CSAT
An indicator that expresses, as a percentage or an average, the score a customer gives to a specific interaction right after it ends.
Also known as: customer satisfaction score · csat score · how to calculate csat · csat index
Try CRPRO free for 7 daysCSAT (Customer Satisfaction Score) is the score a customer gives to a specific interaction, collected right after it ends. It is the most direct of the satisfaction indicators: one question, a short scale, and a number you can break down by agent, channel, and week.
How it is calculated
The most used scale runs 1 to 5. Two formulas circulate in the market, and they do not produce the same result:
Top-2-box, the most common: responses of 4 and 5 divided by the total responses, times 100. With 100 responses and 78 of them between 4 and 5, the CSAT is 78%.
Simple average: the sum of the scores divided by the number of responses, resulting in something like 4.2 out of 5.
Neither is wrong. What is wrong is switching between them, or comparing your 78% with another company's 4.2. Pick one formula, write down which one it is, and keep it.
CSAT, NPS, and CES do not measure the same thing
| Indicator | Scope | Moment | Typical question |
|---|---|---|---|
| CSAT | one interaction | right after contact | "How do you rate this support interaction?" |
| NPS | the relationship with the brand | cyclical, quarterly | "How likely are you to recommend the company?" |
| CES | the customer's effort | after resolving a ticket | "Was it easy to get what you needed?" |
What target makes sense
There is no universal number, and comparing CSAT between companies almost always compares different methodologies instead. CSAT tends to start high because whoever answers tends to be whoever had a good experience or goodwill toward the brand — an 85% score can mean excellence, or simply that the unhappy customers left without answering. What matters is your own historical series and the breakdown by agent, channel, and topic type. A six-point drop in one week says more than the month's absolute value.
Where CSAT misleads
- It measures the agent, not the company. A customer gives a 5 to someone who was polite but did not solve anything, and a 1 to someone who was flawless while delivering a policy-driven "no."
- A small sample becomes noise. With eight responses in the week, a single score of 1 tanks the agent's indicator and does not mean anything.
- The score arrives before the outcome. Measured at the end of the conversation, CSAT captures relief; the order that arrives wrong three days later never makes it into that score.
- It becomes a target and gets contaminated. Once CSAT enters compensation, you start seeing agents asking for a score and hard cases nobody wants to close.
- It does not explain anything on its own. Without an open field for the reason, you are left with a number that rises and falls without anyone knowing why.
CSAT on WhatsApp
Automatically firing the survey when the conversation ends is what makes CSAT viable for a small operation — asking for a score manually always ends up forgotten on the busiest days, which are exactly the days when the score matters most. Use buttons or a numbered list from 1 to 5 with a label, send it within the 24-hour window so it does not depend on an approved template, and save the response to the contact's record, so a low score turns into a follow-up instead of a line in a report.
Frequently asked questions
How do you calculate CSAT?
In the most used formula, you divide the number of positive responses by the total responses and multiply by 100. On a 1-to-5 scale, the positive responses are the scores of 4 and 5. There is also a simple-average version, which sums the scores and divides by the number of responses, but the two are not comparable to each other.
What is a good CSAT?
There is no universal benchmark, because the number depends on the scale, the formula, and who answers. High scores are common precisely because most unhappy customers do not respond. Instead of chasing a market number, track your own historical series and investigate sharp drops by agent and topic type.
CSAT or NPS: which one should I use?
CSAT when the goal is to fix support, because it measures one episode and points to a specific conversation. NPS when the goal is to track the relationship with the brand over the year. A small operation usually starts better with CSAT, since it generates action the next day instead of next quarter.
How often should I send the CSAT survey?
After every closed interaction, automatically, but with a limit on how often you survey each customer so you do not exhaust whoever talks to you every week. A common cap is one survey per contact every fifteen or thirty days. If the same customer gets a survey every single time, the response rate collapses and the base gets annoyed.
Can CSAT be an individual target for the agent?
It can, but that is the fastest way to ruin the indicator. Once the score turns into money, you get agents asking for a rating, avoiding logging hard cases, and choosing which conversation to close. Use CSAT as a team thermometer and a coaching tool, and save individual targets for indicators the agent truly controls.
Related pages
Related glossary terms
Satisfaction survey
A short questionnaire sent after a contact to turn the customer's perception of that episode into a comparable score.
NPS (Net Promoter Score)
A loyalty metric calculated from the 0-to-10 question about the likelihood of recommending, subtracting the percentage of detractors from promoters.
Customer satisfaction
The result of comparing the expectation a customer had against the experience they perceive they received.
Likert scale
A response scale with ordered points, from strongly disagree to strongly agree, that converts opinion into a comparable number.
Support SLA
An agreed deadline for replying to or resolving a support request, used as an operational target and measured against each conversation's real time.