# CSAT vs CES: which should a small team track?

> CSAT asks how satisfied a customer is with an experience, and CES asks how much effort a task took. CSAT suits support and service moments, and CES suits setup and multi-step flows. Either score is more useful when a written comment explains it.

Source: https://escutaproduto.com/resources/csat-vs-ces
Last updated: 2026-10-11

## What is CSAT and what does it ask?

CSAT stands for customer satisfaction score. It asks a customer to rate how satisfied they are with a specific experience, usually on a short scale such as one to five. A typical question is "How satisfied were you with this support conversation?" It measures the mood of one moment, which makes it useful for checking whether an interaction or a change landed well.

## What is CES and what does it ask?

CES stands for customer effort score. It asks how much work a task took. A typical question is "How easy was it to finish setting up your account?" It measures friction. A task that takes too much effort is often the point where people give up, so CES helps you find the steps that need attention.

## Which question does each score answer?

The two scores answer different questions, so they point to different fixes:

| Question | Score | What a low result usually suggests |
| --- | --- | --- |
| How did this feel? | CSAT | The experience disappointed, but the cause is not yet clear |
| How hard was this? | CES | The steps are too many, unclear or confusing |

A low CSAT after a billing change may mean the message to customers needs work. A low CES on a setup flow points at steps, labels or missing guidance. Choose the score that matches the decision you need to make, and avoid asking both questions in every context, because that doubles the effort you ask of customers.

## Why does an open comment matter more than the score?

A number tells you that something went wrong. A comment tells you what it was. Two customers can both rate a task two out of five for entirely different reasons. One could not find the export button, and the other did not trust the result. Only the written message shows which fix matters.

That is why any score should sit beside a text box. The score makes trends visible over time, and the comment makes those trends actionable. A score without a comment is a chart with no explanation, and teams tend to argue about what it means instead of acting on it.

Scales need care as well. A five-point scale with labels only at the ends invites people to pick the middle option when they are unsure, which can make a neutral result look like a verdict. Label the points if you use more than two, and decide in advance what counts as a good result for your team, so the number is read the same way each month. Changing the wording or the scale midway through breaks the trend, and a broken trend is hard to interpret. If you must change the question, start a new series and note the date of the change.

Ask at a consistent moment too. A score collected right after a support reply reflects that reply, while a score collected at random reflects the mood of the day.

## Which one should a small team track first?

Start with the open message, which costs nothing and needs no survey design. Read the messages each week and tag the ones that describe effort or satisfaction in your own words, using internal notes. Then add a single rating question at a moment you choose, such as after a support reply or a completed onboarding step.

Track CES first if your product has a setup flow or a multi-step task where customers drop off. Track CSAT first if your main contact with customers happens through support or another service interaction. In either case, keep the scale identical over time, so a change in the trend means something real. A two-point change between two scores is rarely meaningful on its own, so look for a sustained shift across several weeks.

## How do you capture a score and a comment together?

Put the question and the comment in the same moment. Ask once, right after the task or conversation ends, and let the customer write a sentence if they want to. Waiting a week before asking makes people forget the details that would have explained the score. Keep the question short, and never ask for a score on a screen the customer is using to finish something urgent. For wording the button that opens the form, see [what a feedback button should say](/resources/feedback-button-copy). For wording the button that opens the form, see [what a feedback button should say](/resources/feedback-button-copy).

## Capturing a score and a comment in Escuta Produto

Escuta Produto does not run survey campaigns or send scheduled questions, so the score question is yours to design. The widget can show an optional one to five star rating in the same form as the message, and the dashboard shows the average rating for each product next to counts by type and a 30-day chart. Treat it as a lightweight satisfaction signal, and keep the written message as the main input.

Open the inbox, filter by praise or other, and read the comments that come with low ratings first. Add an internal note that names the effort or satisfaction problem in your own words, so the pattern is visible when you review it later. The [widget documentation](/docs/widget) lists the rating option, and [the comparison of NPS and open-ended feedback](/resources/nps-vs-open-ended-feedback) explains why the comment usually carries more signal than the number.

## Frequently asked questions

### What is the difference between CSAT and CES?

CSAT asks how satisfied a customer is with an experience. CES asks how much effort a task took. CSAT suits support and service moments, while CES suits setup flows and multi-step tasks where friction causes drop-off.

### Which score should a small team track first?

Start with the open written message, then add one rating question. Choose CES if your product has a setup or multi-step flow where people drop off, and CSAT if your main contact with customers is support.

### Is a score enough to understand customer feedback?

No. A score shows that something went well or badly, but not why. A short written comment explains the cause, which is what your team needs in order to decide on a fix.