STATISTICAL APPROACHES
A forecasting method is sometimes adapted for direct use as the initial step in developing a sales Plan. Use any statistical method for this narrow purpose requires a technically trained person. For example, in one medium-size, highly decentralized company, each autonomous plant manager is required to develop (1) a five-year profit plan for the plant and (2) a comprehensive one-year profit plan including a marketing plan and all other related plans for the plant. As the initial step, the economic research staff develops projections of approximately ten different indexes relating to various economic indicators (causal variables), such as gross national product housing starts, regional bank deposits, and a projection of probable sales the plant. These forecasts are given to each plant manager. Each plant manager may decide to Use one or more of them (or none) in developing the profit plans for the plant. Starting with the statistical forecast of sales, the plant menagers develop their own profit plan. Approval is by the president of the company.
Most managers are not familiar with the Theoretical aspects of forecasting and its limitations because they do not have the inclination or time to investi gate it in depth. Yet many companies have a forecasting function because it can help them cope with an increasingly uncertain environment. Management planning involves uncertainty, and reliable forecasts can help reduce the uncertainty in planning.