May 29 2008
Service Level vs. Cost vs. Customer Experience
I’m out at a customer site this week and overheard the following conversation from the Workforce Management Team:
The difference in customer experience between 93% Service Level and 100% Service Level is negligible. But the difference in staffing cost to us is huge.
Now, I’ll spare you my full rant about Service Level (you can find it here) but I think this is indicative of a larger perception and education problem in the call center industry. Simply put: Service Level is NOT a measure of Customer Experience. It’s an opaque metric.
Let’s assume for purposes of this argument that the service level goal is 93% calls answered within 20 seconds. By decreasing the goal from 100% to 93% you’re saying that it’s okay for 7% of your customers to sit in queue for longer than 20 seconds. Logically, and most likely what the Workforce Manager was thinking, the effect on customer experience by being answered in 19 seconds vs. 21 seconds is unnoticeable. However the real impact to Customer Experience between 93% and 100% is actually immeasurable from Service Level alone. You have no way of knowing how many of the calls in queue longer than 20 seconds were answered in 21 seconds or how many were answered in 20 minutes and 21 seconds.
Try explaining to the irate customer that listened to hold music for twenty minutes that his difference in customer experience was “negligible”.
Chris,
I agree with your concerns about the experience of the other 20%.
What is the approach that you advocate (to replace the use of service level)?
D
I thought I might tell you about a story from a call centre visit I had some ten or twelve years ago. It was to a top 20 British company and the call centre manager I visited was bonused for making the 80% of calls in 20 second service level. He was very happy because every month he made his target and collected his bonus, but without the need for overstaffing. He wouldn’t tell me how he did it, but when I looked at his Call Routing rules I soon saw how he was achieving it. All calls that came in were allocated a medium priority level. However once a customer had been waiting for 20 seconds their priority was lowered. This is obviously counterintuitive as you would expect the customer’s priority to be increased the longer they wait. But it’s not counterintuitive if your bonus depends on the service level being met. After all, if you have two customers in the queue and one has been waiting 10 seconds and one has been waiting 30 seconds, why would you answer the longer waiting customer? They’re already outside the service level, so they might as well wait longer. Answer the one that counts towards your bonus. The problem here is the metric – it’s useful for measuring routing efficiency, but it’s one dimensional and it doesn’t reflect the customer experience. It could also be very dangerous. For instance, what if that customer who is forced to wait minutes (if they ever get answered at all) was one of your best customers, or about to place a large order or a blogger with a readership of thousands? The call centre manager’s bonus would be negligible compared to the lost business.
Chris,
I agree with your concerns about the experience of the other 20%.
What is the approach that you advocate (to replace the use of service level)?
D
I like the irony of what is meant by ‘opaque metric’ being fairly . . . well . . . opaque to be honest.
Service Level is a indication of variation from goal. It is only one indicator of the Customer Experience.
But don’t discount the power of regression analysis there are strong coeraltions between service level and ASA & Abandons. Not to mention net staffing.