CSAT · CES Customer Satisfaction Metrics
Customer Satisfaction Score · Customer Effort Score
What is CSAT?
- CSAT measures satisfaction with the experience; CES measures effort in the process
- One question right after the experience, plus a one-line question asking why
- Tying scores to performance reviews inflates the numbers without improving anything
CSAT and CES don't ask 'Do you like our company?' They ask about one specific experience the customer just had. CSAT asks how satisfied they were with that experience; CES asks how much effort it took to resolve a problem or get a request handled. Both measure something different from NPS, which asks how likely customers are to recommend you based on the whole relationship, so the question is not which of the three to pick, but which question to ask at which touchpoint.
They serve different purposes. CSAT goes on right after an outcome, such as the end of a training course, a completed delivery, or a closed support case, to gauge quality. CES goes on processes where customers have to get something done, such as requesting a quote, reporting an outage, or renewing a contract, to find the obstacles. The metric grew out of the finding that customers respond more to having their effort reduced than to being delighted beyond expectations. NPS looks at the overall relationship, once every six months or once a year.
The math is simple, but you need to fix your definitions. CSAT is usually the share of 4 and 5 responses on a 5-point scale; CES is the average level of agreement with 'It was easy to get this handled,' or the share of positive responses. If you change the scale or the way you aggregate, you can no longer compare against earlier results, so document your method once you choose it. Always follow the score question with a one-line open-ended question asking 'why you answered that way'. The score tells you where the problem is; the reason tells you what to fix.
In practice, things go wrong the moment the score becomes the goal. If CSAT is tied directly to an agent's performance review, surveys quietly skip unhappy customers, or agents add 'Please give us a 5,' and only the number goes up. A once-a-year survey is a problem too. Answers about an experience from three months ago reflect impressions, not memories. Watch for the false signal of a low response rate as well. In B2B, respondents often number only a few dozen, so one or two answers can swing the score sharply. Always report the number of responses alongside the percentage.
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum Consulting
Hyerang ParkSpecialist in content storytelling and AI-driven brand content| Category | CSAT | CES | NPS |
|---|---|---|---|
| What it asks | Satisfaction with a specific experience | Effort required to get something done | Likelihood of recommending to others |
| What it measures | Quality of individual touchpoints | Process friction | Loyalty across the whole relationship |
| When to survey | Right after training, delivery, or a consultation ends | Right after a request, intake, or renewal is processed | Every six months or once a year |
| Sample question | How satisfied were you with this service? | Was it easy to get your request handled? | How likely are you to recommend us to a friend or colleague? |
| Departments that typically use it | Training, service, and customer support | Sales support, operations, product | Executive leadership, marketing, and customer success |
CSAT has no single founder; it hardened into a standard metric through customer satisfaction survey practice. CES became widely known through a 2010 Harvard Business Review article by Matthew Dixon, Karen Freeman, and Nicholas Toman, which reported that reducing customer effort matters more for loyalty than trying to delight customers.
Consider a hypothetical case. Company A supplies production management software to mid-sized manufacturers and has 140 clients and 110 employees. It ran one NPS survey every December. The score sat around 30 for three years, and nothing in the results said what to fix.
- The customer success lead proposed replacing the annual survey with touchpoint surveys. "If we only ask whether people would recommend us, we never learn where it hurts," they said, and the marketing lead agreed.
- Both teams added a single CSAT question right after training ended and right after an outage was closed, and a single CES question right after a quote request and right after a contract renewal was completed. Each also added a one-line follow-up asking why.
- Over three months, 212 responses came in. CSAT for training and incident response was high at 88%, but contract renewal CES was the lowest, at 3.4 out of 7.
- Reading the open-ended answers, we found that “we resubmit the same documents every time we renew” came up 19 times. The sales support manager said, "So we have three separate forms across three departments."
- They merged three renewal documents into one and switched to e-signatures. The next quarter's renewal CES rose from 3.4 to 5.6, and the average renewal processing time fell from 12 days to 4.