Most support platforms provide all the essential reports to guide your team in the right direction. But they don’t necessarily tell you why or how.
Whether you’re new to customer service or a seasoned veteran, it’s always worth getting a fresh perspective on your own stats, data, and metrics.
Why? Because even though support is a high-touch discipline that thrives on human connections, elevating good customer service to exceptional only takes place when you know the cold, hard numbers.
That’s why this article exists: To break down the KPI (key performance indicator) metrics as well as to help you understand exactly what it’s measuring and why it matters.
Below, you’ll get a glimpse of each one in the wild through our reporting dashboard. And then, tangible ways to improve them. With a data-meets-action approach at the forefront of your mind…
Here are the 10 customer service KPI metrics that matter.
1. Ticket volume
Ticket volume measures the total number of conversations in your support inbox.
Start tracking this metric to get a general understanding of how many of your customers need help. Track trends over time to see how product or support changes increase or decrease this number.
Let’s walk through an example report from Groove’s dashboard on customer conversations.
Using this chart, we immediately see two big spikes. Correlate those dates with product issues or changes to understand what happened. Brainstorm ways to get ahead of issues to keep volume more steady in the future.
Offer self-service as your frontlines
Improve this metric with a knowledge base. Number of conversations is directly correlated to the visibility and clarity of your knowledge base.
When customers can find answers on their own, they don’t need to reach out to customer support. Optimize it as much as possible, and track ticket volume to see what’s working.
Oh, I should mention that if you’d save all 10 metrics for later, we’ve compiled them—along with the tactical takeaways—into a single PDF:
2. Ticket backlog
Ticket backlog refers to customer support requests left unresolved over a particular time frame.
These tickets remain unresolved due to the performance of your customer service team, abnormally high ticket volume, or business/product related dependencies that require additional time to solve.
Note: Resolution and reply mean very different things in customer service.
- Replying to a customer simply implies you’ve responded to their inquiry.
- Resolving an issue indicates their problem is solved and they (or you) need no further correspondence.
Looking at the Reports dashboard again, we see bars for conversations, customers, and resolutions.
The goal is to get the number of resolutions at or above the number of conversations. This means you’re resolving every inquiry (if you’re above, it means you’re resolving conversations from the prior day or week).
Set up reminders for lingering tickets
Stay on top of unresolved requests by snoozing a conversation for the exact amount of time you need.
When time is up, check on the customer or follow up with your product team. Customer support cannot resolve everyone’s problems all the time (shocking, I know). We can keep open lines of communication, though, and facilitate resolutions as seamlessly as possible.
3. Average resolution time
Average resolution time shows the median amount of time it takes to completely resolve a customer’s issue.
Efficient support teams work to impact resolution time, rather than reply time. A short resolution time typically indicates effective support responses. Customers get everything they need to know, and then some.
Resolution time holds steady in this chart, and even seems to be lower compared to last week. But that spike requires investigation.
Maybe a team member was out of office. Or a bug popped up without a quick fix. Correlate resolution time with other happenings in the company to get the full picture.
Prevent follow-ups with related articles
Link to knowledge base articles within replies to give customers even more information and prevent further questions.
Think of it like offering “related articles” at the end of a blog post. Predict what else may be confusing your customers and address it before shooting out your reply to achieve a resolution.
4. Average reply time
Average reply time displays the median amount of time it takes to respond to a customer.
Customers have very high expectations for reply time. With live chat and auto-replies producing an immediate response, customer anticipate a follow up as soon as they send a support ticket.
The dashboard shows median response time at just over three hours. Plus, we can see the average number went down by over 30% since the previous window of time.
The decrease is definitely a good sign. We always want to be shortening the time between replies. The actual number (three hours) tells us far less. Three hours might be perfectly within the time frame your customers expect, or it might not. Set baselines and benchmarks for your customer service team to interpret this number.
Track ticket volume for proper staffing
View the busiest times for incoming messages across the week.
Staff up for the busy times to reduce reply time. With more hands on deck, customers will get the attention they need as soon as possible.
5. Average first response time
Average first response time tells you how long it takes for a customer to receive an initial reply to their support request.
First impressions mean everything. Good customer service teams excel at responding to customers as quickly as possible. Even if you can’t resolve an issue, send a quick note that you received their message to alleviate the anxiety of an unknown wait time.
We track median first response time using the graph below.
If our goal is to stay under eight hours, then we’re mostly on track. Again, it depends on the expectations you set for your company and your team. Dive into the spikes to see what’s causing them and to work on preventing it in the future.
Automate and templatize your first response
Automate an initial response to reduce this data point to nearly non existent.
Let customers know you received their inquiry and will be in touch shortly. Include some common questions or a link to your knowledge base to try to solve the problem immediately. When there’s a known issue, update this reply with as many details as possible to get ahead of customer inquiries.
6. Customer satisfaction score
CSAT (customer satisfaction score) reveals whether a customer experience was good or bad.
A customer satisfaction survey is typically presented after ticket resolution to gather customer sentiment. It asks the customer to rate their support experience on a scale ranging from good to bad.
In the dashboard, we track both CSAT score and the percent of customers who participated in the survey.
An 80% CSAT is considered excellent in the customer service world. If we’re using general benchmarks, we can be very satisfied with this number. If we’re comparing against the last month, though, the report shows that we’re 13% lower. We’ll want to research why it went down to prevent churn.
Get more survey participants
The second piece of data in the CSAT score report gives us a great starting point for improvement. If you can get more people to complete the survey, you might be able to raise CSAT without putting in too much more effort. More often than not, angry customers are the most vocal. Encourage happy customers to complete the survey to get a more balanced metric.
7. Average handle time
Average handle time calculates how long it takes an agent to respond to a single inquiry.
This measures from the time an agent opens a support email to the time they click send (or from the moment they answer a phone call until the moment they hang up). Customer service lives in the details. Shaving a few seconds off handle time adds up to big savings. Managers should constantly be looking to improve the average time by optimizing and streamlining processes for their team members.
We’re looking at average handle time for one specific agent in this graph.
The goal is to reduce this number overtime, but the actual number to reach for will vary depending on your organization. Managers should be watching this number more than agents. If the number rises, try out a new process or shadow your team member to see what’s slowing them down.
Provide an internal self-service resource
Turn your knowledge base into an internal hub for support reps. Especially when you’re knowledge base is connected to your inbox, this will be the easiest and quickest place for your team to find answers.
Your customers love self-service because it helps them resolve their own issues quicker. Take this same philosophy into the workplace. Arm your support reps with a hefty internal knowledge base to answer any questions they may have during their time in the inbox.
8. First contact resolution rate
First contact resolution rate identifies the percent of total tickets wherein you fully resolve an inquiry within a single response.
In theory, customer service questions can be solved with a single answer. In practice, this rarely occurs. Support agents often need more information to fully understand the question, prompting a second email. Customers have follow up questions. But this metric should always be top of mind. Working to improve first contact resolution will improve both customer happiness and team productivity.
Groove Reporting shows first reply resolved as a percentage of total conversations. We can also see the percent change from the previous month.
We resolve over 50% of issues on the first reply. This cuts the number of tickets needing multiple responses in half. Agents can focus more of their time on challenging tickets.
Insights from this metric are two-fold. You want to keep it high to make sure your responses are clear and effective right out the gate. However, if it gets too high, consider optimizing (or creating) your knowledge base. The inbox should be the last line of defense, not the place to go for easy answers.
Create a database of proven replies
Canned replies rule this metric. Build a database of replies for common questions and shoot them out as needed.
Test what customers respond best to. Ideally, the majority of your canned replies will end in a ticket resolution. These should be tried and tested, and nearly guaranteed to work.
9. Net Promoter Score (NPS)
NPS (Net Promoter Score) measures how likely your customers are to recommend your product or service.
Send an NPS survey to solicit responses and generate the number associated with their recommendation.
After gathering all the data points, break it out into promoters, passives, and detractors.
- Promoters: Customers who selected 9 or 10 on their NPS survey.
- Passives: Customers who selected 7 or 8 on their NPS survey.
- Detractors: Customers who selected 0-6 on their NPS survey.
Subtract the percent of detractors from the percent of promoters for a given time period to reveal NPS.
Borrowing from our in depth article on analytics, we’ll briefly walk through an example of how to interpret NPS in the real world.
Let’s say surveys went out to 30,000 customers, and 3000 responded. Filter the responses by promoters, passives, and detractors to get the percent for each. Then plug it into the NPS formula.
Your NPS score may vary wildly. Again, set a baseline and work from there. Interpret the results using benchmarks.
Correlate all other metrics with NPS
Essentially, everything your customer support team and your organization-at-large does can affect NPS. There’s no direct line to shave time or boost numbers on this metric. But, if you put the work into creating a streamlined and personal support system using the metrics above, it will lead to increased NPS and heightened customer loyalty.
10. Replies per resolution
Replies per resolution identifies how many times an agent and customer go back and forth before coming to a resolution.
We track median number of replies per resolution in the Reporting dashboard using a line graph. We can easily see baselines and observe any abnormalities over time.
Something happened on August 18 to really skew this metric. Correlate it with product issues or changes to understand why it jumped. Think about what could have been done to keep replies more encompassing.
Use notes to communicate before responding
Notes ensure that agents get the full picture before replying to a customer. They can correspond internally with another team member or manager if they need help with a ticket, rather than reaching out to a customer unprepared.
Team members can also use notes as reminders or info cards. Attach certain data or reference materials to a complex conversation so everyone has context when replying. Notes prevent agents from asking the same question twice and from replying without all the needed information.
Take action on your customer service metrics
Metrics alone won’t produce satisfied customers. It’s on you to take these data points and build a customer service experience that works for your organization.
Set KPIs (key performance indicators) based on each customer support metric to guide your support staff in the right direction. Lean on your help desk to track progress. And use all the tools on hand to fulfill customer requests with minimal effort.