Here are two typical conversations I’ve had when consulting with contact centers.
Exchange No. 1
Hi, we’d like to spend the next half day sitting next to your best customer service agent, double-jacking into their calls with your customers or observing how they do their email. In this way, we can learn the most about your systems, your customer issues and how to improve customer care. So, who’s your best agent?
Customer service director (after staring blankly around the room):
Gee, well, I guess that would have to be Kayla, since we have somewhere here a really nice letter from a happy customer. Or no, that’s Gerry, since he’s got the best AHT [average handle time] two months running. Or maybe that’s Sam, since she’s got such great attendance year to date or …
Exchange No. 2
Customer service metrics manager:
Our agent attrition [annualized turnover] last year was 60 percent—not bad, compared to the three benchmark centers we study across town.
“OK, but what’s the attrition for your best agents, the ones you hate to lose? And what’s their attrition year to date [annualized]?”
CS metrics manager:
Uh, well, we don’t calculate it that way. What do you mean by best agents?
What do these two quick exchanges have in common? First, we’ve run into them time and time again with our customers. Second, they fall into the trap we like to call the “Garrison Keillor caveat,” just as the Prairie Home Companion host and author says of his fictional Lake Wobegone, a place “where all the women are strong, all the men are good-looking and all the children are above-average.”
They’re not, of course. Nor are customer service agents. Some are doing an awesome job with customers, overcoming system limitations or minimal post-hire training amidst proliferating products or services. But others are struggling to do their job.
To produce optimal overall performance for the customers who are bothering to call, email, or launch web chats for help or advice, customer service agents need to know requirements, just like employees in any other profession. Customer service agents need to be recognized for their balanced performance across two opposing dimensions, each axis tied into corporate and overall customer service objectives. In this way, companies can finally identify their best agents, figure out how they do what they do, find more agents like them and, most importantly, hold onto them.
So how do you determine best agents?
Speed versus quality
First, you need to figure out the opposing forces that operate on each agent, otherwise called key performance indicators (KPIs). We’d encourage you to start with two (one on each side) before becoming too elaborate. One customer told us, “We have 21 KPIs,” forgetting the important word, “key.” Agents need to know very clearly what’s expected of them, and 21 KPIs are 19 too many.
The classic opposing forces for agents are speed and quality, just as for manufacturing workers, it’s speed and safety. Most companies measure speed all too well, and report it too frequently; it goes by names such as AHT or MPI (minutes per incident). In most cases, we’ve seen it applied equally across all agents, with these three value judgments:
- Faster AHT (or its flipside, shorter MPI) is better.
- The fastest agents get the best shifts or even bonuses.
- The slowest agents get close counseling (to go faster) but eventually are forced out of the company.
Fair? Correct? Not if you consider that customer frustration with customer service is high and getting higher. And so is that of the agents, because of short shrift to quality—and goodness.
This second opposing force is critical and typically follows a “three-stage rocket”:
- Monitor contacts (calls, emails, chat sessions) at least 12 times per agent per month, calibrating a central quality assurance team view with the agents’ supervisors’ scoring. Note: Outsourcing firms need to use their client’s forms and calibrate with them.
- Conduct post-contact customer satisfaction surveys using IVR or email, and then associate results per agent and per contact. Note: This is an excellent vehicle to capture dissatisfied customers right away!
- If the customer makes subsequent purchases, track them to ensure that a “good call” with a “satisfied” customer actually leads to future business/transactions.
Connecting this adjusted AHT and the three-stage quality as opposing KPIs, companies can finally compare speed versus goodness on behalf of their customers; remedy critical situations much more quickly; provide cogent and timely feedback to agents to improve their balanced performance; and identify the best agents.
And those earlier exchanges? This is how they’ll play out now:
Revised Exchange No 1:
My best agent? We have three agents who have consistently scored tops in both speed and quality, and just last Thursday night we recognized them in our all-hands meeting, so let me introduce you to …
Revised Exchange No. 2:
CS metrics manager:
Our best agents turn over [attrite] only 25 percent per year, since we are really paying a lot of attention to how they are able to provide great productivity [AHT] and c-sat!