This article appeared under the title “Sigmund Freud and Contact Center Monitoring” in the September issue of Contact Center Pipeline.

Let’s do a “free association” exercise.  What pops into mind when you consider the phrase “Call Center Monitoring Technologies?”  Just think for a moment.  Depending on how successful the technology monitoring initiatives have been in your center, you might have thought of one or more of these:

• “Scrolling up and down waiting for something to stick out”
• “Dividing one column into another in my head”
• “Trying to remember that obscure queue name”
• “Endless displays of SNMP data”
• “ASA?  SL?  Why can’t I get revenue per agent per hour?”

Many times, we get used to using information our tools provide us, rather than getting creative and either developing or purchasing what we really need.  I don’t mean creating yet another custom report that adds your boss’s favorite KPIs, I’m referring to shifting strategies to maximize your time and improve your visibility into business performance.

I propose three alternatives to the common monitoring solution:  combine disparate systems, focus on outliers, and measure KPIs of metadata.  Let’s take a minute to unpack that.

Combine Disparate Systems

By insisting that monitoring systems be combined, you can create new metrics that would be previously difficult to achieve.  For example, what if you could take call volume forecast data from your workforce management system and combine it in real time with actual call volume? This would let you see how you’re actually performing against forecast in real time!  Or, what if you could access actual, raw sales data and use it to promote contests among your agents during the day? In addition to the productivity bump you’ll achieve with agent performance, the information will be a lot more telling than simply comparing handle times across different pools of agents who work on different types of calls.

Focus on Outliers

Focusing only on outliers has the advantage of not cluttering up your day with “normal” – you’ve spent a lot of money on systems that are supposed to work just the way you want them, with “99.999%” reliability, so why waste your time watching it run?  Monitoring systems should only show exceptions – unless you want to drill down into minutiae for historical analysis.  Multicolored thresholds are useful, but designing a system that hides, by default, any data that isn’t outside a normal range saves time and energy.

Measure KPIs of Metadata

Most call center KPIs are applied to specific agents or agent groups, or queues and groups of queues.  Imagine instead if you could see service levels for a specific line of business or geographic region rather than having to mentally average across several groups of queues.  A properly designed monitoring system should allow you to start from an enterprise view and navigate the call center space with familiar groupings—and subdivide in any way you choose—without having to know that certain business units share certain resources or which queues belong where.

Baking intelligence into call center technology monitoring systems saves time and allows management to react quickly to unexpected turns throughout the day, smoothing out the peaks and valleys of performance.  However, these things aren’t easily accomplished. Combining data feeds from unlike source systems requires a very open architecture with the ability to tie metrics to some common framework (business unit, customer, geographic region, etc.). USAN offers such a platform, and you can learn more about it at usan.com. But, enough of the sales pitch. Back to our free association!

With these thoughts in mind, take a fresh look at how you monitor your call center. Consider what information you need to make decisions, rather than what each technology vendor offers out of the box, and make your monitoring systems infinitely more useful.