Customer service isn’t just about being polite—it’s about being efficient, effective, and impactful—all the time. But how do you know if your team is delivering actually great service or just coasting along?

That’s where various customer service metrics come in. If you pick the right ones, they help you quantify performance, measure success, and identify areas for improvement. 

But let’s be honest—there are tons of metrics you could track, and sorting through them all—let alone picking a few to really focus on—can feel overwhelming.

No worries! We’re here to break it down. Today, we’ll walk you through 10 key customer service metrics, explain what benchmarks to aim for, and show you how to improve them.

1. Total tickets or calls

This metric measures how many support tickets, chats, or calls your team handles within a set time period.

Why does this matter? Because total ticket volume lays the foundation for many other customer service KPIs. It gives you a bird’s-eye view of how busy your team is and can help identify staffing needs, peak hours, and overall workload balance.

How to improve this metric:

• Automate common inquiries using AI-powered chatbots to reduce ticket volume.
• Implement self-service options like FAQs or a knowledge base to handle repetitive queries.
• Analyze ticket trends to find recurring issues and address them proactively.

2. Net Promoter Score (NPS)

Your Net Promoter Score (NPS) is one of the most telling customer satisfaction metrics. It answers the golden question:

"How likely are you to recommend our company to a friend?"

Customers respond on a 0-10 scale, which places them into three categories:

• Promoters (9-10): These are your brand advocates—the customers who love you and will happily spread the word.
• Passives (7-8): They’re satisfied but not enthusiastic. A better offer elsewhere could easily lure them away.
• Detractors (0-6): Not only are they unhappy, but they’re also the ones leaving those scathing one-star reviews (that can really end up hurting your brand reputation).

How to calculate your NPS

Suppose you received 31 survey responses—based on the ratings, 10 respondents were detractors, one was passive, and 20 were promoters. You would divide the number of promoters by the total number of surveys (20/31 = 0.645). Your promoter percentage would be 64.5%.

Now, you can determine your detractor percentage by dividing the number of detractors by the total number of surveys (10/31 = 0.322). Your detractor percentage would be 32.2%.

Finally, subtract the detractor percentage from the promoter percentage to get your final NPS of 32.3 (64.5-32.2 = 32.3).

Determining the definition of a good NPS depends on the industry and many other variables. However, according to global NPS standards, a score over 50 is good, and over 70 is excellent.

How to improve your NPS:

• Follow up with detractors to understand their pain points.
• Empower your team with better training and clearer customer communication.
• Track trends—if your NPS drops after a major change (like a website redesign), investigate ASAP.

3. Customer Effort Score (CES)

How easy is it for customers to do business with you? Customer Effort Score (CES) measures just that.

A high CES means customers had to jump through hoops to get help. A low CES means interactions were smooth and effortless—just the way they should be.

You can figure out your CES by surveying your customers.

However, survey timing matters. Here’s when to do it:

• After an interaction resulted in a purchase
• After an interaction with a customer service associate
• While performing user-experience (UX) and user interface (UI) testing

Likert scale vs. emoticons

There are two common CES surveys. In the Likert scale, the survey asks customers to rate the simplicity of their experience on a 5-7-point scale, with one being the easiest and the highest number being the most difficult.

There’s also a simpler emoticon rating system. Instead of choosing a number, customers choose a frowning face, an expressionless face, or a smiling face to represent how easy their experience was. 

How to calculate your CES

Divide the sum of the effort rating scores by the total number of survey responses. So, if you performed 100 surveys and the scores on your scales totaled 750, your CES is a 7.5. 

If you’re using the emoticon responses, choose a number to correlate with each emoticon and use those numbers for scoring. For example, a frowning face is a 1, an expressionless face is a 5, and a smiling face is a 10.

There are no standardized good or bad CES results. Set a benchmark based on averages within your industry and your company’s average CES over time, and compare your customer service team’s results to that.

How to improve your CES

• Simplify your contact options—nobody should dig through five pages to find a phone number.
• Live chat support = faster resolutions.
• Reduce the number of menu options before customers reach a human agent.

4. Customer Satisfaction Score (CSAT)

CSAT asks customers a straightforward question:

"How satisfied were you with your experience?"

They respond using a 1-6 scale (where 1-3 is negative, 4 is neutral, and 5-6 is positive).

A good CSAT score? Typically 75-80%. But more importantly, it should trend upward over time.

How to calculate your CSAT score

Divide the total number of positive responses by the total number of survey responses and multiply by 100 to get a percentage. 

So, if you received 75 positive responses out of 100 total responses, you’d have a CSAT score of 75%.

CSAT score benchmarks vary by industry, but the standard mark for a good CSAT score is 75% to 80%. 

How to improve your CSAT score:

• Listen to your customers—analyze feedback and implement changes.
• Empower your team to go above and beyond for customers.
• Make self-service more accessible (e.g., knowledge base, FAQs).

(The difference between NPS, CES, and CSAT)

The three metrics described above are all similar in that they each ask customers to rate their experiences with your brand, product, or people. How are they alike and different? 

• NPS: This metric is purely relational. It shows a big-picture view of how your customers feel about your brand of product.
• CES: This metric is purely transactional. It asks customers to rate their specific experience with your product or team.
• CSAT: This metric is a little of both, depending on when and how you ask for the rating, though it is usually transactional.

The bottom line is that NPS offers a wide view that asks customers to consider all of their interactions and experiences with your company, while CES and CSAT are more narrow, asking customers to consider a specific interaction or experience.

It’s important to selectively use each of these survey-based metrics where they make the most sense to give you the relevant data you need to measure your progress toward your goals. 

Also, make sure you don’t bombard your customers with multiple surveys every time they talk to you!

5. First Response Time (FRT)

This metric tracks how long it takes for a customer to get an initial response after submitting a support request.

How to calculate FRT

You can calculate this useful key performance indicator (KPI) by dividing the total first response times in a given period by the total number of tickets in that period.

For example, if your total FRT for today was 60 minutes and you had 10 total tickets, you’d have a six-minute average FRT that day.

Ideal benchmarks:

• Social media: Under 60 minutes
• Email: Less than 24 hours
• Phone: Less than 3 minutes
• Live chat: Under 2 minutes

How to improve your FRT:

• AI-powered chat assistants can handle simple queries instantly.
• Prioritize urgent tickets with intelligent routing.
• Avoid multitasking overload—focus on one inquiry at a time.

6. First Contact Resolution Rate (FCR)

FCR measures how often support tickets get solved on the first attempt—without requiring a follow-up. 

A great FCR score shows your associates are good at understanding and resolving customer issues. Reducing the number of issues that require more than one call or email to resolve reduces frustrating interactions, which can increase trust in your organization and lead to repeat business and referrals. 

It also lowers the resources needed to resolve each issue, which helps you manage your overall costs.

How to calculate your FCR

To calculate FCR, divide the total number of tickets solved on the first agent interaction during a select period by the number of total tickets in that period. Then, multiply the result by 100 to get the FCR percentage. 

For example, if you had 100 tickets in a week and 75 of them were resolved on first contact, you’d have a 75% FCR rate ([100 / 75] x 100 = 75).

While many variables impact FCR rates, the industry benchmark is 75%. 

However, be careful: a great FCR can also potentially indicate that your associates are spending too much time on easy-to-resolve tickets that could be eliminated with better self-serve options or the power of AI.

How to improve your FCR:

• Train agents in active listening to get to the root of issues faster.
• Use AI-driven support tools to provide real-time suggestions.
• Optimize ticket routing so complex queries reach the right expert immediately.

7. Cost Per Ticket (CPT)

Want to know if your support operation is cost-effective? CPT tells you how much you spend per ticket.

How to calculate your CPT

To calculate CPT, divide your customer support team’s total operating expenses during a period by the number of tickets during the same period. 

For example, if your team costs $50,000 per month to operate and handle 5,000 tickets, they would have a $10 CPT.

How to improve your CPT

• Automation + self-service = fewer human-handled tickets.
• Improve agent training to reduce handling time.
• Lower agent turnover rates to cut down hiring & training costs.

8. Average Resolution Time (ART)

Average resolution time (ART) is the average time it takes a customer support associate to resolve customer issues. 

The faster your team can resolve customer issues, the more satisfied the customer is, the more likely they are to return and refer others.

Plus, the faster your associates can resolve issues, the fewer resources you spend per issue.

How to calculate your ART

You can calculate this customer support metric by dividing the total time needed to resolve all tickets in a specific period by the number of tickets during that period.

For example, if you had 100 tickets in a week, and it took 600 minutes to resolve all customer problems during that period, your average resolution time was six minutes.

The benchmark ART across all industries for phone support is 6.46 to 6.73 minutes, but the exact time varies by industry and company size. 

How to improve your ART

• Train agents to resolve issues without escalating unnecessarily.
• Use AI-powered chatbots or assistants to pre-screen customer inquiries.
• Provide an internal knowledge base to help agents work faster.

9. Churn rate

Churn rate—when and how often customers stop doing business with a company—is one of the most important customer service metrics. 

A low customer churn rate shows your customers are generally satisfied, but a high churn rate shows an issue along the customer journey, causing them to cancel or leave. 

However, it will not show why they are leaving, so you’ll need to explore other customer service metrics to find where the issue is.

How to calculate your churn rate

The formula for calculating churn rate percentage is the number of customers lost during a time frame divided by the total number of customers you had at the start of the time frame, multiplied by 100. 

So, if you had 100 customers at the beginning of the month and 10 cancel, you had a 10% churn rate ([10 / 100 = 0.10 x 100 = 10%).

A good churn rate varies by industry and other variables, but the accepted rate across all industries is 2% to 8%.

How to reduce churn:

• Offer loyalty incentives to keep customers around.
• Improve customer experience based on complaint trends.
• Proactively engage with at-risk customers before they leave.

10. Reopen Rate (RR)

Reopen rate (RR) is the percentage of closed or solved tickets reopened by a customer. This means the customer support agent thought they had resolved an issue, but the customer continued having problems.

This can lead to additional frustration for your customer and reduced satisfaction, which can negatively affect repeat business, churn rate, and referrals. 

Also, a high RR shows your customer support team is having issues properly identifying the root of customers’ issues and may need additional training.

How to calculate your RR

The RR formula is the number of reopened tickets in a period divided by the total number of tickets solved during the same period. Then, you multiply that result by 100 to get the RR percentage. 

So, if your team had five reopened tickets and 100 solved tickets in a week, your weekly RR is 5% (5 / 100 = 0.05 x 100 = 5%).

A 10-20% reopen rate is acceptable. Anything higher suggests agents aren’t resolving issues properly.

How to improve your RR:

• Train agents to confirm issue resolution before closing tickets.
• Ensure responses are concise, accurate, and actionable.
• Provide clear troubleshooting steps in initial responses.

Pick the right customer service metrics & elevate your support team and CX

Tracking customer service KPIs is about more than just crunching numbers—it’s about improving the customer experience.

By constantly monitoring and optimizing these key customer support metrics, you’ll:

• Deliver faster resolutions
• Boost customer satisfaction
• Build long-term loyalty

The result? Happier customers, stronger retention, and a better bottom line.

Want more customer service insights, tips, and tricks? Check out the PartnerHero blog.

Need a little help with your customer support or some assistance from Augmented AI? Get in touch, and let’s have a chat—no strings attached!