Forecasting is the process by which management can identify environmental trends and predict future events. It involves a number of common techniques ranging from simple task of monitoring current events in newspapers and journals to the more sophisticated econometric models.
Each manager must identify explicitly how future condition will affect operations. The conditions in the external environment are outside oneâ€™s control, but they must be estimated so that that the organization can quickly adapt to changes that are occurring rapidly.
If one wishes to forecast general economic conditions, the following components of the gross national product (GNP) provide a framework for the approach: consumer purchases (including durable consumer goods, nondurable consumer goods services); private investment expenditure (construction, durable equipment, inventory build up); government expenditures (federal, state, and local spending); and net exports or imports (the difference between imports and exports). Using the GNP model, one can make general estimates of specific values for each of these components and with the help of published forecasts the manager can arrive at an estimate of general economic conditions.
The demand for the industry can be viewed as consisting of such components as sales of products to new customers, sales of additional products to old customers, replacement sales for products that have worn out, and sales affected by recent technological developments.
A number of approaches for specific forecasting are available as follows:
1. Quantitative time series analysis: A study of past data such as monthly sales or shipments made.
2. Derived forecasts: If we can discover another phenomenon that has been forecasted by a government agency or expert, and this phenomenon is closely associated with the variables that we need to predict, a forecast can be derived from these other estimates.
3. Causal models: If an underlying cause for the variable can be determined, the forecast can be handled mathematically and produce quite accurate results. For example, one might find that sales are the direct result of the number of contacts by salespeople and predict that from every five contacts, one sale will result.
4. Survey of plans and attitudes: The University of Michigan has, for a number of years, been successful in using statistical samples of consumers for determining their plans and attitudes about purchasing the future. This is now followed by most Market Research organizations for forecasting demand or similar parameters.
5. Brainstorming: On the assumption that â€˜two heads or more are better than oneâ€™, one method for predicting the future is to assemble a group of people with knowledge and interest in a specific problem and encourage free flow of creative comments. The conditions required for these brainstorming sessions are important (a) No participant may criticize any idea, regardless of how farfetched it might be; (b) each participant is encouraged to supplement the comments of others and to provide inputs for future estimates; (c) after recording the comments during the meeting, a manager may then construct a forecast built on the variety of ideas from the group.
6. Delphi Method: The judgment of experts is sometimes the best and most feasible method of forecasting. The Rand Corporation developed the Delphi Method as a means of forecasting by seeking expert opinions. The method contrasts with brainstorming in securing independent judgment by having experts complete a detailed questionnaire independently and without knowledge of the responses of other experts.
7. Contingent forecasting scenario: One approach for handling the lack of precision in forecasting is contingent forecasting and planning. At the heart of this approach is the development of several scenarios, each scenario providing a different set of assumptions about future events. The scenario describes a logical sequence of events that might occur in the future.
The forecasting techniques do not replace managers or their intuitions and judgments and the manager cannot entirely depend on these techniques, the reasons being the lack of reliability of information used for forecasting.