Customer satisfaction scores are the most direct proxy for support quality. Learn how to measure CSAT, what a good score looks like, and the three levers that move the needle.
Customer satisfaction scores are the most direct proxy for support quality. Learn how to measure CSAT, what a good score looks like, and the three levers that move the needle.
What is CSAT and how is it measured?
CSAT (Customer Satisfaction Score) is measured by sending a short survey after a support interaction: 'How satisfied were you with the support you received?' Responses are typically on a 1-5 or 1-10 scale. Your CSAT score is the percentage of respondents who gave a top rating (4-5 on a 5-point scale, or 9-10 on a 10-point scale).
What does a good CSAT look like?
Industry benchmarks for e-commerce support: below 70% is poor (significant problems to address); 70-80% is average; 80-90% is good; above 90% is excellent. Retenza customers typically see CSAT improve by 8-15 percentage points in the first 3 months, primarily because faster response times are the biggest driver of satisfaction.
Lever 1: Speed
Response time is the single strongest predictor of CSAT in e-commerce support. A reply within 5 minutes scores 22% higher on average than a reply after 4 hours, regardless of the quality of the response. AI-assisted support's biggest CSAT impact is here.
Lever 2: First contact resolution
Customers who have to send a second email to resolve their issue score 30-40% lower on CSAT. Train your AI knowledge base to provide complete answers — not just acknowledgements. If an order can be refunded in the first reply, do it in the first reply.
Lever 3: Language match
Customers who receive a reply in their native language score 18% higher than those who receive a reply in English when they wrote in another language. Automatic language detection and multilingual AI reply generation address this lever entirely without any agent effort.