Customer satisfaction score (CSAT)
CSAT (customer satisfaction score) is the share of clients who rate a specific interaction or purchase as satisfying, usually on a short scale right after it happens.
What it is
CSAT asks the client one direct question about one specific experience: “How satisfied are you with this conversation?” The answer comes on a short scale, for example from 1 to 5. Unlike broader loyalty metrics, CSAT is about the moment: this agent, this chat, this delivery.
How it is calculated
The common method counts only positive answers: CSAT = number of ratings 4 and 5 / number of all ratings × 100%.
An illustrative example: 300 clients answered, 240 of them chose 4 or 5. CSAT = 240 / 300 = 80%. Some teams report the average score instead, but the share of satisfied clients is easier to read and compare over time. Always show how many people answered: 80% out of 15 answers means far less than 80% out of 500.
Why it matters
CSAT is the quickest signal that something went wrong in a specific place: one agent, one type of request, one shift. It is cheap to collect and easy to explain to the team. Comments next to low ratings are often more useful than the number itself, because they name the actual problem.
The typical mistake
Trusting a high score without asking who answered. Satisfied clients and very angry ones answer more often, the silent middle does not answer at all. Agents can also influence the survey by asking only the clients who seem happy. Send the question after every closed contact, read the low ratings one by one, and compare CSAT with repeat contacts and first contact resolution. If CSAT is high and clients keep coming back with the same issue, the problem is in the survey, not in the clients.
More in Signals
- Call center KPIs for messaging apps: what to measure when agents type instead of talkIn messaging apps, classic call center KPIs change meaning: average handle time loses weight, while first response time, service level, missed conversations and resolution in a single conversation move to the front. New metrics appear that calls never had: concurrent conversations and promises kept. Measure them in working hours and together, never one at a time.
- Sales rep performance review: how to judge the conversation, not the personReview a rep’s work with a scorecard of five to seven criteria and a simple yes, partly, no scale, on a small sample of conversations: a few per person each week, some random and some that matter. Reviewers regularly compare their scores with each other, and the result is discussed one on one: a good example first, a question instead of a verdict, one thing to improve. You judge the conversation, not the person, and the team knows the criteria in advance.