Let’s consider a call centre that is offered 100,000 calls every week but has been suffering from poor service levels for quite some time. Unsurprisingly, the forecasts going forward are also for 100,000 calls per week given that recent call forecast accuracy has been pretty good, despite the poor service levels.
The call centre has been recruiting with vigour, and all of a sudden there are now enough people on the phones to deliver the targeted service level of 80% answered within 20 seconds. At the point where service levels were achieved, it’s seen that the call volume offered drops, first to 95,000 for a week and then to a new steady-state volume of 90,000 calls per week. Thereafter, the forecasts are revised down to the new 90,000 calls per week level.
Why? How is it that the poor service levels in the call centre have caused the forecasters to “overshoot” their forecast so that the centre is now staffed up to deliver 100,000 calls of capacity when the actual underlying volume is only really 90,000? The answer lies in the abandon rate of the call centre which is much more informative than the service levels in such circumstances.