Open any quarterly CX review and the same three numbers are on the slide. CSAT at 4.4. NPS at 42. AHT at five minutes and change. Your competitor’s slide, if you could see it, would look almost identical.
None of those numbers answer the question a customer would ask: how much work did this take me? However, a customer effort score does, and it tends to be missing from exactly the dashboards that are most confident about their CX.
In this article, we look at what the standard metrics leave out, what effort measures instead, how to start tracking it without rebuilding your reporting, and how interaction data surfaces effort without asking anyone a single survey question.
What the Standard Metrics Don’t Capture
Each of the usual three measures something real. The problem is what falls between them.
- CSAT captures how someone felt about one interaction, usually asked right after it ended, when relief and satisfaction are hard to tell apart.
- NPS captures stated intent toward the brand as a whole, which moves with pricing, product and advertising as much as with service.
- AHT captures your cost. A five-minute call says nothing about whether it was the customer’s first call or their fourth.
What none of them record is the work the customer did.
Finding the number in the first place. Repeating account details to a second person. Switching from chat to phone because the bot couldn’t complete the request. Waiting for a callback that never came.
Most of that work never reaches your survey
Survey-based CX metrics have a coverage problem that keeps getting worse. That is exactly why some underused CX metrics deserve a second look.
Qualtrics XM Institute’s 2026 research found fewer than one in three consumers now give companies direct feedback, an all-time low. Meanwhile, 34% reduce spending after a negative experience and 13% stop spending with that company entirely.
Customers rarely announce that you were hard to deal with. They route around you instead.
Gartner found in 2025 that more than half of customer service journeys now start on third-party platforms, rising to 74% among Gen Z customers. They look for an answer on a search engine, a forum, or an AI assistant first.
Someone who solves their problem on Reddit never appears in your CSAT. Your score stays healthy while your channels quietly become the last resort, and a blind spot.
Customer Effort Score: What It Measures and Why It Predicts Loyalty
The metric came out of research by CEB, now part of Gartner, which set out to test a comfortable assumption: that delighting customers builds loyalty.
The finding was less flattering. Reducing the effort customers spend to get a problem solved predicted loyalty better than exceeding their expectations did.
The measurement itself is a single statement, rated on a seven-point agree-to-disagree scale: the company made it easy for me to handle my issue. Everything interesting is in what that question catches that the others don’t.
The table below puts CES next to the two metrics most teams already report. Read the last column first: it shows what each score can’t see, which is the reason to add another one.
| Scores | What it asks | What it captures | Blind spot |
| CSAT | How satisfied were you? | Feeling about one interaction | Politeness bias, and no memory of the three previous attempts |
| NPS | Would you recommend us? | Brand-level sentiment | Moves with price and product, slow to react to service |
| CES | Was this easy for you? | Work the customer had to do | Needs a defined journey to be meaningful |
CSAT and NPS aren’t wrong; they just share a blind spot: neither records how much work the customer did to get help. CES is built to cover that gap, so treat it as a companion to your existing scores, not a replacement. Use CSAT to see how a single interaction felt, NPS to track the brand relationship, and CES to find the processes creating friction.
CES has a limit of its own, too. It only means something when it’s tied to a defined journey, which is why we recommend starting with one later in this article.
Why it’s harder to game
An agent can influence CSAT by asking nicely at the end of a good-humored call. Effort is less cooperative. A customer who has called three times about the same issue will register that, however pleasant the third agent was.
That’s also why effort correlates with behavior you can see elsewhere. Repeat contacts, transfers, and channel switches are effort made visible in your own operational data, whether or not anyone fills in a survey.
It shows up in language too, which is why conversation analysis picks up friction that never reaches a form. We wrote about that mechanism in how AI reads between the lines of customer conversations.
How to Start Tracking It Without a Big Process Overhaul
Nobody needs a new research program. Start narrow and let the number earn its place.
- Add one question, don’t build a new survey: Put the effort statement into the survey you already send. Survey fatigue is real, and a second questionnaire will lower response rates on both.
- Pick one journey: Billing disputes, returns, or cancellations are good candidates, because they carry effort naturally and matter commercially. A company-wide average tells you nothing you can act on.
- Use operational proxies alongside it: Repeat contact rate within seven days, transfers per resolution, channel switches per case, and repeated authentication are all measurable today with no survey at all.
- Set a baseline before you change anything: Four weeks of data gives you something to compare against when you fix the process you suspect is the problem.
- Read the low scores with the conversation attached: A score of 2 is a signal; the transcript tells you it was a broken self-service flow sending people to the phone line.
If you’re refreshing the survey side of this, keep it to one or two questions and send it in the channel where the interaction happened. That alone usually separates a 6% response rate from a workable one.
The mechanics of setting that up sit in customer satisfaction surveys.
How CX Insights Surfaces Effort Signals From Interaction Data
Surveys sample the customers willing to answer. Interaction data covers everyone who contacted you, which is the more interesting population when you’re measuring friction.
CX Insights detects effort and friction signals at the interaction level: the caller who says this is the third time, the chat that repeats information already given, the conversation where sentiment drops the moment a transfer is mentioned.
These are the moments a survey would have recorded as a 2, if the customer had bothered to fill it in.
Signals worth watching
- Repeat contact language: “As I explained last time” is an effort score in plain speech.
- Channel switching inside one issue: Chat, then email, then a call, all in two days.
- Re-explaining: The customer restating the same problem after a transfer.
- Process friction: Long holds, repeated verification, references to a form that wouldn’t submit.
Trend tracking is where this stops being anecdotal. Effort signals are tracked over time, by topic and by queue, so you can see whether last month’s self-service change reduced friction or simply moved it.
That’s the same shift described in from reactive to predictive CX work: catching patterns while they’re still forming.
Interaction data also gives you the why, not only the score. A drop in effort scores on billing tells you something is wrong; the conversations tell you an invoice template changed and stopped showing the due date.
See What Your Customer Effort Score Might Be Telling You
If your CX dashboard looks like everyone else’s, it will produce the same decisions everyone else is making. Effort is the number that points to the specific processes making your customers work.
Book a demo to see how CX Insights detects friction signals across your conversations and tracks them over time, so effort becomes something you can manage rather than something you find out about later.
FAQ
What is customer effort score (CES)?
Customer effort score (CES) measures how easy it was for a customer to get their issue resolved. Customers rate one statement, “the company made it easy for me to handle my issue,” on a seven-point agree-to-disagree scale. Lower effort predicts loyalty better than exceeding expectations.
What is the difference between customer effort score, CSAT, and NPS?
CSAT measures how a customer felt about one interaction, and NPS measures how likely they are to recommend your brand. Customer effort score measures the work the customer had to do, such as repeating details or switching channels. That makes it the metric that catches friction the other two miss.
How do you measure customer effort score?
Add the customer effort score statement to a survey you already send, and start with one journey such as returns, billing disputes, or cancellations. Set a four-week baseline before changing anything. Repeat contact rate, transfers per resolution, and channel switches can run alongside it as operational proxies.
Can you measure customer effort without surveys?
Yes. Customer effort shows up in interaction data as repeat contact language, channel switching within one issue, re-explaining after a transfer, and long holds. CX Insights detects these friction signals across all conversations and tracks them over time by topic and queue, without relying on survey responses.
