NPS vs CSAT vs CES: which metric should you use?

NPS, CSAT, and CES are the three most common customer-experience metrics — and teams constantly use the wrong one for the question they're actually asking. They measure genuinely different things. This guide explains what each one measures, how to calculate it, what a good score looks like, and when to reach for which.

The one-line difference

NPS (Net Promoter Score) — measures loyalty and word-of-mouth. "Would you recommend us?"
CSAT (Customer Satisfaction) — measures satisfaction with a specific experience. "Were you happy with that?"
CES (Customer Effort Score) — measures how hard something was. "How easy was it to get this done?"

Loyalty, satisfaction, effort. Different questions, different metrics.

NPS (Net Promoter Score)

The question: "How likely are you to recommend [product] to a friend or colleague?" on a 0–10 scale.

How to calculate:

Promoters = 9–10
Passives = 7–8
Detractors = 0–6
NPS = % Promoters − % Detractors (a number from −100 to +100)

Worked example. 200 responses: 110 promoters (55%), 50 passives (25%), 40 detractors (20%). NPS = 55 − 20 = +35.

Benchmarks: above 0 is net-positive; 30–50 is good; 50+ is excellent; 70+ is world-class. Benchmarks vary a lot by industry, so track your trend more than the absolute number.

Best for: overall relationship health, loyalty, and predicting growth via word-of-mouth. Measure it periodically (quarterly), not after every interaction.

CSAT (Customer Satisfaction Score)

The question: "How satisfied were you with [specific experience]?" typically on a 1–5 (or 1–7) scale.

How to calculate: CSAT = (# satisfied responses ÷ total responses) × 100, where "satisfied" usually means the top 2 box (4–5 on a 5-point scale).

Worked example. After a support chat, 150 responses, 120 rated 4 or 5. CSAT = 120 ÷ 150 = 80%.

Benchmarks: 75–85% is typical; above 90% is strong. Like NPS, it's industry-dependent.

Best for: measuring a specific touchpoint right after it happens — a support interaction, an onboarding step, a feature launch. It's immediate and intuitive, but only captures short-term sentiment about that one moment.

CES (Customer Effort Score)

The question: "How easy was it to [complete a task]?" on a scale from Very Difficult to Very Easy (commonly 1–7).

How to calculate: typically the average score, or the % of respondents choosing the "easy" end of the scale.

Worked example. After a returns flow, average CES = 5.8 / 7, with 78% choosing the top-3 "easy" boxes.

Why it matters: research popularized by the Harvard Business Review article "Stop Trying to Delight Your Customers" found that reducing effort is a stronger predictor of loyalty than delighting customers. Low-effort experiences keep customers; high-effort ones churn them, even if they're otherwise satisfied.

Best for: transactional, task-based experiences — checkout, onboarding, support resolution, self-service. If a task should be effortless, measure its effort.

Side by side

	NPS	CSAT	CES
Measures	Loyalty / advocacy	Satisfaction	Ease / effort
Question	"Recommend us?"	"Satisfied?"	"How easy?"
Scale	0–10	1–5 (or 1–7)	1–7
Timing	Periodic / relationship	After a touchpoint	After a task
Predicts	Growth, word-of-mouth	Short-term happiness	Loyalty / churn risk
Score range	−100 to +100	0–100%	Avg or % easy

When to use which

Use NPS for a periodic pulse on overall loyalty and to benchmark your relationship health over time.
Use CSAT right after a specific interaction to see if that moment landed.
Use CES for task-based flows where friction predicts churn — the metric most tied to behavioral loyalty.

Many mature teams use all three: NPS quarterly for the relationship, CSAT after key touchpoints, and CES on critical task flows. They answer different questions, so they complement rather than compete.

Common mistakes

Using NPS for everything — it's a relationship metric, not a touchpoint metric. Firing an NPS survey after every support ticket annoys customers and muddies the signal.
Reading a single number in isolation — track the trend and segment it; a blended score hides which customers are unhappy.
Ignoring the follow-up "why" — always pair the score with an open-ended question and code the responses; the number tells you what, the comments tell you why.
Comparing your score to unrelated industries — benchmarks vary widely; compare to yourself over time and to close peers.

Measure NPS, CSAT, and CES free

ResearchRocket's survey builder includes NPS, CSAT, CES, and Likert presets built in — so you don't have to hand-build the scales or calculation. Field the survey via shareable links or email lists, then analyze the results (including the open-ended "why") with a real statistics engine and AI thematic coding, all in one place.

FAQ

What's the difference between NPS, CSAT, and CES? NPS measures loyalty and likelihood to recommend (0–10), CSAT measures satisfaction with a specific experience (usually 1–5), and CES measures how much effort a task took (usually 1–7). They answer different questions and are best used together.

How do you calculate NPS? Subtract the percentage of detractors (scores 0–6) from the percentage of promoters (scores 9–10). The result ranges from −100 to +100. Passives (7–8) are excluded from the calculation.

How do you calculate CSAT? Divide the number of satisfied responses (typically the top 2 box, e.g., 4–5 on a 5-point scale) by total responses, then multiply by 100 for a percentage.

What is a good NPS score? Above 0 is net-positive; 30–50 is good, 50+ is excellent, 70+ is world-class. Benchmarks vary by industry, so tracking your own trend matters more than the absolute number.

Which metric is best for predicting loyalty? CES (Customer Effort Score) is a strong predictor of loyalty — reducing effort tends to retain customers better than trying to delight them. NPS predicts advocacy and growth; CSAT captures short-term satisfaction.

Can I use all three? Yes — mature teams use NPS periodically for the overall relationship, CSAT after specific touchpoints, and CES on key task flows. They complement each other.

Measure NPS, CSAT, and CES free on ResearchRocket →