In this post we are going to touch on one of the most important KPIs that any demand planner should focus on: Forecast bias
Forecast bias is the general tendency for a forecast to be higher or lower than the actual value.
Forecast bias is distinct from forecast error in that a forecast can have any level of error and still be completely unbiased. For instance, even if a forecast is fifteen percent higher than the actual value half of the time and fifteen percent lower than the actual value the other half of the time it has no bias. If the forecast is on average fifteen percent higher than the actual value has both fifteen percent error and fifteen percent bias.
Bias can exist in statistical forecasting or in judgment methods. With statistical methods, the forecasting model must be adjusted or switched to a different model. For judgment methods however, bias can be conscious and driven by certain incentives provided to the forecaster or it can be unconscious.
How to simply calculate forecast bias at an aggregated level?
BIAS = Historical Forecast Units minus Actual Demand Units.
The impact of bias can mean that either an organization is holding too much inventory (over-forecast bias) or missing sales due to service issues (under-forecast bias).