NPS vs CSAT vs CES: Which One to Use?

NPS, CSAT, and CES measure completely different things. Using the wrong one for the situation gives you meaningless data.

Here is when to use each and how they complement each other.

What each metric measures

When to use each

Use NPS when:

• Measuring overall sentiment quarterly
• Segmenting customers by advocacy level (promoters vs detractors)
• Identifying at-risk accounts (detractors are 3-5x more likely to churn)
• Benchmarking against category peers

Read: What is NPS?

Use CSAT when:

• After a support ticket resolves
• After a feature launch or product release
• After onboarding step completes
• Rating a specific product area

Read: What is CSAT?

Use CES when:

• Measuring how easy a process is (onboarding, complex task completion)
• Diagnosing friction in a customer journey
• Predicting churn from specific interactions
• Prioritizing which processes to redesign

Read: What is CES?

Which metric predicts churn best

Research on this is clear: CES predicts churn better than NPS or CSAT.

The reason: effort is the strongest driver of disloyalty. Customers who find things hard are more likely to churn than customers who are just neutral or unenthusiastic. "Delight" moments do not offset "difficulty" moments in memory.

That said, NPS is more useful for account triage. A detractor response gives you a specific customer to call. A low CES gives you a specific process to fix. Different operational value.

How to use all three together

Most SaaS teams should run all three, but not on every touchpoint. Survey fatigue is real.

A recommended cadence:

• NPS: Quarterly, sent to a segment of active customers
• CSAT: After every support ticket, after major feature releases
• CES: After onboarding completes, after high-effort processes (billing changes, migrations)

Common mistakes picking between them

1. Using NPS for specific product feedback

NPS is too broad to diagnose specific issues. If you want to know how a new feature is received, use CSAT for that specific feature.

2. Using CSAT for overall sentiment

Averaging CSAT across all interactions is not the same as knowing overall sentiment. NPS asks the right question for that.

3. Not using CES at all

Many SaaS teams skip CES because it feels redundant with CSAT. But effort is a different signal from satisfaction. A customer can be satisfied with the outcome but frustrated with the process.

4. Running all three on the same touchpoint

Do not send NPS, CSAT, and CES surveys after every interaction. Pick the metric that fits the moment.

The takeaway

Different metrics for different questions:

• Want to know how customers feel overall? NPS.
• Want to know if a specific thing worked? CSAT.
• Want to know where friction is hurting you? CES.

For the full retention picture and how to fix satisfaction gaps that predict churn, take the 60-second Churn Health Check.