TruCallCenters: World-Class but Still an Error Machine

TruCallCenters.  It is my tongue-in-cheek version of TruTV.

My first installment was about how Call Centers use yellow smiley balloons and laminated cards as Quality Improvement strategies: http://ifyouwanttoscream.blogspot.com/2013/10/yellow-smiley-balloons-laminated-cards.html

The second was about a call center in Florida with 350% annualized turnover and suggested that we had better plan for the reality of high turnover (OK, maybe not 350% everywhere...but north of 50% is not uncommon at all) and figure out how to drive continuous improvement even with new agents:  http://ifyouwanttoscream.blogspot.com/2013/12/latest-installment-of-trucallcenters.html

This installment is about how you can be a world-class call center and still be an error-producing machine.  I ran the call centers for a Fortune 500 company a few years ago.  We had a great reputation in the industry.  But as good as we were, honestly, we still sucked.  Every single day we made hundreds of errors…wrong email addresses entered, wrong diagnostic steps, missed consumer disclosures, in appropriate/incorrect transfers, incorrect call documentation, inaccurate After Call Work documentation, accent escalations from customers, poor warranty compliance, failure to cross-sell when we were supposed.  I literally could go on and on with the list of errors we made, day in and day out.  And we had centers that were genuinely looked up to with good CSAT and Loyalty scores!

As an aside, this idea that even a great call center can be an error producing machine is related to a previous post I did called What is an Acceptable Contact Center Error Rate?  In that post, I present example after example of ridiculously bad contact center performance that is rather matter-of-fact.

Two insights have come into graphic relief for me is since I was running centers of my own.  The first is that the root cause of all the inefficiencies, errors, and customer satisfaction issues is between-agent variation in processes and therefore outcomes.  You get astonishingly different experiences from call to call, from agent to agent...even from the same agent between the beginning of her shift and the end of it...because, for a dozen different reasons, the agents don't consistently follow the correct process. 

The second blindingly clear insight is that the ability to control/reduce between-agent variation in process and outputs with monitoring and coaching is severely limited.  You can read a much deeper discussion here Call Center Coaching Remains a Labor-in-Vain)

Someday, in the not too distant future, call center leaders are going to wake up and say, "I'm tired of this.  I am tired of living the Myth of Sisyphus.  I am tired of the turnover.  I am tired of calibration meetings.  I am tired of agreeing to standards and having agents fail to comply 20% of the time.  I am tired of agent output metrics charts that never improve but just continue to tread water.  I am tired of the best call centers in the world still producing thousands of errors a day.  What we are doing is not working and I have to find a better way."

There of course is a better way and it is no more complex than baking automation into the process the agents are using to virtually guarantee that the agents do and say what they are supposed to, when and how they are supposed to do it.

When call center leaders get tired of beating their heads against the wall and implement agent-assisted automation, you will see a step function change in call center outcomes that will create a better work environment, better results for customers, and profitability improvements that will surprise the skeptics.