metrics
CSAT (customer satisfaction score)
What is CSAT and what does it actually measure?
CSAT - customer satisfaction score - asks people to rate their satisfaction, usually on a short scale, right after an experience, and reports the percentage who answered positively (typically the top one or two ratings). It's a quick, intuitive read on "did that go well?" for a specific interaction - a purchase, a support chat, a feature.
Also known as: csat, customer satisfaction score, satisfaction rating
Prefer to watch?
Watch the recap 0:54
The demo
The simplest research metric there is. Rate the experience below, and watch the CSAT headline - the share of people who said they were satisfied - move.
How satisfied were you with checkout?
— CSAT (% satisfied — top two boxes)
What this demo shows (text version)
A CSAT survey asks how satisfied you were with checkout, on a five-point scale from very dissatisfied to very satisfied. Your answer joins a small illustrative set of earlier responses, and the headline CSAT figure updates: it's the percentage of all respondents who chose one of the top two ratings ("satisfied" or "very satisfied"). A small distribution bar shows where everyone landed.
This shows both how CSAT is calculated and what it hides. The single "% satisfied" number is quick and intuitive, tied to a specific recent experience - but it lumps "fine" in with "delighted", skews upward because unhappy people often don't answer, and never tells you why. Use it as a fast pulse alongside an open "why?" and deeper, behavioural measures. (The prior responses here are illustrative, not real data.)
CSAT is the simple one: ask "how satisfied were you?", count the share who said satisfied. It's great for measuring a specific, recent interaction and easy for people to answer - but it's shallow and biased upward (unhappy people often just leave without rating), it doesn't tell you why, and "% satisfied" hides the difference between "fine" and "delighted". Use it as a fast pulse, paired with an open "why?" and with deeper measures.
Pick a rating and watch the headline number move - CSAT is just the share of people in the top boxes. That simplicity is its appeal and its trap: it's easy to read and easy to game, tells you something went well or badly but never why, and quietly omits everyone so unhappy they didn't bother to answer at all.
Mechanically, CSAT is the percentage of respondents who chose a positive rating (e.g. 4-5 on a 5-point scale, the "top-two box") out of all who answered. It's tied to a specific, recent experience, which makes it sharper than a vague overall-happiness question - ask it right after the moment you care about, and keep the scale consistent so the number is comparable over time.
Its limits are real. Response bias skews it upward, because dissatisfied people are more likely to abandon than to rate; cultural and scale differences make absolute numbers slippery; and a single percentage flattens the gap between merely-adequate and genuinely-delighted. A high CSAT with low retention is a classic warning that people say they're satisfied and behave otherwise.
So treat CSAT as one fast signal, not the verdict. Always pair the rating with an open "what's the main reason for your score?" to get the why, watch the trend rather than obsessing over a point change, and triangulate with behavioural metrics (retention, task success) and other measures like NPS or CES. The number tells you there's something to look at; the follow-up tells you what.