NPS vs CSAT vs CES: Which Customer Feedback Metric Should You Use?

4 min read

Three metrics dominate customer feedback: NPS, CSAT, and CES. Each measures something different. Used correctly, they're powerful. Used interchangeably — which is what most teams do — they'll send you optimizing for the wrong thing.

Here's what each one actually measures and when to use it.

NPS — Net Promoter Score

Question: "How likely are you to recommend us to a friend or colleague?" (0–10)

What it measures: Long-term loyalty and word-of-mouth potential.

Scoring:

9–10 = Promoter
7–8 = Passive
0–6 = Detractor
NPS = % Promoters ? % Detractors

NPS is the headline metric for most SaaS companies. It's a leading indicator of growth — promoters drive referrals, detractors drive churn. A single number, easy to benchmark across industries.

When to use NPS

Measuring overall product sentiment
Tracking trends quarter-over-quarter
Comparing your company to industry averages

When NPS misleads

NPS is a relationship metric, not a transaction metric. Sending it after a specific support ticket is a category error — respondents will conflate the ticket experience with their overall opinion of your product, and you'll get noise.

CSAT — Customer Satisfaction Score

Question: "How satisfied were you with [specific interaction]?" (1–5)

What it measures: Satisfaction with a specific, recent experience.

Scoring: Usually % of respondents giving 4 or 5 out of 5.

CSAT is a transactional metric. It tells you whether a specific interaction went well. Great for support tickets, checkout flows, and post-delivery surveys.

When to use CSAT

After a support ticket resolution
After a specific feature use (did the onboarding go well?)
After a purchase or subscription upgrade

When CSAT misleads

CSAT shows satisfaction with the current experience but tells you nothing about future loyalty. A customer can rate a ticket resolution 5/5 and still churn next month because your pricing changed. It's a snapshot, not a trend.

CES — Customer Effort Score

Question: "How much effort did it take to [accomplish X]?" (1–7, usually inverted so low = good)

What it measures: Friction. How hard did the customer work to get what they wanted?

Scoring: Average score across responses, or % who rated "low effort."

CES is the metric that actually predicts churn most reliably — more than NPS or CSAT. Customers don't remember great experiences; they remember painful ones. CES finds the pain.

When to use CES

After onboarding completion
After self-serve feature use
After a cancellation attempt ("how hard was it to cancel?")
Any high-friction, high-effort moment in the customer journey

When CES misleads

CES is purely about friction. A delightful, easy, but useless product can score well on CES and still fail. Use it alongside NPS, not instead of it.

Which one should you use?

If you're just starting out and can only pick one, use NPS — it's the most flexible and benchmark-able. But if you're serious about customer feedback:

NPS once per quarter (relationship health)
CSAT after every specific interaction (transactional quality)
CES at high-friction moments (reduce churn drivers)

The mistakes most teams make

Mistake 1: Sending NPS after a support ticket. This measures how the ticket went, not loyalty. Use CSAT instead.

Mistake 2: Tracking CSAT as "customer satisfaction" overall. CSAT is per-interaction. Averaging all your CSAT scores into one number and calling it "satisfaction" is like averaging restaurant reviews across their breakfast, lunch, and dinner — technically possible, mostly meaningless.

Mistake 3: Using CES only after problems. CES after a broken experience just tells you the obvious (it was hard). CES after a supposedly smooth flow is where you find the hidden friction.

Mistake 4: Writing follow-up questions that anchor. "Why do you feel that way?" is good. "What went wrong?" is terrible — it primes negative answers.

Make them anonymous if you want honesty

Regardless of which metric you pick, the respondent will self-edit if they think you'll retaliate or follow up based on low scores. That's why B2B NPS skews high (customers fear damaging the relationship) and DTC NPS tends to be more honest.

If you want the real number, run it anonymously. Anonymeter supports all three formats (0–10 NPS, 1–5 CSAT, 1–7 CES) with zero-IP anonymity — so you see the honest score, not the diplomatic one.

Bottom line

NPS = long-term loyalty ? quarterly
CSAT = specific satisfaction ? after each interaction
CES = friction ? after effortful journeys

Pick the right tool for the right job, run them anonymously, and your metrics will finally tell you something useful.

Create your first NPS/CSAT/CES survey ?