Tips for Measuring the Accuracy of Forecasted to Actual Call Volume

One of the most common measures of the WFM team is forecast accuracy.  But there are lots of variables and options that should be taken into consideration when measuring this statistic.   Here are a couple of things that need to be taken into account:

Forecasting Horizon – Whether it is 18 months, 12 months, 90 days, or one week, each time frame should probably have a goal, ideally getting tighter and tighter as the planning horizon gets closer (e.g., by Friday you should have an extremely high confidence level in your forecast for the following week).

Variability – Each industry/line of business has different challenges.  The goal or expectation needs to be established based on availability of historical information, awareness of the drivers that cause variability, any seasonality that may or may not exist, etc.  It is a lot harder to forecast in an industry that is marketing or sales driven than perhaps a pure servicing environment with a stable and well established billing cycle.

Exception Process – There needs to be some dialogue and agreement regarding the underlying assumptions that make up the forecast and the established goal.  Forecasters need a “Court of Appeal” and a process to explain/defend forecasts that miss the stated goal due to significant deviations or variances from those assumptions that may be out of their control.   This establishes two things – a commitment to understand the causal impacts that created a forecast variance outside of acceptable tolerance ranges, and probably more importantly, helps establish buy in/accountability from your forecasting team that they will not be penalized for things beyond their control.

Note: This tip provided by SWPP Board Member Andy Wainwright of Toyota Financial Services.  He may be reached at andy_wainwright@toyota.com.