Forecasting for Sales and Materials Management

This chapter covers the family of techniques known as time series analysis. These are a set of techniques which all make the same underlying assumption, which is that the only independent that needs to be considered is time. The use of time as a proxy variable to simplify the complexity that exists in reality results in methods that are arithmetically straightforward and simple for the beginner to use. It should also be borne in mind that even though some of the time series techniques are extremely simple, they are capable of yielding very accurate results, provided they are applied in the correct circumstances.

This chapter covers the family of techniques known as Causal forecasting techniques. These are a set of techniques which again all make the same underlying assumption, but in this case, unlike time series analysis, the assumption is that there is a discernible relationship between the forecasted dependent variable and a measurable independent variable. Once this relationship has been established, the value of the forecast can be arrived at by inserting the known value of the independent variable into the relationship. A tacit assumption of this type of objective approach is that the independent variable can be measured to at least the same accuracy as that to which the forecast is desired. Rather than use a complex relationship between the forecast variable and an independent variable (which has also to be guessed), it is simpler, and quite probably more accurate, merely to guess the forecast. It will usually be cheaper and quicker as well.