Importance Of Forecasting And Controlling Errors

"I have seen the future and it is very much like the present, only longer." says Kehlog Albran in his book The Profit. This pseudo-philosophy is actually a concise description of forecasting, the science of predicting future events. From an operational point of view, market opportunities are the driving force behind production decisions and these opportunities are compiled in the form of demand forecasting which then provides the input for planning production: process design, capacity planning, aggregate planning, scheduling, and inventory management. But why forecasting is so important for operations?

In order to understand the factors of forecasting, one should imagine himself as a part of a supply chain - e.g. a factory. A factory's job is to be able to supply the market demand with lowest operating costs possible. Forecasting in a factory plays the hardest role of knowing what to produce now in order to supply the demand in the future and containing the resources available on hand to do this. The challenge is not only to come up with the future demand and the efficient manufacturing design but also to beat the lead times in between the chains in the systems. The errors can be costly in this process. Overshooting in the forecasts will result in inventory costs in the factory, where underestimating will cause late orders, extra labor costs, missed sales opportunities, stockout costs, and even production close downs due to the lack of raw materials since not being ordered on time. Moreover, as the variety and the lifetime cycles of the products increase, (lifetime cycles actually decrease) the process becomes even more sophisticated since now you need to know the production queues and where to stop producing a certain type in addition to the challenge of knowing the right numbers to produce. And unfortunately one should note that in reality the only certain thing about a forecast is that it will be wrong and that's why there will always be some costs. That is why the question for operations management has been the degree of error in forecasting process and the focus is to reduce the bias and deviation in the forecasts.

With today's globalized world - the increased variety and the numbers, the job of forecasting is getting even tougher. As Marshall L. Fisher argues in his article of Making Supply Meet Demand in an Uncertain World; the old-popular systems like quick response programs, just in time inventory systems, manufacturing resource planning, and alike are simply not up to task . In order to reduce the cost of manufacturing, companies started look for better strategies. The accurate response system is one of these new approaches to the forecasting process which provides a way to figure out what forecasters can and cannot predict well, and then making supply chain fast and flexible so that managers can postpone decisions about their most unpredictable items until they have some market signals1. Accurate response system takes into account two new elements in forecasting process: missed sales opportunities plus the distinction of predictable and unpredictable products. It demands