Demand data can almost always be organized into several levels of aggregation. Suppose that the lowest-level forecasts you need to generate represent SKUs (Stock Keeping Units). SKU-level forecasts are often needed to support production planning and inventory control. The corporation might aggregate these SKUs first into products and then into product lines for marketing and sales. These might be aggregated further into geographical regions for the benefit of top management.
For the purposes of this discussion, we will be referring to end-items and groups. End-items are the lowest-level data in your hierarchy. In the above example, each SKU would be an end-item. Groups are aggregations of end-items. In the above example, products, product lines and geographical territories would all be examples of groups.
Forecast Pro allows you to define product hierarchies and create one set of self-consistent forecasts. It allows top-down, bottom-up or user-defined reconciliation, seasonal adjustment based upon aggregate data and model selection at the aggregate level.
It is not necessary that the end-item histories begin and end at the same time. Thus group-level data may consist of end-items that have been retired or replaced by new end-items. Obsolete end-items will contribute to the group-level history but will not themselves be forecasted. You will notice that the starting and ending dates for the overall historic data consist of the starting date for the oldest end-item and the ending date for the newest. Forecasts will be prepared for all end-items and groups that are “alive” at the end of the data set. Those whose histories terminate before that time are considered “dead”—they contribute to the historic aggregates (and therefore influence aggregate forecasts), but they are not themselves forecasted.
See Also