Demand planning & control -handling multiple perspectives through a holistic approach to hierarchical planning

Hierarchical Demand Planning (HDP) is an intricate part of most companies today. HDP is based on the assumption of independence among variables, and this allows for simple and easy aggregation and separation of plans and data. However, the most commonly used arguments for grouping and subsequent aggregating is shared traits contrary to the assumption of independency. One of the predominant issues is the conflicting objectives on different decision levels. An example of this is found in hierarchical forecasting of demand. When forecasting on e.g. a product family level to establish capacity requirements, the objective is usually to achieve a Mean Error (ME) of zero. This conflicts with forecasting for Demand Planning (DP) purposes on SKU level, where minimization of the Standard Deviation of Error (SDE) might be more important. In this paper these issues are addressed through a simple example of hierarchical forecasting and use of a Goal Programming (GP) approach to satisfy both objectives. It is found that some general guidelines for handling multiple objectives within HDP can be inferred from this, leading the way for a holistic demand planning framework.