CSAT

Customer Satisfaction Metrics: CSAT, NPS, CES and More

There is no single number that captures how satisfied your customers are. CSAT, NPS and CES each measure a different angle, and the quantitative scores only tell half the story — the open-text comments tell you why.

This guide explains the main customer satisfaction metrics, what each one is good for, and how to combine them so you understand both what customers feel and why they feel it.

If you are just getting started, read the customer satisfaction survey guide first, then use this page to decide which metrics to track.

CSAT (Customer Satisfaction Score)

CSAT measures satisfaction with a specific interaction, product or experience. Respondents rate satisfaction, usually on a 1–5 scale, and the score is the percentage who chose one of the top ratings.

CSAT = (number of satisfied responses ÷ total responses) × 100

Use CSAT for transactional moments: after support, after purchase, after onboarding. Compute it from raw counts with the free CSAT calculator. For the concept-versus-metric distinction, see customer satisfaction vs CSAT.

NPS (Net Promoter Score)

NPS measures loyalty and likelihood to recommend on a 0–10 scale. Respondents are grouped into promoters (9–10), passives (7–8) and detractors (0–6), and the score is the percentage of promoters minus the percentage of detractors.

NPS = % promoters − % detractors

Use NPS for relationship-level loyalty tracked over time. Calculate it with the NPS calculator, and see customer satisfaction vs NPS for how the two differ.

CES (Customer Effort Score)

CES measures how much effort a customer had to expend, typically by asking how easy it was to get something done. Low effort strongly predicts loyalty, especially in support and service.

Use CES right after a task-based interaction — resolving an issue, completing a signup, returning a product. Calculate it with the CES calculator.

Average rating

The mean of all satisfaction ratings is simple and intuitive, but it hides distribution. An average of 3.5 could mean everyone is lukewarm or that half your customers love you and half are furious — very different situations that need different responses.

Satisfaction distribution

Looking at how ratings spread across the scale reveals what an average conceals. A bimodal distribution (lots of 1s and lots of 5s) signals a polarizing experience worth investigating, while a tight cluster suggests a consistent one.

Create your survey with AI

Describe what you want to learn and get a ready-to-edit survey in seconds — free to start.

Response rate

Response rate is not a satisfaction metric itself, but it shapes how much you can trust the others. Low response rates raise the risk of bias — often only the very happy or very unhappy reply. Track it and use the survey response rate calculator to interpret your results in context.

Qualitative themes and sentiment

Open-text answers are where the 'why' lives. Grouping comments into recurring themes (speed, price, quality, communication) turns raw text into priorities, and sentiment analysis flags whether the tone is positive, neutral or negative at scale.

AI can summarize hundreds of comments into themes and sentiment quickly — see analyzing open-text survey responses for the approach.

Comparing CSAT, NPS and CES

CSATNPSCES
MeasuresSatisfaction with an experienceLoyalty / likelihood to recommendEase / effort of a task
Typical scale1–50–101–5 or 1–7
Best momentAfter an interactionRelationship, recurringRight after a task
Question focusHow satisfied were you?How likely to recommend?How easy was it?
Best forTransactional feedbackLong-term loyalty trendsReducing friction

For deeper guidance on choosing, see how to measure customer satisfaction.

No single metric tells the whole story

Each metric answers a different question, and none of them alone explains why customers feel the way they do. The strongest programs pair one quantitative metric (so trends are comparable) with open-text feedback (so the reasons are clear), reviewed by segment and over time.

Frequently asked questions

Keep reading