1. Forecasting

The forecast is perhaps the most important function in controlling the inventory. In a typical inventory entity, forecasts are needed for each of the future months up to the planning horizon, typically 12 months. Data from the past demands is needed to generate the forecasts. Assuming monthly time buckets, the demand for a fixed number of history months, (usually 12, 24 or 36), is saved in the database. The monthly demands represent the number of pieces ordered on an item for each month. Many entities also save the lines per month. These are the number of customer orders that arrive during a month. The history months are mostly saved in calendar months, but some entities use fiscal months of the 4, 4, 5 type. It behooves the entity to process the demand history prior to forecasting, by running a routine to seek if any outlier demands are found, and if so, to adjust them prior to forecasting. Another useful routine would seek to adjust the demand history if any demands are returned due to a ship-in-error by quantity or by part.