How to Calculate CSAT & What It Means for Your Business

How to Calculate CSAT & What It Means for Your Business

Your customers and the customer experience (CX) should always be at the center of everything you do – it’s Business 101. You need to ensure that your customers are content or, even better, enthusiastic about your brand, and that your products and services are meeting their needs.

There are a number of metrics to measure customer happiness, like Net Promoter Score (NPS) and Customer Effort Score (CES), among others. Most of these calculations produce great results, focusing on slightly different aspects of the customer journey. However, one of the most used and most trusted ways to evaluate your customer support efficacy and overall customer experience is by determining your customer satisfaction (CSAT) score.

What Is a CSAT Score?

A customer satisfaction (CSAT) score is a measurement of overall customer satisfaction with a brand or its products and services. It’s a form of survey analysis, usually based on a single, scale-rated survey question like, “How would you rate your experience today?”

CSAT surveys ask close-ended questions, with a predetermined set of possible responses: Yes/No, on a scale of 1 to 10, 1 to 5 stars, or “enthusiasm rated,” e.g., agree strongly, agree somewhat, neither agree nor disagree, etc. Because CSAT scores deal with close-ended, quantitative, or structured data, each response has a corresponding number value, so they’re easy to administer and easy to calculate.

CSAT surveys are often asked via email or in-chat, immediately after a customer interacts with customer service, or at certain customer journey touchpoints, like after purchase or after onboarding. They can pop up on a company’s website, on in-store touchscreens, in-app, in chatbots, and more.

CSAT survey questions and possible responses can vary. They, essentially, just need to serve the purpose of finding out if customers are satisfied with what they have purchased or the service they have received, or not. A few examples of CSAT survey questions:

  • Did we help you achieve everything you needed today?
  • On a scale of 1 to 10, how would you rate [our company, product, customer service, etc.]?
  • Does our product/service help you reach your goals?

From “extremely satisfied” to “not at all satisfied” how would you rate the following services? (With different company services listed and choices to rate with a click below.)

CSAT scores are a fairly basic calculation of customer feedback and customer loyalty, but they can be a powerful metric for understanding general customer happiness and measuring your business against your competition.

And when it comes to a good customer experience, good customer service is key. Customer support analysis is the main focus of CSAT – it’s a major pain point for many customers, but companies with great customer service tend to outperform their competitors.

According to Zendesk’s global CSAT index report, 82% of Americans said they’ve stopped doing business with a company due to poor customer service. So, just a single bad experience can really hurt your bottom line.

Zendesk’s CSAT index, answering the question, “How would you rate the service you received?” shows global customer satisfaction at 86%. Australia and Canada lead the the English-speaking world at 93% each, with the US scoring 87% and the UK coming in last with 83%.

The global CSAT index by company size (number of employees) shows us some interesting data:

Company sizeCSAT Score
1 - 991%
10 - 9984%
100 - 49988%
500 - 4,99993%

Small, mom-and-pop businesses have the time to give good customer service because they tend to deal with customers face-to-face and don’t have customer support tickets, per se. Medium-to-large companies (10 to 500 employees) tend to struggle with customer service, possibly because they have grown too fast, or haven’t sufficiently focused on how it affects their customers.

For your CSAT score related to customer service, it all boils down to (1) how many support tickets come through your support desk, (2) first response time, and (3) the percentage of tickets resolved. Your company’s CSAT calculation can help understand where you’re failing your customers. And later, we’ll show you what machine learning can do to transform your customer service feedback into action.

A picture of an equation for a CSAT Score.

If you asked the question: “Did we help you with everything you needed today?” after a customer service call to 200 customers, with 140 saying “Yes” and 60 saying “No” your CSAT score is 70%:

This is for binary or dichotomous questions: Yes/No, True/False, thumbs up/thumbs down, etc., when customers are only expressing approval or disapproval.

If you are calculating on a number scale, the calculation is slightly different. You add up all of the response scores given, divide by the maximum scores possible and multiply by 100.

A picture of an equation for a CSAT Score

Let’s say you asked “On a scale of 1 to 10, how would you rate our customer service today?” to five customers and received these results:

CustomerResponse Score 1 to 10Maximum Possible Response
Response 1910
Response 2710
Response 3510
Response 41010
Response 5410

If you’re working with ratings from very satisfied to not at all satisfied or similar, you would simply give each rating a number value and perform the same calculation.

Although not as tied to customer service as most CSAT scores, star rating systems, like you see on Yelp and Amazon work similarly. You simply multiply by 5 instead of 100 because the highest possible score is 5 stars, rather than 100%.

A picture of a star rate CSAT Score Scale

Pros & Cons of CSAT

The main benefits of CSAT are that it’s fast and easy to calculate, and because CSAT surveys generally only ask a single question, you’re likely to get high response rates from your customers. Simply add your CSAT question to your email signature or have it pop up in your app or on your website after a customer service interaction.

CSAT benchmarks, by industry, are readily available, so you can easily see how your business stacks up. And easy data collection and quick results mean you can follow your CSAT as it rises or falls, day-by-day.

Unfortunately, as you’re not necessarily getting responses from every customer, your results could end up skewed. Customers that had a terrible experience, for example, may simply want nothing more to do with the company and will quickly log off or avoid emails.

Furthermore, CSAT scores only give you quantitative information that measures customer satisfaction with numbers and percentages. Quantitative data tells you what has happened, i.e., the percentage of your customers that are satisfied with your product or services.

Qualitative data, on the other hand, can dig into the why behind these statistics. Qualitative survey data analysis deals with open-ended questions that allow the respondent to answer the question in their own words, giving you the data of ideas, feelings, and opinions, and often leading to new information you had never considered.

Because qualitative data is free-form text, it’s harder to analyze – you can’t simply allocate corresponding whole numbers and plug-in formulas. With advances in machine learning, however, AI text analysis software can structure your unstructured survey text data and automatically mine it for keywords, opinion polarity, categories and themes, and more, for real-time results and qualitative data-driven decisions.

MonkeyLearn is a SaaS text analysis platform with a suite of tools to automatically analyze survey data – thousand to hundreds of thousands of open-ended responses – for huge insights.

When performing a CSAT survey for example, you can ask, “On a scale of 1 to 10 how was your customer support experience?”

Then, you can follow that question up with an open-ended question: “What can we do to improve our customer support?” or “How did you experience problems today?”

This gives you both qualitative and quantitative data, so you can find out what happened (good or bad experience) and why it happened.

Sentiment analysis of survey responses, for example, automatically analyzes surveys for “opinion polarity” – showing the feeling or emotion of the respondent. Try out this pre-trained sentiment analyzer to see how it works:

Test with your own text



From there you can sort your text by topic or category. Like in this survey feedback analyzer that classifies open-ended survey responses by aspect: Pricing, Customer Support, Ease of Use, and Features:

Test with your own text


Customer Support88.9%

This, ultimately, gives you aspect-based sentiment analysis, so you can automatically analyze your survey responses to find out which aspects of your business are Positive and which are Negative. In this example, Customer Support is Negative.

With MonkeyLearn, you can custom-train these tools, and more, to the needs and criteria of your business and your industry, for even more accurate results. Manually reading and annotating open-ended text data is a thing of the past. Machine learning allows you to supplement your quantitative CSAT calculations with quantitative, open-ended survey responses, for deeper data and more fine-grained results.

Enhance Your CSAT Score

CSAT surveys are a standard measurement across most modern industries for customer satisfaction and customer service efficacy. They offer solid results and quick and easy measurements, but their quantitative data will only get you so far.

Enhance your CSAT score with open-ended questions, for results that could lead to wholly new insights, backed by real opinions. MonkeyLearn’s text analysis tools work constantly and in real time, so you’ll never leave your customers in the cold.

Take a look at MonkeyLearn’s pricing page to see what we have to offer or schedule a demo to try the tools before you buy and learn about all of our text analysis and data visualization tools.

Rachel Wolff

January 21st, 2021

Posts you might like...

MonkeyLearn Logo

Text Analysis with Machine Learning

Turn tweets, emails, documents, webpages and more into actionable data. Automate business processes and save hours of manual data processing.

Try MonkeyLearn
Clearbit LogoSegment LogoPubnub LogoProtagonist Logo