Customer Satisfaction Score (CSAT): Formula, Survey and Benchmarks

The Customer Satisfaction Score (CSAT) tells you how happy customers are with your product, service or experience. You ask one question, customers rate their satisfaction (usually from 1 to 5), and CSAT is the percentage who answer 4 or 5.

Running a business is more than Key Performance Indicators, sales figures, tax returns or product suites. When stripped down to the core, all companies should revolve around one goal: satisfying the customer. Only when you provide undeniable value will you earn loyal customers who stick by your side in difficult times. That is why customer experience teams deserve the same consideration as sales or marketing teams. In their hands lie the positive experiences customers are craving.

Fortunately, we have many tools and metrics to track the performance of a great digital customer experience. This guide covers what CSAT means, how to calculate it correctly, how it differs from other CX metrics, and how it contributes to improved Customer Lifetime Value.

What is CSAT (customer satisfaction score)?

You can use CSAT to understand the reasons behind customers' happiness (or unhappiness) and then take the necessary steps to improve critical areas inside your business. Because you can ask the question at any touchpoint, CSAT works for a whole relationship ("How satisfied are you with our store?") or for one moment ("How satisfied are you with your delivery?").

CSAT is one of several satisfaction metrics. The Customer Satisfaction Index (CSI) is a different one: it combines ratings on several attributes, weighted by importance, into a broader score, while CSAT is the share of satisfied answers to a single question.

How do you calculate CSAT?

As with other customer feedback metrics, you measure CSAT with customer satisfaction surveys deployed at several stages of the customer lifecycle. A typical CSAT survey consists of one close-ended question:

To take full advantage of customer feedback, add an open-ended question that asks customers to give the reason behind their rating. You can also add CSAT questions that narrow in on specific parts of the transaction, such as the shopping experience, product quality, the UX of your online store or the shipping process.

After the responses come in, use the CSAT formula:

Worked example

You send a post-purchase survey and receive 200 responses:

• 5 (very satisfied): 90 responses
• 4 (satisfied): 60 responses
• 3 (neutral): 25 responses
• 2 (unsatisfied): 15 responses
• 1 (very unsatisfied): 10 responses

Satisfied responses = 90 + 60 = 150. CSAT = (150 ÷ 200) × 100 = 75%.

Neutral answers count in the total but not as satisfied, so they pull the score down. Some tools report the average rating instead (the sum of all scores divided by the maximum possible score, × 100). In the example above, that gives 745 ÷ 1,000 × 100 = 74.5%. It is a different number, so always state which method you use and keep it the same over time.

As you can see, CSAT is calculated as a percentage. Still, you can also see the satisfaction of each individual customer. As you would do with NPS, if a high-value customer scores lower, your customer service team can quickly step in and do some damage control. This way, you turn your CX department into a profit center rather than a cost center, and you prevent customer churn instead of merely reacting to it.

The happy/unhappy alternative

A second approach is less nuanced, but it still gives a general idea of customer happiness and can even bypass cultural differences in how people use rating scales. Instead of a scale, customers pick one of three options:

• Happy
• Neutral
• Unhappy

Here, CSAT is the share of "Happy" answers. You can also ask customers to explain their choice in an open-ended question. However, it is easier and more common to work with numbers than with visual elements, so the happy/unhappy version is used less often than the classic rating question.

What is the difference between CSAT, CES and NPS?

With so many customer experience metrics, a comparison is due. It shows what each metric represents and when to use each one to improve customer loyalty. For a deeper side-by-side, read NPS vs. CSAT vs. CES.

CSAT measures customer satisfaction alone. It doesn't tell you anything about customer loyalty or the effectiveness of your support center. It can also be tricky or misleading: respondents usually score in the extremes, so customers who are satisfied but sit somewhere in the middle of the scale risk falling through the cracks. Use CSAT to evaluate how happy you make your customers, but don't stop there.

The Customer Effort Score measures how much effort a customer puts in when interacting with your organization. It is generally used to evaluate customer support, as it shows how complicated it is for a customer to have an issue solved by your service reps. Customers don't always have the patience to go through an intricate call support process or wait more than a few minutes until a representative picks up the phone.

The Net Promoter Score measures the likelihood of customer recommendations. Like CSAT, NPS is measured through a survey with a close-ended question. The similarities end there, since NPS revolves around word of mouth, and there's no guarantee that a satisfied customer will tell their friends about you. Our Net Promoter Score guide covers it in full.

Should one be more important than the other? We advise using the metrics like a storyboard. Let them tell the story of why your customers love you, how smooth the relationship is, and how likely they are to bring their friends to the party. Customers' expectations aren't easy to meet: your products must bring real progress, the price point should be affordable, and their issues should be solved fast. The metrics are your mirror, showing how good a job you're doing and highlighting improvement opportunities you might have overlooked.

Why do CSAT scores matter?

Your long-term goal is to monitor and increase Customer Lifetime Value (CLV) and get more profitable relationships from customers you have already acquired. Service quality is the only way to reach a satisfactory CLV. Over-delivering on your promises should be the new normal, and CSAT is the metric that tracks this goal.

Here is how CSAT works for you:

• It shows what works and what must change. Depending on how you design the survey, you get insights into the parts of the customer experience that work well and the ones that need attention.
• It shows how customers perceive a change. Measure CSAT before and after a new delivery partner, checkout or returns policy. Because the answers are nuanced, you can detect even small shifts in satisfaction.
• It works as a bug detector. Negative responses and open-ended answers uncover friction points, issues and problems within your organization, so you can prioritize what to fix.
• It measures team performance. CSAT after support conversations is a good gauge of your teams' effectiveness.
• It is easy to run. CSAT surveys are easy to implement and send, and the questions don't take up much of your customers' time.

One of the best skills a professional can learn is asking the right questions, such as "Why are customers leaving us?" CSAT helps answer it, showing where and why you aren't meeting customer expectations.

What is a good CSAT score?

To evaluate your CSAT score:

• compare it with similar companies' scores,
• use external industry benchmarks,
• or look at global customer experience trends.

When you look for benchmarks inside your industry, compare with companies that share a similar size, product suite and customer base. Besides industry benchmarks, compare your scores with your team's historical performance or with similar teams in different locations. Looking at CSAT from a historical point of view is the most reliable way to track your progress.

How do you create a CSAT survey?

CSAT is only valid when customers actually want to respond to your survey. With this in mind, here are the key steps for creating a successful one:

1. Be intentional about the information you need

   Pick your objectives from the start and be sure of the information you want to collect. Talk with your teams first: anyone who interacts directly with customers can point to expectations, red flags or areas of dissatisfaction. For example, if customers from one location churn quickly and buy only once or twice, there might be a shipping issue. Design questions around that assumption, or ask customers to rate the delivery process. Keep a core set of questions and swap in others for what you want to uncover.
2. Don't be afraid to go into specifics, but put the key questions first

   Start with overall satisfaction, then continue on different paths to discover the details of the experience. By the end of a survey, customers are usually at the end of their patience and want to get it over with, so their ratings might not be as sincere. Treat the answers at the beginning of the survey as the most valuable ones.
3. Use clear language and keep it short

   Use simple, easy-to-understand words. People might multitask while they answer, so you don't have their full attention. The survey should take at most a couple of minutes. The longer it is, the more people quit before they finish, and the less relevant the answers you get.
4. Don't overdo it with open-ended questions

   Respondents tend to say everything on their minds in the first open-ended question, so extra ones lead to blank fields or repetitive answers. Include only the questions that matter at the time. If you want to gauge sentiment about a new product, leave out questions about delivery or site speed.
5. Use different groups for your survey distribution

   If you survey the same group of customers every quarter, you risk tired and even annoyed customers. The insights will also be one-sided and won't give you the holistic view of your organization that you need.

CSAT survey question examples

Here are CSAT questions you can adapt for different moments in the customer journey. Pair each rating question with an optional "What is the main reason for your score?"

• Overall: How would you rate your overall satisfaction with [brand]?
• Shopping experience: How satisfied are you with how easy it was to find what you were looking for?
• Checkout: How satisfied are you with the checkout process?
• Delivery: How satisfied are you with the delivery of your order?
• Product: How satisfied are you with the quality of [product]?
• Support: How satisfied are you with the help you received today?

You can run these as on-site and post-purchase surveys with Omniconvert Explore, then test fixes for what customers complain about on the same platform.

How do you analyze CSAT feedback?

To analyze your CSAT surveys effectively, go through these two steps.

Step 1: Split your respondents into satisfaction groups

Segment respondents into sub-groups based on their answers, and use a different approach for each. First, check where you stand in numbers: do you have more satisfied customers or unsatisfied ones? If unsatisfied customers outweigh the satisfied ones, act as soon as possible to tip the balance.

• Unsatisfied customers: the main goal is preventing churn. Solve their issues or the aspects that disappointed them, and you might turn them into satisfied customers. After all, there's a thin line between love and hate.
• Neutral customers: ask an open-ended question to see what you could have done differently to deliver a better experience.
• Satisfied customers: follow up and identify exactly what they loved about your brand. That's your strong point, your USP against the competition. Satisfied customers can be turned into promoters, even if you use a CSAT survey rather than an NPS one.

Satisfaction groups become much more useful when you cross them with customer value. An unhappy customer who orders often and spends a lot deserves a faster, more personal response than a one-time discount buyer. RFM segmentation shows you who is who, and reducing churn starts with that high-value, at-risk group.

Step 2: Use text analysis to make sense of your data

Text analysis helps you work through open-ended responses without burning resources or staff time. Use text analysis software to extract the recurring themes in your customers' answers, then act on those insights. Customer sentiment analysis works the same way: classify each comment as positive, negative or neutral, and follow the trend.

Let's also shine a light on your customer support teams. CX representatives are usually among the most strained people inside a company, and support departments often face high turnover and more stress than needed from dealing with customers daily. Text analytics can help them automate tasks like sorting CSAT answers, and improve their problem-solving while they keep listening to the voice of the customer.

FAQ: CSAT