 responses. In practice, aim for at least 30 responses per team or channel before drawing conclusions. Response rates for in-app CSAT surveys typically range from 10–30%.