How to Calculate CSAT Score (Customer Satisfaction)

To calculate a CSAT score (Customer Satisfaction Score), ask customers to rate their satisfaction with a specific interaction on a 1 to 5 scale, then divide the number of 4 and 5 ratings by the total number of ratings: CSAT = (4s + 5s) ÷ all responses × 100. This "top-2-box" method gives you the percentage of satisfied customers. For example, 125 satisfied answers out of 150 give a CSAT of 83.3%. Some teams report the average rating instead, so always state which method you use.

This guide explains both methods, walks through a worked example, and covers the timing and bias issues that make the difference between a useful CSAT and a misleading one. The free CSAT calculator calculates both versions from your data.

The CSAT question

How satisfied were you with [the support you received / your delivery / your appointment today]?

Rating	Label
1	Very dissatisfied
2	Dissatisfied
3	Neither satisfied nor dissatisfied
4	Satisfied
5	Very satisfied

Make the question specific. "How satisfied are you with us?" is vague; "How satisfied were you with today's delivery?" tells you exactly what was measured.

Follow it with one optional open question:

What is the main reason for your rating?

Method 1: Top-2-box (percentage satisfied)

CSAT = (number of 4 and 5 ratings) ÷ (total number of ratings) × 100

Why only 4 and 5? Because a neutral 3 is not satisfaction. Counting only clearly satisfied customers produces a stricter, easier-to-interpret figure.

Method 2: Average rating

Average = sum of all ratings ÷ number of ratings

The average uses every answer, including how dissatisfied unhappy customers are. A shift from 2s to 1s changes the average, but not the top-2-box score.

Worked example

After one month of deliveries, you receive 150 ratings:

Rating	Count	Rating × count
1	4	4
2	6	12
3	15	45
4	45	180
5	80	400
Total	150	641

Top-2-box CSAT: (45 + 80) ÷ 150 = 125 ÷ 150 = 83.3%

Average rating: 641 ÷ 150 ≈ 4.27 out of 5

Bottom-2-box (dissatisfied): (4 + 6) ÷ 150 = 6.7%

Reporting the dissatisfied share alongside CSAT is useful: two months can both show 83% CSAT while the dissatisfied share doubles, because neutral answers moved into the negative.

Which method should you use?

	Top-2-box %	Average
Easy to communicate	Yes ("83% satisfied")	Moderately ("4.3 out of 5")
Sensitive to severity of dissatisfaction	No	Yes
Comparable across 5- and 7-point scales	Roughly, if defined consistently	No
Common in support teams	Very	Common

For most small businesses, top-2-box as the headline with the distribution next to it works well. Whatever you choose, never switch between methods without noting it.

When to send a CSAT survey

Timing affects both response rate and accuracy.

Interaction	When to ask
Support chat or email	Right after the ticket is marked solved
Phone call	Same day, by email or text, if the customer has agreed to be contacted that way
Delivery	A day or two after delivery, once they have opened it
Appointment or visit	Same day or the next morning
Project milestone	Right after handover

Ask too late and customers either do not respond or answer based on their overall feeling rather than the specific interaction.

Biases that distort CSAT

1. Selection bias

Who answers is not random. Very happy and very unhappy customers are more likely to respond than those in between. Your CSAT describes respondents, not necessarily all customers. Track your response rate alongside the score; if it drops sharply, the score may shift for that reason alone.

2. Surveying only good outcomes

If surveys go out only after resolved tickets or successful deliveries, failures never get measured. Survey after every interaction of a given type, not just the ones that went well.

3. Leading wording

"How satisfied were you with our fast and friendly service?" tells the customer what to think. Neutral wording keeps the answers honest.

4. Pressure from staff

"Please give me a 5, it really helps me" inflates scores and makes them useless. If individual staff members are judged by CSAT, this pressure appears quickly. Use CSAT to find process problems rather than to rank people.

5. Scale direction and labels

Changing scale direction (5 = best vs 1 = best) or labels between surveys breaks comparability. Keep the scale fixed.

6. Small numbers

With 20 responses, each answer moves the score by 5 percentage points. Look at monthly or quarterly totals rather than reacting to daily swings.

Using CSAT to improve service

A CSAT score on its own does not tell you what to fix. Make it actionable:

1. Tag the open comments by theme (speed, friendliness, accuracy, product quality, price).
2. Break CSAT down by interaction type, channel, product or team.
3. Follow up on low ratings personally within a day or two. A dissatisfied customer who receives a genuine response often becomes more loyal than one who never had a problem. Use the process in how to handle customer complaints.
4. Fix the top theme before the next reporting period, then check whether that segment's CSAT moves.

Worked example: finding the problem behind a score

Say a small online shop's delivery CSAT falls from 86% in one quarter to 78% in the next, with around 300 responses each quarter. The overall number says something is wrong but not what.

Breaking the responses down by delivery region shows that most regions are unchanged, but one region dropped sharply. Reading the comments from that region, the most common theme is "parcel left outside" and "arrived damaged". A check with the courier reveals a new local depot with a different drop-off practice.

The shop asks the courier to require a signature for that region, emails affected customers with an apology, and adds a line to its order confirmation about safe delivery options. The next quarter, CSAT for that region recovers.

Two lessons. First, the open comments did the diagnostic work; the score only raised the alarm. Second, breaking CSAT down by a dimension you can act on (region, product, channel, team) turns a vague worry into a specific fix. If a public review mentions the same problem, reply using the approach in how to respond to negative reviews.

CSAT and the other metrics

CSAT tells you about individual moments. It does not tell you how customers feel about the relationship overall, or how easy your processes are. For those questions, use NPS and CES respectively; our guide to NPS vs CSAT vs CES compares all three. The CSAT calculator also includes a CES mode for 1–7 effort ratings.

Common mistakes

• Comparing your CSAT with figures found online. Survey methods and customer groups differ too much for meaningful comparison. Compare with your own history.
• Mixing scales. A 1–5 CSAT and a 1–10 rating cannot be averaged together.
• Measuring without acting. Customers notice when feedback disappears into a void.
• Asking too many questions. Long surveys get fewer and less thoughtful answers.
• Ignoring the 3s. Neutral ratings are often the earliest sign that something is slipping.

Your next step

Pick the one interaction your customers care most about, add a single CSAT question right after it, and collect responses for a month. Then enter the counts per rating in the CSAT calculator, note the top-2-box score and the dissatisfied share, and read every comment from customers who answered 1, 2 or 3. For more ideas on what else to ask, see our list of customer feedback survey questions.