Demand forecasts are attempted by several organizations and individuals. For instance, the planning Commission undertakes systematic forecasts for the demands for all major goods and services in the economy for the subsequent next five years or so. The industry organizations are engaged in forecasting the demand for their corresponding industrial products and firms in their corresponding brands. Researchers undertake forecasts of all kinds, including the worldwide forecasts, which are also carried out by international organizations like the World Bank, International Monetary Fund, United Nationsâ€™ Organization and Asian Development Bank.
The significance of these forecasting studies can hardly be exaggerated. In particular, these are needed to plan future production and thereby future needs for various resources, including manpower, raw materials and funds. Production of most goods and services takes time, and sometimes this period, called the gestation period, is spread over decades.
For example, production of hydro electric power from scratch might well over a decade and one very knows that Maruti Car came to the market only about ten years after the project was conceived. The case of production of services may be similar. For example, if more MBA graduates are to be turned out, might take several years to achieve that target New institutions might have to be started or / and facilities in terms of buildings, furniture, faculty, staff, library etc may have to be provided to enlarge seats in the existing institutions. Unless the future demand is known well in advance there may not be enough time and execute the production to meet that demand. And if demand is not met, firms may be able to attain their objectives. For example, if production is short of demand, firms might have gone with little or no consumption of the good in question.
If production exceeds the demand leading to high stocks, firms might not be able to sell all their production at reasonable or any prices. Thus, both under and over production are undesirable. To avoid such a situation, accurate demand forecasts are essential.
There is no choice between forecasting and not forecasting. Not to forecast is to assume an indefinite continuation of the status quo. To expect no change in these days of fast changing consumption habits in the face of emerging new products seem very short sighted and highly unrealistic. The area of choice only concerns the way the forecasts is made, who does it and what resources are devoted to it.
Forecasts (genuine or ex-ante), by its very definition, involves the future, which is uncertain. Thus, no forecast can be expected to be cent percent correct. This together with the essential feature of forecasts argues for a paradox: This is true but efforts have to be mounted to obtain as accurate a forecast as possible. The basic question then is how to get good forecasts?
There is perhaps no unique method for anything, and there is not likely to be one either. This is because methodology is always plural. Forecasting methods are no exception to this rule. As is generally the case, there are several methods of demand forecasting basically for three reasons:
(a)no method is perfect and no method (worth its name) is useless
(b)no method is the best under all circumstances
(c)the best method may not be available in a particular situation due to constraints from data and 1/ or resources (time and money)
Broadly speaking there are two methods for demand forecasting. Survey method and Statistical method. Under the first surveys are conducted about the intentions of consumers, opinions of experts and markets, based on which demand forecasting is made.
Under statistical methods, historical data are extrapolated or analyzed through economic models and through them demand forecasts are worked out.