# CSAT: what it is and how to calculate customer satisfaction

> CSAT reports the share of valid answers that meet a defined positive threshold on a stated scale.


**CSAT, or Customer Satisfaction Score, measures how satisfied someone says they are with a defined product, service, or interaction.** A common way to report it is the percentage of positive answers among valid responses to that question. State the scale, the positive-response threshold, and when the survey was sent so readers can interpret the number.

## What does CSAT measure?

A typical question is: **“How satisfied were you with [this interaction]?”** followed by a scale, for example, from 1 (very dissatisfied) to 5 (very satisfied). “How satisfied are you with your purchase?” and “How satisfied were you with today’s support?” measure different things. Name what you are evaluating and when you ask.

CSAT is different from [NPS](/en/guides/net-promoter-score/). CSAT asks about satisfaction with a specific touchpoint or outcome; NPS asks about willingness to recommend. Someone may be satisfied with a purchase but still not recommend the brand, or recommend it while being unhappy with one particular interaction.

## How do you calculate CSAT?

There is more than one way to summarize a scale. In the *top-two-box* percentage method for a 1–5 satisfaction scale, responses of 4 and 5 count as positive. Qualtrics describes this approach in its guide to [CSAT and satisfaction measurement](https://www.qualtrics.com/en-au/articles/customer-experience/what-is-csat/). This is a calculation convention, not a rule for every survey: define the threshold from the response labels and keep it consistent when comparing results.

**CSAT (%) = valid positive responses ÷ valid answers to the question × 100.** Invitees who do not answer are not negative responses and do not belong in that denominator. Report how many people responded out of those invited separately.

### Example: 50 invitations and 36 responses

Imagine a store sends 50 invitations after a purchase. It receives 36 valid answers on a 1–5 scale: 27 people choose 4 or 5, while 9 choose 1, 2, or 3.

| Measure | Calculation | Result |
|---|---|---:|
| CSAT (4–5 on a 1–5 scale) | 27 positive answers ÷ 36 valid answers × 100 | 75% |
| Response rate | 36 responses ÷ 50 invitations × 100 | 72% |

CSAT is **75%**, not 54%: the 14 people who did not answer are not counted as dissatisfied. The response rate was **72%**. That rate does not by itself correct for response bias; consider whether people with unusually good or bad experiences are more likely to reply.

## What should you do with a 75% CSAT?

The percentage summarizes those responses, not the satisfaction of every customer or the cause behind each rating. To turn it into a decision, suppose the survey followed order pickup at a café and several negative responses say the order was not ready at the agreed time. Check the promised and recorded pickup times, review where the delay occurred, and ask a few people whether they received an update. Then decide whether to correct the pickup-time communication or the preparation process. Check the process change as well as surveying again: a later score alone does not prove that the delay has been resolved.

When comparing periods, keep the same question, scale, threshold, population, and survey timing; record any change. There is no universal “good CSAT” benchmark that makes different samples comparable. If you switch from 1–5 to 1–10, or from a positive-response percentage to an average, label the new method instead of merging the figures.

*The invitations, responses, and comments in this example are fictional. The 75% describes only the case’s 36 valid answers.*

The cover is an AI-generated illustration.

## Plan your next post in HeyMark.

Keep the idea, review the draft with your team, and see how it performed in the accounts you connected.

[Start free](https://app.heymark.ai) · [See how it works](https://heymark.ai/en/#product)

## Developer and agent resources

- [HeyMark MCP documentation](https://heymark.ai/en/mcp/)
- [llms.txt](https://heymark.ai/llms.txt)
- [Full site content for language models](https://heymark.ai/llms-full.txt)
- MCP protocol endpoint: `POST https://mcp.heymark.ai`
- OAuth protected-resource metadata: [/.well-known/oauth-protected-resource](https://mcp.heymark.ai/.well-known/oauth-protected-resource)
- MCP server card: [/server-card](https://mcp.heymark.ai/server-card)
