How To Improve Forecasts With Hybrid Forecast Inputs

HOW TO IMPROVE FORECASTS WITH HYBRID FORECAST INPUTS By John A. Gallucci Hybrid forecasting inputs (statistical, sales-driven, POS, etc.) enhance forecast accuracy ... provide a mechanism to discard inputs that do not add value ... by ensuring planners that forecasts are based on the “best of the best” input data, they will be assured that they will have the most accurate demand plan. T T he information used to derive best practice forecasts can come from many places. In the consumer packaged goods environment, viable inputs include statistical forecasts, sales forecasts, marketing estimates, annual budgets, syndicated data, and CPFR (Collaborative Planning, Forecasting, and Replenishment), to name just a few. These inputs certainly create an overwhelming amount of information. The best data in this set will lead us towards the best forecasts, while the worst will lead us to disaster. The difference between these scenarios can be several million dollars. How can we decide which inputs will give us the best results? It is a common practice for many demand management teams to measure error by forecast input. When considering future demand, a heavier weight is placed ...