Quick answer
Customer effort score (CES) measures how much effort a customer had to put in to get something done, such as solving a problem or completing a purchase. It is based on one question asked right after the interaction, answered on a numbered scale, and it is usually reported as the average score.
Formula: CES = sum of all effort scores ÷ number of responses
What is customer effort score?
Customer effort score (CES) measures how much work a customer had to do to get something done: solve a problem, find an answer, complete a purchase or change a setting. It comes from one question asked right after that task, answered on a numbered scale. Low effort means the customer found it easy; high effort means they had to push.
The idea was popularised in a 2010 Harvard Business Review article by Matthew Dixon, Karen Freeman and Nicholas Toman of the Corporate Executive Board (now part of Gartner), based on an analysis of more than 75,000 people who had dealt with contact centres or self-service channels. Their argument: delighting customers does not build loyalty as reliably as removing the obstacles that make them work, and effort tends to say more about whether a customer stays than whether they were satisfied.
How the question is worded
CES has no single official wording, which is the first thing to settle before you compare numbers.
| Style | Example wording | Scale | Higher score means |
|---|---|---|---|
| Original effort question | "How much effort did you personally have to put forth to handle your request?" | Five- or seven-point | Depends on how the ends are labelled; check |
| Agree statement | "The company made it easy for me to handle my issue." | Commonly 1–7, strongly disagree to strongly agree | Easier (better) |
| Direct ease question | "How easy was it to get your request handled?" | 1 (very difficult) to 7 (very easy) | Easier (better) |
Notice that an effort question and an ease question run in opposite directions. Sources disagree on scale direction too, so write on the report which direction is good, and never merge answers collected with different directions or different scales.
Ask it immediately after a specific interaction: a support case closed, an order placed, a returns request finished. A CES asked a week later measures memory, not effort.
How to calculate CES
| Method | Formula | Notes |
|---|---|---|
| Average | sum of all scores ÷ number of responses | The most common method; the number stays on the 1–7 scale |
| Top-box share | responses of 5, 6 or 7 ÷ all responses × 100 | Easier to read as "percent who found it easy" |
Zendesk's guide uses the average and illustrates it with 12 responses adding up to 71, which gives about 5.9.
A worked example
Sample data: 120 customers answer "The company made it easy for me to handle my issue" on a 1–7 scale right after a support case.
| Score | 7 | 6 | 5 | 4 | 3 | 2 | 1 | Total |
|---|---|---|---|---|---|---|---|---|
| Responses | 30 | 34 | 24 | 14 | 9 | 6 | 3 | 120 |
| Points | 210 | 204 | 120 | 56 | 27 | 12 | 3 | 632 |
- Average CES: 632 ÷ 120 = 5.27.
- Top-box (5–7): (30 + 34 + 24) ÷ 120 = 88 ÷ 120 = 73.3%.
Now split the same 120 answers by what the customer was trying to do:
| Issue type | Responses | Points | Average |
|---|---|---|---|
| Password reset | 50 | 310 | 6.2 |
| Order change | 40 | 200 | 5.0 |
| Billing dispute | 30 | 122 | 4.1 |
| All | 120 | 632 | 5.27 |
The overall 5.27 looks fine and hides the real story: billing disputes average 4.1, a full two points below password resets. The decision is to look at the billing flow, not to celebrate the average. CES is most useful when you cut it by journey like this.
A survey you can copy
Send one statement right after a support case closes or an order is placed: "The company made it easy for me to handle my issue." Labels: 1 Strongly disagree, 2 Disagree, 3 Somewhat disagree, 4 Neither agree nor disagree, 5 Somewhat agree, 6 Agree, 7 Strongly agree. Add one optional line, "What made it harder than it should have been?", and tag every answer with the issue type so that you can split the results as in the table above. Higher is easier; write that on the report.
When to use CES instead of CSAT
CSAT asks whether the customer was happy; CES asks how hard it was. Use these rules:
- If the thing you want to change is a process (a form, a hand-off, a returns flow), ask CES. Effort points at the step that causes it.
- If you want to judge the quality of the conversation (tone, helpfulness), ask CSAT.
- If a customer can be satisfied with an agent and still have had to contact you three times, CSAT alone will miss it. CES, or a repeat-contact count, will not.
Both are single-interaction metrics. Neither replaces a measure of long-term loyalty.
What is a good CES?
It depends on the scale and the direction, so an external number is weak guidance. As a rough vendor rule of thumb for a 1–7 "easier is higher" scale, an average above 5 is treated as good and above 6 as strong, but that is a heuristic, not a standard. Trust your own comparisons: the same question, the same scale, split by issue type, channel and month. In the example above, the movement of billing disputes from 4.1 to 5.0 after a fix would be a far better signal than any benchmark.
Common mistakes
- Mixing scale directions. Averages from "1 = low effort" and "1 = strongly disagree" surveys cannot be combined.
- Reporting one overall number. The example shows how an average hides the problem; always split by journey.
- Asking too late. Ask right after the interaction.
- Surveying only people who succeeded. If failures never reach the survey, CES looks better than reality.
- Ignoring the sample. Apply the margin-of-error logic from CSAT before acting on a small group.
- Using CES as a score for individual agents. Effort is mostly caused by process, not by one person's skill.
For work on reducing the friction itself, see user friction, and for raising satisfaction, improve CSAT.
Looking for effort in TapCub
TapCub does not run CES surveys and has no built-in effort score; it only measures. What it can do is show where effort probably comes from. Applying the example above:
- Find where customers ask for help. In Live chat › Reports, the hot pages list shows the page where visitors started a chat (without the query string, top 10). If the billing page leads the list, that is where people get stuck.
- Check whether the problem got solved. The same report gives the resolution rate: resolved conversations ÷ all conversations. A low rate on one topic is a high-effort signal.
- Check how long they waited. First response is the average time to the first human reply. AI replies are not counted in it.
- Read the nearest rating. The report's satisfaction score (1–5 within 24 hours after the conversation ends, averaged to one decimal) is a satisfaction measure, not effort, so read it as a neighbour of CES rather than a substitute.
- Watch the page itself. All-platform analytics › Behavior › Heatmaps shows clicks, rapid repeated clicks and scroll depth for a page; All-platform analytics › Behavior › Session replay (off by default) shows how a visit went. Neither is a CES.
If you collect CES with a survey tool, keep that number separate and compare it with these signals by journey.
Sources
- Matthew Dixon, Karen Freeman and Nicholas Toman, Stop Trying to Delight Your Customers, Harvard Business Review, July–August 2010.
- Zendesk, Customer effort score (CES): what it is and how to measure it.
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Product Analyst
Focuses on event analysis, conversion funnels, and retention, using clear metric definitions and structured analysis to explain product performance.