What words come to mind when we think of great customer support? Empathy, patience, and positivity perhaps. But what about metrics? While I agree that support is ultimately about people, metrics are how we measure our performance and inform improvements. To boost your customer support strategy, take a left and right-brain approach by tracking these essential support metrics.

You’re probably familiar with the first four. Consider them the universal, time-tested metrics to measure the effectiveness of your customer support system. They can be identified quickly with great helpdesk software.


1. Response times

Just four years after tech-leaders declared that software was eating the world, we’ve seen sweeping changes in customer expectations and support practices. By today’s standards, speed isn’t a bonus, it’s a necessity. Take these stats:

  • Phone: 53% of customers find 3 minutes to be a reasonable time while waiting for a support agent.
  • Email: Oracle declared in 2011 that one week was the longest that customers would wait before ending business with a company. By today’s standards, 24-48 hours can be unacceptable.
  • Social: 42% of consumers expect a response on social media within one hour, and 32% think it should be within 30 minutes.

With that in mind, we can naively push our teams to “have all tickets answered within 12 hours”. But as support professionals, we’re also aware of the delicate balance in setting poor expectations and underdelivering. Speed is priority most of the time, but not without context.

  1. First reply – Speed is king for the first reply. Contrary to what we would assume, most customers prefer a quick but “ineffective” response over a calculated, delayed answer. Immediate “your inquiry has been received” responses are standard, so wow your customers with a quick, personalized first response.
  2. Time between replies – Cadence is what’s important here. For complicated issues with longer resolution cycles, customers need to know their requests aren’t dropped. Perform regular check ins, but make sure you are updating customers with tangible progress.
  3. Resolution time – The ideal time taken to resolve customer issues should be minimized. Set clear and realistic deadlines with customers and consider tracking time undelivered promises to better train your team.
  4. Number of replies per ticket – This should be minimized in tandem with resolution time. Aim for as few replies as possible to get your customers to their resolutions.


2. Ticket volume

In some ways, more support tickets can feel good. It means your support network of collection forms, live chat, etc. are accessible and that customers are invested enough to get in touch instead of jump ship.

But being that support tickets are direct feedback for instances where your product fell short or was confusing, we should always aim to minimize the number of support tickets for the following reasons.

  • Getting to “inbox zero” is a herculean task for many. The more time your company spends putting out fires, the less time you have to build better fireproofing (i.e improving documentation, the product, etc).
  • Frequent, recurring tickets indicate issues in how customers derive value from you product. To get to the heart of product friction, maintain clear labeling and ticket tagging schemas.


3. Channel attribution

Take a page from marketing and use channel attribution (i.e tying user actions or sources to outcomes) to get more clarity on customer complaints.

Say your 40 support tickets this week broke down to:

  • 20 from live chat widget
  • 10 from support@yourcompany.com
  • 10 from knowledge base

Find patterns throughout your sources. Maybe inquiries from the knowledge base are more technical while live chat requests are simple onboarding questions. Use this info to deliver the right solutions at the right time.


4. Customer experience rating

Customer ratings are necessary to gauge your support team’s effectiveness but consumers generally hold binary oppositions when prompted for feedback (for/against, like/dislike, etc.). Also consider the following:

  • For every customer who bothers to complain, 26 other customers remain silent.
  • It takes 12 positive service incidents to make up for 1 negative incident.
  • The average “wronged customer” will tell 8-l6 people about it. Over 20% will tell more than 20.

Source: Lee Resource, Inc.

Customers don’t just make quick judgements about how they feel,they tend to be more vocal when upset. Striving for higher support ratings is a given, but don’t dismiss the unhappy customers as a lost cause. Smooth out negative conversations and take them as learning opportunities, even if there’s no chance of winning a customer back.


5. Knowledge base traffic

Getting these numbers will require joint work with marketing. Here are three battle-tested web page metrics that will measure the effectiveness of your knowledge resources.

  • Bounce rate is when someone visits your site, then leaves it without navigating to other pages within it. Logically, we assume people who “read and leave” found a page to be ineffective. However, it can be a good indicator when customers get their answers and can move on.
  • Time on page is important when coupled with bounce rate. Short durations point to clear, concise documentation. Long times on a page could mean your documentation is difficult to understand. This really depends on the technical level of your product.
  • Conversion rates are indicators of success when customers perform an intended action on a web page (sign up, click-through, subscribe, etc.). Think about what the goal of your knowledge base is. Work with marketing to track activities beyond your knowledge base. Should customers return to the application when they find the info they need? Or are they clicking through on the “contact us” button because they need more help?


6. Response lengths and sentiment analysis

Take a page from the field of data science and perform sentiment analysis on your inquiries and responses. And yes, you can do this without highly technical methods. To start, add sentiment tags to your process (happy, unhappy, dead-ends, delighted, etc.) and analyze your ratios.

Go one step further, take a sampling of support transcripts and analyze the actual language. Think of questions on improving your process and answer them with data.

  • Can your team maintain effectiveness with shorter responses?
  • How often are your support reps needing to apologize?
  • How often are your customers left with “dead-end” responses?


7. Proactive support tally

Our final metric is not easily gleaned from any analytics dashboard. Seeing how many times your support team goes from “reactive” to “proactive” can be the difference between having a good vs great support team. While this encroaches into the area of customer success, I believe proactive reach out is here for the long run.


Go beyond metrics. Get insights

As you may have noticed, the last three support metrics were less obvious than the first four. In some cases, they might require you to go outside your scope as support to gather and combine data from different departments and sources.

But by being data-driven and being open to unique strategies, you can gain new perspectives to improve the quality of your customer support experience.