A customer has a billing issue. They spend 20 minutes navigating your help center, fail to find the answer, open a chat, explain the issue, get transferred to another agent, re-explain the issue, and finally get a refund. The issue is resolved. They rate it 4 out of 5 on CSAT.

But they're exhausted. And the next time they see a competitor's ad, they click it. Because even though the outcome was fine, the effort was awful.

CSAT didn't catch this. Customer Effort Score would have.

What CES Measures

Customer Effort Score measures how easy it was for the customer to get their issue resolved. The standard question is: "On a scale of 1 to 7, how easy was it to get help today?" (1 = very difficult, 7 = very easy.)

Some companies use a 5-point scale or a simple "agree/disagree" with the statement "The company made it easy to handle my issue." The scale matters less than the consistency. Pick one and stick with it.

CES captures something CSAT misses: the process, not just the destination. A customer can be satisfied with the resolution (CSAT 5/5) but deeply frustrated by the process (CES 2/7). CSAT measures the outcome. CES measures the experience of getting there.

Why CES Predicts Loyalty Better Than CSAT

The original CEB (now Gartner) research on CES found that 96% of customers who had a high-effort service interaction reported being disloyal, meaning they would spread negative word of mouth, reduce spending, or switch providers.

Compare that to CSAT, where the correlation with future behavior is much weaker. A satisfied customer is more likely to stay, but plenty of satisfied customers still churn. A low-effort customer is much more likely to stay. The effort metric is a stronger predictor because it's about friction, and friction drives behavior more than satisfaction does.

Think about your own experiences. The subscription you keep paying for isn't always the best product. Usually it's the one that's easiest to use and easiest to get help with. Low effort is sticky. High effort repels.

How to Measure CES

Ask the question immediately after the support interaction is resolved. Same timing principles as CSAT: right at the end of the conversation, not in a follow-up email hours later.

The question: "How easy was it to get your issue resolved today?" with a 1-7 scale. Or: "The company made it easy to handle my issue." Strongly disagree to strongly agree.

One question. No follow-up survey. No demographic questions. Just the effort question and an optional free-text "anything we could improve?"

Calculate your CES as the percentage of respondents who score 5 or above (on a 7-point scale). This is your "easy" percentage. Track it weekly.

Target: 70%+ of respondents should score 5 or above. Below 60% means your support process has significant friction. Above 80% means you're doing very well.

What Drives High Effort

High effort almost always comes from process problems, not people problems. Your agents can be wonderful, but if the process makes customers jump through hoops, effort stays high.

The top effort drivers:

Channel switching. Customer starts on chat, gets told to call. Or starts on email, gets a link to a form. Every channel switch adds friction and frustration.

Repeating information. Explaining the issue to one agent, getting transferred, explaining it again. This is the number one effort driver across all industries.

Multiple contacts for the same issue. The customer contacts you, gets a partial answer, has to come back for the rest. Or the first agent can't resolve it and says "someone will follow up" but nobody does.

Waiting. Hold times, delayed responses, "we'll get back to you in 24-48 hours." Every hour of waiting is effort.

Self-service failure. Customer tries to help themselves, can't find the answer, and has to contact support anyway. The self-service attempt was wasted effort.

How to Reduce Effort

Fix channel switching. Let customers resolve their issue in the channel they started in. If they message you on chat, resolve it on chat. Don't redirect them to email, phone, or a form.

Fix transfers. When a ticket needs to be handed off, transfer the full context. The receiving agent should know what the customer already said and what's been tried. Supp's classification and conversation history automatically travel with the ticket.

Fix multi-contact issues. Track "reopen rate" (tickets that were marked resolved but get reopened or followed up on). A high reopen rate means you're not resolving issues completely on the first try.

Fix waiting. AI helps here directly. Instant classification and auto-responses for simple issues eliminate wait times for the majority of queries. For complex issues, instant acknowledgment ("I've received your message and a specialist will review it within 2 hours") reduces perceived effort even when actual resolution takes time.

CES vs CSAT: Use Both

CES and CSAT measure different things. You need both.

CSAT tells you: are customers happy with the outcome? If CSAT is low, your answers are wrong, incomplete, or delivered rudely.

CES tells you: is the process smooth? If CES is low but CSAT is high, your answers are good but your process is painful. Fix the process.

If both are low, you have a fundamental support problem.

If both are high, you're doing great. Keep monitoring.

The most common pattern is high CSAT with moderate CES. Agents are doing their best, but the systems and processes create unnecessary friction. This is the pattern where AI has the biggest impact: reduce steps, reduce wait times, reduce channel switches, reduce repetition.

Supp's dashboard tracks resolution metrics that map to CES drivers: messages per resolution (fewer = lower effort), time to first response (faster = lower effort), and escalation rate (lower = fewer handoffs). While Supp doesn't include a CSAT survey UI, the analytics give you the operational data to keep effort low.