KPIs: The Good, The Bad and The Ugly

Key Performance Indicators are nothing new to anyone in the contact center world. Everyone uses them in different ways. Some use them as the basis for bonuses, some for yearly reviews, and some just to measure how well their contact center is doing either for internal tracking, or comparing to competitors. The questions I usually ask people are:

  • Are you using them in a way that makes sense?
  • Is there a way to game the system?
  • Could you accomplish the same with less?

The good - when it comes to KPI’s is that they are called ‘key’ for a reason, they contain a lot of really useful data that when used properly can help you maintain consistent service and stay above your competitors. The bad - when it comes to KPI’s is that they are easy to mislead you and have you think you’re doing great when in reality you don’t look at the whole picture. You need to be sure you’re measuring your contact center against a multiple set of KPI’s, using them as a system of checks and balances against each other.

  • Are you measuring and monitoring the correct KPI’s?
  • Which ones do you think are the most important?

I generally find that Contact Centers are measuring against a single KPI. Many times, this KPI is calls handled, average talk time, customer satisfaction rating or something similar. I find that any time there is a single factor KPI, agents find a way to cheat the numbers in their favor. If average talk time is the only factor, they will do or say anything to get off the phone as quickly as possible. When this occurs, most often customer satisfaction is poor. Conversely, if customer satisfaction is the only factor, then you’ll find people spending forever on calls that should have taken just a few minutes in order to keep the customer happy. This will usually cause hold times to skyrocket and abandoned rates to climb. The ugly thing about KPI’s is that they can be cheated if you don’t watch how you’re measuring them. It can quickly become overwhelming. As with any data collection, the numbers are only as good coming out as they are going in. Gather the data accurately; make sure what you measure and how you’re measuring makes sense and serves to balance each other out so that you’re not misleading yourself with data that doesn’t actually represent what you are trying to show. Just because you have an average talk time of under 2 minutes, would it be correct to assume your customers are happy and that you have high call resolution? What if a high percentage of calls meet your SLA? Hopefully it means your agents are doing a great job keeping customers happy…but how do you know for sure? It is also easy to get sucked into wanting to compare 50 - 100 different stats, when a lot of the information included may not actually affect your business. Look for the stats that matter for you and your customers based on how you do business. So how do you stay in the good and avoid the bad and the ugly? Make sure you have planned out the KPI’s that you will measure your contact center performance against in way that creates checks and balances. I suggest combining KPIs into groups that make sense based on your business needs. An example: by combining calls handled and customer satisfaction… This way the agents are encouraged to keep the customer happy, but limit the length of time they are on the phone. Some other combinations might be:

  • cost per call and agent utilization
  • customer satisfaction and first call resolution
  • service level and customer satisfaction.

One last note, when I say combining, I mean requiring both to be above predefined levels, not just one or the other. Remember to do what makes sense for your business, don’t just blindly follow trends. Measuring more stats that don’t really matter to your business is a waste of your time, energy, and money.