Loyalty vs friction - two metrics that predict churn from different angles

NPS vs CES

Net Promoter Score (NPS) measures whether customers love you. Customer Effort Score (CES) measures how hard you make them work.

Both predict churn - but from different angles and at different timescales. CES catches friction before it shows up in NPS decline.

NPS vs CES at a glance

Net Promoter Score (NPS)

“How likely are you to recommend us?”

Type: Relational - measures the overall relationship
Scale: 0–10 (Promoters 9–10, Passives 7–8, Detractors 0–6)
Output: Score from -100 to +100
Timing: Quarterly or after major milestones
Best for: Loyalty trends, churn prediction, referral likelihood
Benchmark: +30 is good, +50 is excellent (SaaS)

Customer Effort Score (CES)

“How easy was it to resolve your issue?”

Type: Transactional - measures a specific interaction
Scale: 1–7 (higher = easier, less effort)
Output: Average score from 1–7
Timing: Immediately after each interaction
Best for: Friction detection, support quality, process improvement
Benchmark: 5.5+ out of 7 is good (SaaS)

When to use NPS vs CES

The most common mistake is using a relational metric (NPS) to diagnose transactional friction - that's what CES is for.

Use NPS when you want to know...

Is overall customer loyalty improving or declining?

Run quarterly. Track the trend over 4–8 quarters, not individual scores.

Which customers are most likely to refer others?

Promoters (9–10) are your referral engine. Study what they have in common.

Which customers are at churn risk before it happens?

Detractors (0–6) are warning signals. Follow up within 48 hours.

How does the relationship change after a major product update?

Run NPS before and after a significant release to measure overall impact.

How do we compare to industry loyalty benchmarks?

NPS has well-published SaaS industry benchmarks. Use them for competitive context.

Use CES when you want to know...

How easy was our support interaction to navigate?

CES on support is one of the highest-signal uses. High effort = likely to churn.

Was onboarding smooth or did users struggle to get started?

Trigger CES after the onboarding flow completes, not days later when the context is gone.

Is the checkout or upgrade flow creating unnecessary friction?

High effort at purchase conversion = immediate revenue leakage.

Where in our product are users working too hard?

Map CES across key workflows to find where effort spikes. Each spike is a retention risk.

Are we reducing friction over time as we improve the product?

Track CES on the same workflows quarter-over-quarter to measure improvement.

NPS vs CES vs CSAT

All three core CX metrics - what they measure, when to send them, and what they predict.

Dimension
NPS
CES
CSAT
Question asked
"How likely to recommend?"
"How easy was that?"
"How satisfied were you?"
What it measures
Loyalty / advocacy
Effort / friction
Satisfaction / experience
Type
Relational
Transactional
Transactional
Scale
0–10
1–7
1–5
Output
-100 to +100
1–7 average
0–100%
When to send
Quarterly / after milestones
After support / onboarding / checkout
After each interaction
Good benchmark (SaaS)
+30 to +50
5.5+ out of 7
80–85%
Predicts
Referral likelihood, renewal
Churn (high effort = high churn)
Repeat purchase, loyalty
Weakness
Doesn't pinpoint problems
Narrow scope (friction only)
Not predictive of long-term loyalty

CES as a leading indicator

CES catches churn before NPS shows it

NPS declines after users have already decided to leave. CES spikes at the moment users struggle - giving you a chance to fix the friction before it becomes a cancellation.

Warning pattern

NPS looks healthy but early-stage churn is rising

Support volume increasing but NPS has not moved yet

Onboarding completion rate dropping - CES spike you haven't measured

NPS declining quarter-over-quarter with no obvious cause

CES diagnosis

Run CES on support, onboarding, and checkout separately

Track CES trend for each touchpoint alongside quarterly NPS

Low CES at onboarding + declining NPS = fix the activation flow first

Low CES at support + stable NPS = localized problem, fix before it spreads

Frequently asked questions

NPS vs CES - common questions answered.

Get started today

Run NPS and CES together to catch churn before it happens

Every response linked to the real user who gave it. Segment by plan, role, or cohort. No spreadsheets required.

Run NPS and CES Surveys Free

No credit card required