What is customer satisfaction score (CSAT)?

Customer satisfaction score (CSAT), explained

CSAT is the most direct feedback a support team gets. Right after a conversation, the customer is asked to rate it, often with stars, faces or a thumbs up and down. Because it is tied to one interaction, it is good at showing which topics, channels or replies leave customers unhappy.

The standard calculation counts only the satisfied responses. On a 1–5 scale, that is usually 4s and 5s. The result is a percentage, not an average score, which makes it easier to compare over time.

Its main weakness is who answers. Response rates on post-chat surveys are often low, and the customers who respond tend to be the very pleased or the very annoyed. Treat CSAT as a signal to investigate, not a precise measurement, and always read the comments.

For AI support, report CSAT separately for conversations the bot handled alone and those that reached a human. Blending them can hide a poor bot experience behind a strong human team, or the reverse.

How to measure it

Keep the question and scale fixed over time, and track response rate alongside the score.

Common mistake

How SuperBot relates

SuperBot's analytics report CSAT together with resolution rate, sentiment and knowledge gaps, so a dip in satisfaction can be traced to the topics and missing content behind it.

• SuperBot features
• AI chatbot evaluation checklist

Frequently asked questions

Related terms

• Net Promoter Score (NPS)Net Promoter Score (NPS) measures customer loyalty by asking how likely someone is to recommend a company to a friend or colleague on a 0–10 scale. Scores of 9–10 are promoters, 7–8 passives and 0–6 detractors. NPS is the percentage of promoters minus the percentage of detractors, giving a result from −100 to +100.
• Customer effort score (CES)Customer effort score (CES) measures how easy it was for a customer to get an issue resolved or complete a task, usually by asking them to rate a statement such as “The company made it easy to handle my issue” on a numbered agreement scale. Lower effort is associated with stronger loyalty, which is why support teams track it.
• Sentiment analysisSentiment analysis is the automated detection of the emotional tone in text, typically classified as positive, neutral or negative. In customer support, it is applied to messages and conversations to spot frustrated customers, track how people feel about topics over time, and flag conversations for a human to review, without waiting for a survey response.
• Resolution rateResolution rate is the percentage of customer conversations or tickets that end with the customer's issue solved. For AI support, it usually means conversations the bot resolved without a human. Because vendors define “resolved” differently, the number is only meaningful when you know the definition behind it and check it against repeat contacts and satisfaction.
• First contact resolution (FCR)First contact resolution (FCR) is the percentage of customer issues fully resolved in the first interaction, with no follow-up, transfer or repeat contact needed. It is one of the most customer-centred support metrics because having to contact a business again about the same problem is a common source of frustration.