Metrics & KPIs
CES (Customer Effort Score)
CES measures how much effort customers spend resolving issues. Strongest predictor of churn for support interactions.
CES measures how much effort customers spend resolving issues. Strongest predictor of churn for support interactions.
Customer Effort Score (CES) is a customer experience metric that measures how easy or difficult it was for a customer to resolve their issue, complete a task, or get what they needed. It is calculated from a single survey question — "How easy was it to handle your issue today?" — answered on a 1-5 or 1-7 scale.
CES has emerged as the strongest predictor of customer loyalty for support and service interactions. Research from CEB (now Gartner) found CES is more predictive of repeat purchase and likelihood to recommend than CSAT.
Different survey designs produce different formulas. The two most common:
5-point scale (Likert): "How easy was it to resolve your issue?" → 1 (Very Difficult) to 5 (Very Easy). CES = average score.
7-point scale (Disagree/Agree): "The company made it easy to handle my issue." → 1 (Strongly Disagree) to 7 (Strongly Agree). CES = average score, or % responses scoring 5-7.
A higher CES means lower customer effort, which is good. Most contact centers report the average score (e.g., "CES of 5.8 out of 7").
| Score Range (1-7) | Interpretation |
|---|---|
| 6.0+ | Top-quartile — minimal customer friction |
| 5.5 - 6.0 | Strong — most issues resolved easily |
| 5.0 - 5.5 | Average — room to reduce friction |
| Below 5.0 | High friction — investigate root causes |
For 5-point scales, multiply by 7/5 to roughly compare. Top-quartile in a 5-point system is around 4.3+.
These three metrics answer different questions:
CES is the strongest predictor of churn for support interactions. CSAT is the easiest to measure. NPS is the most strategic for executive reporting. High-performing contact centers track all three.
| Driver | What's Happening |
|---|---|
| Multiple transfers | Customer has to repeat the issue to each agent |
| Repeat verifications | Same security/identity questions each call |
| Channel switching | Customer pushed from chat to phone to email |
| Required callbacks | Issue not resolved on first contact |
| Confusing IVR | Customer gets routed wrong or can't reach an agent |
| Holds without context | Long holds without explanation |
Each of these behaviors increases effort and decreases CES.
Customer effort is signaled by call characteristics that AI can analyze on every call: number of transfers, repeated phrases, customer frustration markers (sighs, "I already told someone"), dead air duration, hold time, and whether the issue resolved on first contact (FCR). AI-powered call auditing scores effort as a composite metric across 100% of calls — surfacing the agents, queues, and issue types where effort is highest. Contact centers using automated call scoring typically reduce average customer effort 10-20% within 90 days because root causes (KB gaps, transfer patterns, slow systems) become visible at scale.
On a 7-point scale, 5.5+ is good and 6.0+ is top-quartile. On a 5-point scale, 4.0+ is good and 4.3+ is top-quartile. Compare against industry baselines and track trend over time more than absolute number.
Within 24 hours of the interaction, while the experience is fresh. Most contact centers send CES via SMS or email immediately after call completion.
CES measures effort (process). CSAT measures satisfaction (emotion). A customer can rate CSAT highly because the agent was friendly, while still rating CES poorly because the issue took multiple calls to resolve. Both metrics give different views.
No, but it complements them. AI QA infers effort indicators from call audio (transfers, repeat info, frustration markers) on 100% of calls. Survey response rates are typically 5-15% — AI fills the gap on the other 85-95%, predicting effort from observable call characteristics.
Last updated: April 2026
