CSAT: What It Is, How to Measure Customer Satisfaction and How to Improve It

CSAT (Customer Satisfaction Score) is a metric that measures immediate customer satisfaction after a specific interaction, using a 1-5 or 1-10 scale. It's a snapshot of a concrete moment: how the customer feels right after talking to support, completing a purchase or finishing onboarding.

The question is always a variation of the same thing: "How satisfied are you with [interaction]?". The customer responds on a numerical scale or with options like "Very dissatisfied" to "Very satisfied", and the company gets direct data, without ambiguities or complicated interpretations.

What differentiates CSAT from other metrics like NPS is its transactional nature. NPS asks about the overall relationship with the brand--"Would you recommend this company?". CSAT asks about the last touchpoint. They're different questions that measure different things. A customer can have a high NPS because they love your product, but a low CSAT because the last time they contacted support it took you two days to respond.

That granularity is precisely the value of CSAT: it tells you exactly where the operational problems are. It doesn't tell you if the customer loves you; it tells you if you've solved their problem. And in customer support, that's what matters. If you want to understand the complete ecosystem of metrics, check our guide on customer success metrics.

How to Calculate the CSAT Score

The formula is straightforward with no room for interpretation:

CSAT = (Positive responses / Total responses) x 100

Positive responses are those that fall in the top two values of your scale. If you use a 1 to 5 scale, positive responses are 4 and 5. If you use 1 to 10, they're 9 and 10.

Practical example: you've received 200 responses to your CSAT survey after support interactions. Of those 200, 156 scored with 4 or 5. Your CSAT score is (156 / 200) x 100 = 78%.

Is that 78% good or bad? It depends on the sector. These are the reference benchmarks:

• SaaS: 75-85%. The product is complex, expectations are high and customers are technically demanding.
• E-commerce: 80-90%. The transaction is simpler and expectations are more contained.
• Technical support: 70-80%. Solving technical problems generates inherent friction; a 75% here is a solid result.

Don't compare yourself with companies in other sectors. A 78% in SaaS is a competitive result; a 78% in e-commerce indicates clear room for improvement.

When and How to Measure CSAT

Timing is everything. If you launch the survey too late, the customer no longer remembers the interaction precisely. If you launch it too soon, they haven't had time to evaluate whether the solution really worked. The sweet spot is between 5 minutes and 24 hours after the interaction.

Key moments to measure customer satisfaction with CSAT:

• Post-support ticket: immediately after closure. It's the most common use case and the one with the best correlation to the team's operational quality.
• Post-live chat: when finishing the conversation, before the customer closes the window.
• Post-purchase: after order confirmation, especially if the checkout process has several steps.
• Post-onboarding: when completing the product's initial setup, when the first impression is fresh.

Channels matter as much as timing. Email surveys have response rates of 10-15%. In-app surveys rise to 20-30%. WhatsApp surveys can exceed 40% because the channel is already open and the response requires a single tap.

Common tools: Zendesk includes integrated CSAT in its ticket system. SurveyMonkey offers customizable surveys with conditional logic. GuruSup automates CSAT directly on WhatsApp after each interaction resolved by the AI agent, eliminating secondary channel friction and multiplying response rates.

How to Improve CSAT

Measuring without acting is wasting time. These are the five levers that have the greatest impact on customer satisfaction:

Reduce wait times. The correlation between first response time and CSAT is almost linear. Every minute of waiting reduces the probability of a positive score. The standard in chat support is responding in less than 60 seconds. If you don't reach that, you have a capacity or prioritization problem.

Resolve at first contact. First Contact Resolution (FCR) is the indicator that most impacts CSAT. When the customer has to write again, call back or explain the same problem to another agent, satisfaction plummets. An FCR above 70% is the acceptable minimum; the best operations exceed 85%.

Personalize responses. Generic responses like "Thank you for contacting us, your case is being reviewed" destroy the perception of real attention. The customer wants to feel that someone understands their specific problem and is working on it, not that they're a ticket number.

Post-resolution follow-up. A brief message 24-48 hours after closure--"Has it been resolved correctly?"--demonstrates commitment and allows detecting resolutions that didn't actually resolve anything. It's a minimal gesture with measurable impact.

Train the team in communication. Technical skills solve the problem; communication skills determine how the customer feels at the end of the interaction. Train your team in active listening, expectation management and positive language.

There's a sixth lever that's transforming results: AI for CSAT. AI agents resolve queries instantly, consistently and without mood variability. A well-trained customer service chatbot maintains response times of seconds, high FCR on standard queries and uniform tone 24 hours a day. The direct result is a CSAT that rises because the base experience structurally improves.

CSAT vs NPS vs CES: Which to Use and When

The three metrics measure different things and are complementary, not interchangeable:

CSAT measures point-in-time satisfaction with a specific interaction. It's the most tactical metric of the three. It tells you if you've solved a specific problem well. Typical scale: 1-5. Question: "How satisfied are you with this interaction?".

NPS (Net Promoter Score) measures long-term loyalty. Question: "Would you recommend our company?". Scale: 0-10. It's a strategic metric that reflects the customer's overall relationship with the brand, not a specific moment. Check our complete guide on NPS to dive deeper.

CES (Customer Effort Score) measures the effort the customer had to make to resolve their problem. Question: "Was it easy to resolve your inquiry?". It's the metric that best predicts future behavior: the more effort, the more likely they'll churn.

Mature companies use all three. CSAT after each support interaction. NPS quarterly to measure the overall relationship. CES at identified friction points (return processes, complex escalations, technical configurations). Together, they give a complete picture of customer experience that no individual metric can offer. If you want to understand how churn rate relates to these metrics, the correlation is direct: sustained low CSAT predicts churn.

Frequently Asked Questions

What is a good CSAT?
It depends on the sector, but as a general rule, a CSAT score above 80% is considered good and above 90% excellent. What matters is not the absolute number but the trend: a CSAT that improves month to month indicates your actions are working, even if you haven't reached the sector benchmark yet.

Is CSAT or NPS better?
It's not a choice. CSAT measures satisfaction with specific interactions; NPS measures general loyalty. You need both. If you can only implement one, start with CSAT because it gives immediate operational feedback you can act on today, not in a quarter.

Can CSAT be automated with AI?
Yes, and it's one of the most effective ways to increase response rates. GuruSup sends automatic CSAT surveys via WhatsApp right after each resolution, collects responses and centralizes them in a dashboard without manual intervention. Combined with a solid customer success approach, automated CSAT feeds a continuous improvement cycle.

GuruSup automates CSAT measurement on WhatsApp with AI agents that resolve queries instantly and launch automatic satisfaction surveys after each interaction. No friction, no emails that nobody opens, with response rates that triple those of traditional channels. Try GuruSup free.