Customer Effort Score: The Survey That Shows Where Support Feels Hard

Customer effort score (CES) is a metric built on a single question: how easy was it for the customer to get a specific issue handled? It does not measure affection for the brand or overall satisfaction, only the work the customer had to do themselves. Hunting for an order number, retyping the same details for a second agent, waiting for another department to reply. Such obstacles rarely lower a politeness rating, yet discourage the next contact.

In this guide you will settle the wording, the scale, the moment to ask and the way to calculate the score. You will also see how to read the answers so they lead to fixes rather than another slide. A comparison of CES with CSAT and NPS is in the guide to customer satisfaction surveys; here the focus is the effort question alone.

What customer effort score is and when it makes sense

CES refers to one completed case. The customer judges whether filing a complaint, changing their details or getting an answer took little or a lot of work on their side. That is why the metric is most useful where a case has several steps: a form, a handoff to someone else, sending a document, a confirmation.

Effort and satisfaction are not the same thing. A customer may appreciate a warm tone and still feel that giving their contract number three times was pointless. The opposite happens too: a quick, terse reply can get a cool rating even though the issue was closed within a minute. Only a separate effort question shows which obstacles come from the process rather than from the way the conversation was handled.

Matthew Dixon, Karen Freeman and Nick Toman challenged the belief that loyalty is built mainly by delighting customers in a 2010 article. It is the background to the effort question, but it does not give you a ready threshold for your industry: Harvard Business Review, Stop Trying to Delight Your Customers.

How to word the effort question

You will come across two common forms. The first is a statement the customer agrees or disagrees with, for example "The company made it easy for me to handle my issue." The second is a direct question, "How easy was it to get this issue resolved?", with a scale from "very difficult" to "very easy." Both work as long as higher values always mean less effort. A reversed scale, where 7 means "a lot of work," confuses customers and report readers alike.

Name the specific case in the question, not the whole company. "Changing my delivery address was easy" gives a score you can attribute to a process; "Working with us is easy" mixes everything the customer has ever experienced. Avoid double-barreled questions such as "quick and easy" too, because you cannot tell which half the rating refers to.

Salesforce gives "The company made it easy for me to handle my issue" as the typical wording and lists 1 to 5, 1 to 7 and percentage-agreement scales, with the score calculated as an average. It also notes that scores are only comparable within the same scale. This describes practice; it does not prove that any one scale produces better results: Salesforce, Customer Effort Score.

Scale and how to calculate the score

Pick one scale and one calculation method and keep both while you compare periods. The average is the simplest, but it hides the distribution: a score of 5.0 can mean everyone answers 5, or that some customers rate the case 7 and others 2. The share of "easy" answers is clear for leadership, but it loses the middle of the scale.

↔ Scroll the table sideways to see all columns.

A fictional example: a team collected 40 responses in a month after return cases. The average came to 5.4, but 9 people chose 1 to 3. The numbers are neither a benchmark nor a result from any Meetlead customer; they only show why you always report the count of low ratings and the sample size next to the average.

When to ask and whom to include

Ask after the case is closed, not after the agent's first reply. A customer rating effort halfway through does not yet know whether they will have to come back. A survey sent several days later meets a fading memory, and the answers may come mostly from the very pleased or the very disappointed. For cases handled over several messages, a good moment is the confirmation that the problem has been resolved.

1. Pick one case type.Start with one multi-step process, such as returns or address changes, not every conversation at once.
2. Set the moment to ask.Ask after the resolution is confirmed, once per case, without same-day reminders.
3. Write down exclusions.Decide in advance whether you skip spam, tests and cases the customer abandoned, and apply that rule in every period.
4. Show the sample size.Next to every score, report the number of responses and the number of closed cases they come from.

There is no single response count that is "enough." The smaller the sample, the more carefully you should talk about month-to-month change and the more attention individual comments deserve. Also check who does not respond. If the survey is answered mostly by customers with simple questions, the score will look better than the real process for difficult cases.

How to read the score without fooling yourself

Compare CES within the same case type. A complaint about damaged goods always takes more work than a question about delivery dates, so a combined average depends mostly on the mix of the two. Split the results by case category and channel first, and only then draw conclusions.

Ease does not prove resolution. A customer can find a conversation simple and come back two days later with the same problem. Pair CES with the effectiveness measure described in the guide to first contact resolution, and check whether low ratings cluster in cases with repeat contacts.

Qualtrics describes CES as a single-item metric and notes that it does not always give the full picture of the customer relationship, so it should be used in tandem with NPS. The source names no universal "good" score: Qualtrics, Customer Effort Score.

From answers to a process fix

Once a month, read every conversation rated 1 to 3 and assign each one main obstacle. The same categories tend to repeat: the customer gave their details twice, waited for another department, did not know which document to attach, or had to switch channels. Pick one obstacle, make a specific change and watch the next cases of the same type. Tagging conversations and tickets keeps the obstacle names consistent.

• Did the customer have to repeat information they had already given?
• Did the handoff to another person keep the full context of the case?
• Did the customer know from the start which details or documents you needed?
• Did they have to switch channels to finish the case?
• Did they get a clear message about what happens next and when?

Meetlead analytics has a native CSAT rating with an optional comment, shown with its sample size and with a way into the underlying conversations. The product has no separate CES survey. If you collect effort answers in another tool, you can tag the conversation and add the rating in an internal note so you see it next to the full case history during review. A monthly overview needs only one line with the score and response count, as in the customer service report.

Questions about customer effort score

Does CES replace CSAT? No. CSAT tells you how the customer rates the service; CES tells you how much work it took them to get the issue handled. If you already ask for satisfaction after a conversation, add the effort question only for selected multi-step processes.

What is a good CES score? There is no fair threshold shared by every company, because the result depends on the scale, the wording, the type of cases and who responds. Treat a drop in low ratings after a specific change as success, not getting closer to a number you found online.