Category Archives: Customer Support

The Customer Service KPIs Every Manager Must Know

“What gets measured, gets managed” –Peter Drucker (father of modern management)

“Ok…but what the hell am I supposed to measure?” -Everyone else


The million-dollar question: “What should I be measuring?” A customer service department has endless possibilities of data points to track and use as Key Performance Indicators (KPIs). Every service ticket is (hopefully) logged in a support-ticketing system like ZenDesk or SalesForce – but that just adds to the confusion. As someone who’s gone through this problem firsthand, I’ve written this post to remove the cloud of confusion around customer service metrics by giving you guiding principles for establishing an effective KPI program and specific KPIs to use.

9 Guiding Principles


  1. KPIs will never replace talking to your customers. You can only fix the problems you know about and not everything is logged as a support ticket. If a customer raises an issue and your data shows otherwise, you can review the data together to find the disconnect. Conversely, you may find that customers think everything is OK while their KPIs suggest otherwise. These discussions are all part of relationship building and shouldn’t be thought of as a “got ya.” Instead, use the conversation as a way to build on your relationship and flesh out any problems with your processes.
  2. You won’t get it right the first time. You will never be “done.” Don’t let perfect get in the way of good. When initially developing and building KPIs, fight the urge for perfection and release a KPI once it’s 80% complete.
  3. Set thresholds for what’s good, average, and not so good. Also known as RAG (red, amber, and green) ratings, these help determine what is working well and what needs attention. Your company’s executive team will also rely on these ratings when quickly assessing the health of the organization.
  4. 100% team alignment. Everyone on your team must know what’s being measured and how it aligns with department goals. This means everyone should be knowledgeable of every KPI and why it matters (yes, this means having a formal discussion with your team). When the “why” is established, KPI adoption will skyrocket.
  5. Full transparency. The KPIs you’re measuring must be fully visible to everyone on the team and the entire company (gasp!). Your first reaction may be, “Why should Sally be able to see how John is performing? Isn’t that a bit much?” Think of sharing KPIs across the team as a form of open communication that removes barriers. Exposing KPIs to the entire company will hold you accountable for your performance and also cut down on the number of “how’s it going?” questions.
  6. Consistently track every KPI across every level of your team. If you are tracking X KPI (e.g. volume of tickets per month) at the department level, make sure you’re also measuring X for each individual contributor. This allows everyone on the customer service team to see how her performance is contributing to overall team objectives. Think of the alternative scenario: measuring X for the department but measuring Y for each individual. This is a problem because everyone will be trying their hardest to improve Y, while having little or no impact on X.
  7. Review metrics regularly. KPIs are only useful if you look at them! Decide how often each KPI needs to be reviewed, then make that information public knowledge. Set a recurring calendar reminder to review the metrics as part of your day-to-day activity—you’ll be able to use the data to improve performance.
  8. Publish results & corrective actions. Your team will want to see how everyone is doing and what has changed based on KPI performance. Consider creating a monthly report that reviews the prior month’s metrics and what actions have been taken. Once you have a template for this analysis, it will take you 15 minutes to compile and internally distribute monthly reports. When things are going well, the report shows the company that your team is kicking ass. When things aren’t going well, it gives you a chance to proactively share what corrective actions are in place.
  9. Real-time data & monthly trending. There are two key groupings of metrics you’ll need to track:
    • Real-time data capture data as it stands now and allow for course correction. These KPIs keep things humming on a day-to-day basis: What does my volume look like right now? How many high-priority tickets are currently open? Does Team A have too much work? Does Samantha have too many tickets in her queue? The real-time KPIs are always a reflection of the current state of the team and should capture data that can be actioned immediately.
    • Historical trending captures and trends performance over several prior periods (e.g. months, quarters): How many tickets do we handle each month? Did we meet our contractual requirements last month? What is the trending of ticket age over the past X months? It’s normal to have KPIs that are important enough to review in real time and as a historical trend.


Specific Customer Service KPIs You Need to Know (and Should Use)


Every customer service KPI boils down to two key categories: Quantity (how many, how often) and Quality (how good).



Of the two key areas, Quantity is the easier concept. Quantity in support is typically measured in the number of tickets worked in a certain unit of time.

Tickets Opened – This is the baseline quantity KPI, which measures how many tickets are opened each month. Set RAG thresholds to determine when you’re in the danger zone. If you have the data, it’s worthwhile to review historical trendsto determine spikes due to seasonality (and plan accordingly).

Number of Tickets Opened KPI


Tickets by Customer (or Customer Segment) – If you have less than 100 customers, you should track by customer. When you reach 100 customers, create segments based on common attributes.

Tickets by Segment KPI


Ticket by Root Cause – It’s important to have a field that captures root cause on every ticket. This will let you dig deeper and ask, “How can we eliminate this type of ticket from ever coming up again?” It’s also a great KPI to review with your VP of Product Management. I also recommend trending this over time.

Tickets by Root Cause KPI


Tickets by Severity – It’s extremely important to capture this data in real time to allow your teams to jump on the highest severity tickets. It’s also valuable to have each employee review them before they leave for the day to ensure high and critical cases are handed off. I also recommend trending this over time.

Tickets By Severity KPI


Tickets by Product Management Category – The purpose of this KPI is to identify product problem areas. I recommend grouping tickets according to the way your Product Management team (PM) is structured. If every PM owns a particular module, track tickets by module. If every PM owns a specific integration, track by integration area. I also recommend trending this over time.

Tickets by PM Category KPI



Quality is more difficult to measure than quantity. Why? Because you have to think about what quality and the definition of success mean to YOUR company. Most high-quality KPIs also require configuration in your ticketing system because they’re not captured out of the box. I’ve found the below high-quality KPIs to have the most bang for their buck:

Age of Resolved Tickets – Measure the time it takes to resolve a ticket after it’s been opened. Track this over time and set red, amber, or green (RAG) ratings to define success.

Age of Resolved Tickets KPI


SLA Performance – Create a KPI for every contractual SLA you have with your customers (e.g. initial response time, resolution time). If you don’t have any contractual SLA’s (lucky you!), create service levels that you want your team to follow and track. I like to track these by calculating the percentage of tickets that meet their SLA, because it levels the playing field in cases where different SLAs exist.

SLA Performance KPI


Support Quality Survey Results – The above KPIs are a good indicator of quality but may not always capture the full sentiments of your customers. Again, talk (and listen) to your customers. Ask them a simple question after every ticket is closed—”How do you rate the quality of support for this ticket?”—and give them choices between 1 (poor) and 5 (excellent). I like this approach for a few reasons:

  • It amplifies the customer voice, which may clarify KPIs like ticket age and SLA performance.
  • Each survey result corresponds to a single support ticket, which gives feedback on an individual performer. This data can also be rolled up at the support-team or customer level for trend analysis.
  • It gives you the chance to follow up with a customer when ratings are subpar. I recommend calling or emailing every customer who gives a 1 or 2 rating. This will give you the opportunity to dig deeper into the root cause of the problem and find ways to improve.

Quality Survey KPI


Closing Thoughts

Don’t worry about getting KPIs perfect the first time around. Pick a few, implement them, see how they’re working, and iterate. Think about how you want to measure quantity and quality. Make sure you communicate what you’re measuring and why to your team. Follow these basic guidelines and you’ll be better off than 90% of other customer service managers.

What KPIs have you found useful in your business?


Leave a Comment

Filed under Customer Support, Thoughts

The Psychology Behind Software Customer Support


photo by gisellecfernandes


Social Rank has a great blog post about customer support and perception vs. reality. The post references the Customer Satisfaction Equation, brought to light by David H. Maister in his brief article on the topic titled “The Psychology of Waiting Lines“. Unlike most academic publications, this eight page article cuts right to the point. As a leader of a software support team, I couldn’t help but make connections between Maister’s psychological insights and the realities of software support. I recommend reading the entire article, but I’ve put together the customer support Spark Notes version below:


It’s hard to plan catch-up ball.

The corollary to this law is the proposition that there is a halo-effect created by the early stages of any service encounter, and that if money, time and attention is to be spent in improving the perceived quality of service, then the largest payback may well occur in these early stages. (page 2)


First impressions matter in customer support. Not only that, a good initial experience with the support team will positively influence all perceptions of your company. Your customer discover a bug? It won’t seem as bad if they’ve had positive experiences with support. I recommend your support team makes the most human connection possible with a new customer. In descending order of value: Face-to-face meeting, video conference, phone call, email. The opposite is also true – botch the customer’s initial perception of support and you’re fighting an uphill battle across the board.


Again, I appeal to common experience to reflect the fact that one’s ‘anxiety’ level is much higher while waiting to be served than it is while being served, even though the latter wait may be longer. There is a fear of ‘being forgotten’. (How many times has the reader gone back to a maitre d’ to check that his or her name is still on the list?). (page 4)


Some support issues or questions take a long time to solve – period. This doesn’t change the fact that the customer still gets very anxious while waiting for a solution. Giving the customer consistent updates at committed intervals will help quell the fear that the problem has slipped into a black abyss. Even if you don’t have a good update to give, there is still immense value in telling the customer “we are still digging into your issue and will report back every 4 hours until it’s resolved”.


On a cold and snowy morning, when I telephone for a taxi, I begin with the expectation that my wait will be longer than on a clear, summer day. Accordingly, I wait with a great deal more patience because I understand the causes for the delay. Similarly, if a doctor’s receptionist informs me that an emergency has taken place, I can wait with greater equanimity that if I do not know what is going on. Airline pilots understand this principle well; on-board announcements are filled with references to tardy baggage handlers, fog over landing strips, safety checks, and air-traffic controllers’ clearance instructions. The explanation given may or may not exculpate the service provider, but is it better than no explanation at all. (bolded by Dan for emphasis) (page 5)


We live in the real world where many factors can impact the level of support. It’s important to recognize these factors and give the customer a heads up on how it will impact service before they complain about it. Proactively telling a customer “Our engineers are currently prepping for this weekend’s release and will be slower to respond to your question” is much better than delivering the same message after the customer asked “why hasn’t my issue been resolved yet?!”


Naturally, justifiable explanations will tend to soothe the waiting customer more than unjustifiable explanations. A subtle illustration of this is provided by the practice of many fast food chains which instruct serving personnel to take their rest breaks out of sight of waiting customers. The sight of what seems to be available serving personnel sitting idle while customers wait, is a source of irritation. (page 6)


How many times have you been in Best Buy or Home Depot to see a crowded store of customers desperately looking for help and a pack of four employees talking in a group and not helping customers. Beyond frustrating.

People need to take breaks to stay sharp, but do it outside of the customer’s glare. On-site at with a customer? Avoid the desire to check your Facebook feed. The customer will build that into their expectations the next time they have a problem to solve. “Greg has time to check Facebook so I expect him to have plenty of time to resolve this issue”.


It follows from this principle that waiting for something of little value can be intolerable. This is amply illustrated by the eagerness with which airline passengers leap to their seats when the airplane reaches the gate, even though they know that it will take time to unload all the passengers ahead of them, and that they may well have to wait for their baggage to arrive at the claim area. The same passenger who sat patiently for some hours during the flight suddenly exhibits an intolerance for an extra minute or two to disembark, and a fury at an extra few minutes for delayed baggage. (page 8)


When a customer asks a easy question that can be answered easily, resist the urge to let it fall to the bottom of the pile. You may say, “…but the same customer has five other open questions that are much higher priority”. That may be true, but the customer perceives the “easy” easy question as low value and their expectations adjust accordingly. To meet that expectation, it must be answered quickly. This will show the customer that you’re providing solutions and buy time for those five other difficult questions.




Leave a Comment

Filed under Customer Support, Thoughts, Tips