The forecasting of demand is fundamental to many commercial and operational decisions. It facilitates critical business activities such as financial planning, sales and marketing plans, material requirement planning, and production planning to name a few. Poor forecasting creates no end of issues in any supply chain.
Forecast accuracy decreases with granularity and diminishes the further into the future the estimate is made. Using stock-keeping unit (SKU) forecasts to plan and execute supply chain operations is to plan and execute knowingly in error. Much activity is then devoted to trying to correct operations to respond to actual demand; the alternative is to gear operating models to demand in the first place, that is, be demand-driven. It is in setting resource levels and parameterising demand-driven operating models that demand forecasts should play a critical role, not the triggering of replenishment and production.
The design and operation of the demand forecasting and planning process, including roles and responsibilities, systems, and required skillsets, should promote accuracy and, critically, reduce bias. The consistent generation of a reliable, data-driven, statistical forecast will help to drive supply chain and financial efficiencies. We apply Flow-First design principles to adapt the demand forecasting and planning process.
