CSAT, NPS, and CES are common ways to summarize customer feedback, but they measure different things. CSAT asks about satisfaction with an experience or interaction. NPS asks how likely someone is to recommend an organization or product. CES asks how easy or difficult it was to complete a task. Choosing a metric starts with the question the team needs to answer.
None of these measures is a complete score for customer health. A response is a person’s answer to a question at a particular moment. It does not by itself establish why an outcome happened or predict what every customer will do next.
What each metric measures
CSAT: satisfaction with a defined experience
Customer Satisfaction Score (CSAT) usually comes from a question such as “How satisfied were you with this support interaction?” The organization defines which response options count as satisfied. A common calculation is the number of satisfied responses divided by the total number of valid responses, multiplied by 100. The report should state whether “satisfied” means the top response or a broader group of positive responses.
CSAT is most interpretable when it refers to a specific moment, such as onboarding, a support resolution, or an order. A broad question about overall satisfaction can be useful for a defined research purpose, but it may combine several unrelated experiences.
NPS: likelihood to recommend
Net Promoter Score (NPS) typically asks how likely a respondent is to recommend the organization or product on a 0–10 scale. Responses of 9–10 are classified as promoters, 7–8 as passives, and 0–6 as detractors. NPS is calculated as the percentage of promoters minus the percentage of detractors. The result ranges from –100 to 100.
NPS can provide a relationship-level pulse when the question, audience, and collection timing are consistent. It is not a direct measure of retention, referrals, or revenue; those behaviors should be observed separately. Interpret comparisons in context, especially when markets, customer mix, survey channels, or response rates differ.
CES: effort required to complete a task
Customer Effort Score (CES) asks how easy or difficult it was to do something specific, such as resolve a support issue or complete setup. Question wording and response scales vary. Some surveys ask respondents to rate ease; others ask about effort. Define the direction of the scale in reports so a higher value cannot be mistaken for a worse experience.
CES is most useful when the task is clear and the organization can change the steps involved. It can reveal friction that a general satisfaction question might hide, but it does not show the whole relationship or explain every cause of difficulty.
Which metric should you use?
- Use CSAT when you need a reaction to a particular interaction or milestone.
- Use NPS when you need a consistent relationship-level question about recommendation intent.
- Use CES when you need feedback about the effort involved in a specific task or process.
A software company, for example, might ask CSAT after a support case, CES after users complete onboarding, and NPS in a periodic relationship survey. The metrics should not be combined into one score unless the organization has a clear, tested reason and explains what the combined number means.
Design a survey that produces interpretable feedback
- Ask the question close enough to the experience that respondents can recall it, while avoiding unnecessary survey prompts.
- Use one clear question at a time and avoid combining satisfaction with ease, speed, or outcome in the same sentence.
- Define the eligible population, invitation method, response window, scale, and calculation before comparing results.
- Report the number invited and the number of valid responses where possible; response patterns may differ from the full customer group.
- Provide an optional comment field when follow-up context is useful, and protect personal information in collection and reporting.
- Review small samples and subgroup differences cautiously rather than treating every score change as meaningful.
Turn a score into a useful next step
A feedback metric is a signal to investigate, not an instruction to reward or penalize a team. Review comments and operational records, compare similar cohorts, and check whether the question or survey channel changed. If a low score points to a complicated process, map the steps and locate where customers get stuck. If responses are positive but retention is falling, investigate product fit, pricing, adoption, and service factors rather than assuming the survey is wrong.
Customer feedback becomes more useful when it has a clear owner and a path to action. The guide to building a customer feedback loop explains how teams can capture, route, and close the loop on what they learn. For outcome measurement, keep customer retention by cohort distinct from survey scores.
Common questions
Are NPS and CSAT interchangeable?
No. NPS asks about recommendation likelihood and CSAT asks about satisfaction with a defined experience. Their questions, scales, and interpretations differ.
Can CES replace customer support analysis?
No. CES can help identify perceived effort, but support volume, resolution quality, repeat contacts, and customer comments may explain different parts of the experience.
Should every interaction trigger a survey?
Usually not. Repeated prompts can create fatigue and skew who responds. Choose a sampling approach that provides useful coverage without asking the same people too often.
How did this article land?
Choose one reaction. You can change it anytime.