Customer Effort Score (CES) is the customer experience metric that consistently predicts churn and repurchase better than customer satisfaction. Originally published in the Harvard Business Review article Stop Trying to Delight Your Customers, CES captures the single dimension customers actually remember: how hard it was to get something done. This guide covers the formula, the scale, benchmark ranges, when to use CES instead of NPS or CSAT, and example questions you can deploy today.
The CES question and scale
The canonical CES question:
"To what extent do you agree with the following statement: The company made it easy for me to handle my issue."
Answered on a 1–7 agreement scale:
- 1 — Strongly disagree
- 2 — Disagree
- 3 — Somewhat disagree
- 4 — Neutral
- 5 — Somewhat agree
- 6 — Agree
- 7 — Strongly agree
Some platforms offer a 1–5 variant. Both are valid as long as you don't compare across scales.
The CES formula
CES = Sum of all scores ÷ Number of responses
That gives you the average score. A second, more intuitive way to report CES is as the percentage of "easy" responses — scores of 5, 6, or 7 on the 1–7 scale:
CES (%) = (Responses scoring 5, 6, or 7 / Total responses) × 100
Both are valid; the percentage version reads better in non-technical reporting, while the average is more sensitive to small movements.
CES benchmark ranges
| Average (1–7) | Easy % (5–7) | Band |
|---|---|---|
| 6.5+ | 90%+ | World-class |
| 6.0 – 6.4 | 85 – 89% | Excellent |
| 5.5 – 5.9 | 75 – 84% | Good |
| 5.0 – 5.4 | 60 – 74% | Average |
| 4.0 – 4.9 | 40 – 59% | Needs work |
| < 4.0 | < 40% | Critical |
When to use CES instead of NPS or CSAT
The original CEB / Gartner research found that 96% of high-effort customers become disloyal, while only 9% of low-effort customers do. That predictive power doesn't show up in CSAT — a customer can be satisfied and still churn — and it doesn't show up in NPS until months later.
CES is the right metric when:
- The customer just completed a support interaction (ticket close, chat end, call resolution).
- The customer just finished a self-service flow (password reset, account upgrade, return).
- You're diagnosing friction in onboarding or activation.
- You're measuring a redesigned checkout, signup, or upgrade flow.
Stick with NPS for the relationship-level question and CSAT for satisfaction-with-a-thing (a delivery, a product, a service visit).
CES question examples by moment
| Moment | Question |
|---|---|
| After support ticket close | "How easy was it to get your issue resolved?" |
| After service call | "How easy was it to schedule and complete your appointment?" |
| After onboarding | "How easy was it to get set up and ready to use the product?" |
| After self-service flow | "How easy was it to complete what you came here to do?" |
| After billing or payment change | "How easy was it to update your billing information?" |
| After return / cancellation | "How easy was it to return your item / cancel your service?" |
Always pair the score with a single open-text follow-up: "What made it easy or hard?" The score tells you where friction is; the words tell you what to fix.
Common CES mistakes
- Sending CES too late. Effort fades from memory in 24–48 hours. CES surveys should fire within hours of the interaction, not days.
- Mixing scales. Don't compare a 1–7 CES to a 1–5 CES — convert one to the other before any benchmarking.
- Reporting only the average. The distribution matters: an average of 5.5 could be everyone scoring 5 or 6, or a polarized split between 7s and 2s.
- Treating CES as a vanity metric. CES exists to find friction. If you're not tying low scores to specific flows or agents, you're not running CES, you're just collecting it.
Frequently asked questions
Keep reading
- • What is CES? — the short explainer
- • NPS vs CSAT vs CES — pick the right metric for the moment
- • Closing the loop framework
- • Try CES surveys free in AwareCX →