Measuring Success in Customer Service

What is the best way to measure customer support levels?







By: Alon Cohen EVP & CTO of @alonc

Many organizations see customer support as a cost center and, there is certainly cost associated with delivering good or bad service. Many companies try to reduce that cost by looking at different KPI (Key Point Indicators) and try to improve different aspects of those measurements.

Some standard measurements might be:

  • Average and peak time a caller is on hold.
  • Fraction of calls answered in less than 1 minute.
  • Number of calls required to close a ticket.
  • Average length of a call.
  • Support level as graded by feedback from customers (by e-mail or phone).
  • Average time it takes to resolve a problem.
  • First Call Resolution percentage.
  • Cost per ticket.
  • Percentage of calls being abandoned.
  • Agents’ time utilization.

And so on.

As much as some of those KPIs are common sense and seem logical, optimizing one KPI can harm another and optimizing all can become expensive.

However there is another way to approach the problem. First, the company needs to take a more global perspective and look at the contribution of customer support to the bottom line.

For instance, it is a known fact that, for most companies, keeping an existing customer is more cost efficient than acquiring a new customer. Therefore, if better customer support can reduce customer attrition, it is clearly good for the bottom line of the company.

As a call center manager, if you stick to KPIs, you can lose sight of the way you affect the bottom line of the company, even if your cost is going down. You can do more harm if you stick to the wrong KPIs. So, how can one tell what KPI to improve and which one to ignore?

The solution is relatively simple.

The first thing you need to do is define the goal or the global measurement of the company (not the support department) that you want to improve upon. For a phone company like one such logical measurement would be churn. Now, without getting into complex optimization math, if you draw a graph (and I know Jeremy (@jtwatkin) is doing that on a regular basis) of the past levels of the different KPIs you measured and correlate that to a graph of your defined goal, you can usually see very quickly which of those graphed KPIs are most “negatively” correlated to the global goal you plan to improve.


For example, if you see that in a week where average hold time went up for more customer terminated accounts and when the average hold time goes down, less customers leave, you can easily conclude that average hold time is a KPI you probably need to improve on.

I am sure you will find some KPIs that have no effect or no relation to your global goal. Those should not need to be a priority for you to improve on.

If you have an analyst at your disposal you can probably calculate what the optimal point of improvement of the KPI in terms the bottom line is. You can also calculate what the point of diminishing return or even what other KPIs could be compromised in order to help the one you need to improve, while keeping the support budget in check; all that while contributing positively to the bottom line of the company.

We would love to hear what KPIs you found to be most important at your support organization, and if a similar or different process worked for you!

Share this post:

One comment

  • Customer effort is also a great KPI to keep an eye on, especially since it can tell you about process issues on your end. Does a customer go through 2 or 3 channels before contacting the call center? Why did they do that? Were they forced to because those other channels failed?

Leave a Reply

Your email address will not be published. Required fields are marked *