Forecasting can be broadly considered as a method or a technique for estimating many future aspects of a business or other operation. Planning for the future is a critical aspect of managing any organization, and small business enterprises are no exception. Indeed, their typically modest capital resources make such planning particularly important. In fact, the long-term success of both small and large organizations is closely tied to how well the management of the organization is able to foresee its future and to develop appropriate strategies to deal with likely future scenarios. Intuition, good judgment, and an awareness of how well the industry and national economy are doing may give the manager of a business firm a sense of future market and economic trends. Nevertheless, it is not easy to convert a feeling about the future into a precise and useful number, such as next year's sales volume or the raw material cost per unit of output. Forecasting methods can help estimate many such future aspects of a business operation.
The goal of forecasting is to come as close to possible to an accurate picture of the future. But, as with other forms of fortune telling, it can never be fully accurate. There are simply too many interactive variables. A change in any one of these may cause the forecasted scenario to change. For example, unexpected shocks to the economy, as occurred after the terrorist attacks of 9/11, are extremely difficult to anticipate and plan around. Such extreme situations are, happily, very rare. But there are far more subtle events that may also cause major changes in the assumptions upon which a forecast is based, things like: sharply increased material costs resulting from storms or wars, the unexpected demise or buyout of a large competitor, and/or an increase in demand due to an unexpected fashion trend shift. Despite the fact that forecasting is an imprecise art, a company must do the best it can to plan for the future and an important part of this planning is forecasting.
The task of forecasting can be approached in a number of ways and the best forecasting outcomes are usually the result of applying several forecasting methods. To supplement their judgment, forecasters rely on a variety of data sources and forecasting methods. For example, forecasting may involve the use of econometric models that can take into account the interactions between economic variables. In other cases the forecaster may employ statistical techniques for analyzing sets of historical data referred to simply as time series. Other frequently used data sources are recent consumer surveys and forecasts produced by other institutions—industry associations, investment banks, and economists generally.
In an era where forecasts drive entire supply chain networks forecasting is an increasingly critical organizational capability. Forecasting the future may sound like a lofty and theoretical activity when in reality it is a practical business tool like many others. Here is an example. How should a business go about preparing the quarterly sales volume forecasts for their primary product, say, window-glass? The company will certainly want to review the actual sales data for window glass over the last few years. Suppose that the forecaster has access to actual sales data for each quarter over the 15-year period the firm has been in business. Using these historical data, the forecaster can see the general level of sales but more importantly, he or she can also determine what pattern the sales history produces, what trends are visible. A thorough review of the data may reveal some type of seasonal pattern, such as peak sales occurring in the spring as people do spring-cleaning and others prepare to sell their homes during the summer school break. In addition, if the forecaster is able to identify other factors that influence sales, like weather patterns or housing starts, historical data on these factors can also be used in generating forecasts of future sales volumes.
Academics divide forecasting methods into two broad categories: qualitative and quantitative. The division of forecasting methods into qualitative forecasting and quantitative forecasting is based on the availability of historical time series data. If historical data and time series are available, than quantitative methods may be used. If not, qualitative methods are the only option.
Qualitative forecasting techniques generally employ the judgment of experts to generate forecasts. A key advantage of these procedures is that they can be applied in situations where historical data are simply not available. Even in situations where such data are available, quantitative forecasting methods are a useful addition to successful forecasting. Three important qualitative forecasting methods are: the Delphi method, scenario writing, and the subject approach.
In the Delphi method, an attempt is made to develop forecasts through "group consensus."
A group of experts in a particular field participate. Usually, a panel of these experienced people is asked to respond to a series of questionnaires. The panel members, who should ideally come from a variety of backgrounds (marketing, production, management, finance, purchasing, etc.) are asked to respond to an initial questionnaire. A second questionnaire is then created which incorporates information and opinions gathered in the responses to the first questionnaire. The second questionnaire is then distributed. Each panelist is asked to reconsider and revise his or her initial response to the questions based on the new information. This process is continued until some degree of consensus among the panelists is reached. It should be noted that the objective of the Delphi method is not to produce a single answer at the end. Instead, it attempts to produce a relatively narrow range of opinions—a range into which most of the panelists' opinions fall.
Under the scenario writing approach, the forecaster starts with different sets of assumptions. For each set of assumptions, a likely scenario of the business outcome is charted. Thus, the forecaster generates several different future scenarios (corresponding to different sets of assumptions). The decision maker or business person is presented with the different scenarios, and has to decide which scenario is most likely to prevail.
The subjective approach allows individuals participating in the forecasting decision to arrive at a forecast based on their feelings, ideas, and personal experiences. Many corporations in the United States have started to increasingly use the subjective approach. Internally, these subjective approaches sometimes take the form of "brainstorming sessions," in which managers, executives, and employees work together to develop new ideas or to solve complex problems. At other times, the subjective approach may take the form of a survey of the company's sales people. This approach, which is known as the sales force composite or grass roots method, is relied on because salespeople interact directly with purchasers and it is assumed therefore that they have a good feel for which products will or will not sell and in what quantities.
The advantage of using the salespeople's forecasts is that salespeople are highly qualified to explain the demand for products, especially in their own territories. The disadvantage is that salespeople may tend to be optimistic in their estimates since optimism is a characteristic often found in good salespeople. Also, those working in sales may fear that a low sales forecast will lead to layoffs in the sales area. The opinions of salespeople should not be relied on to the exclusion of all else for one additional reason. Salespeople may not be aware of impending changes in other related areas, such as availability of raw materials, national economic developments, or the arrival of a formidable new competitor.
Quantitative forecasting methods are used when historical data on variables of interest are available—these methods are based on an analysis of historical data concerning the time series of the specific variable of interest. There are two quantitative forecasting methods. The first uses the past trend of a particular variable in order to make a future forecast of the variable. In recognition of this method's reliance on time series, it is commonly called the "time series method." The second quantitative forecasting method also uses historical data. This method is often referred to as the causal method because it relies on the use of several variables and their "cause-and-effect" relationships. Examples of variables that may have this cause-and-effect relationship are: 1) interest rate levels and levels of disposable income; 2) winter weather patterns and demand for heating oil; 3) increasing gas prices and a decline in demand for sports utility vehicles (SUVs). By studying the time series data on two or more variables that have a cause-and-effect relationship with the item for which a forecast is needed, effort is made to incorporate as many relevant factors as possible into the forecast.
In practice, most business people use some combination of these methods and techniques in trying to plan for the future and put together accurate forecasts. With each cycle of forecasting, more is learned about what factor to consider and how to weight their importance in projecting future events.
Aston, Adam, and Joseph Weber. "The Worst Isn't Over: Smarter science is helping companies and insurers plan for hurricanes. The Bad News: This year could be another doozy." Business Week. 16 January 2006.
Chase, Charles W. Jr. "Composite Forecasting: Combining Forecasts for Improved Accuracy." Journal of Business Forecasting. Summer 2000.
Engerman, Stanley. "On the Accuracy of Some Past and Present Forecasts." International Monetary Fund Staff Papers. Annual 2005.
Evans, Michael. Practical Business Forecasting. Blackwell Publishing, 2002.
Gaber, Tal, Jacob Goldenberg, Barak Libai, and Eitan Mullerray. "From Density to Destiny: Using spatial dimension of sales data for early prediction of new product success." Marketing Science. Summer 2004.
Gray, Andi. "How Forecasting Can Help the Bottom Line." Fairfield County Business Journal. 27 June 2005.
Jones, Vernon Dale, Stuart Bretschneider, and Wilpen L. Gorr. "Organization Pressures on Forecast Evaluation: Managerial, Political, and Procedural Influences." Journal of Forecasting. July 1997.
Mentzer, John T., and Mark A. Moon. Sales Forecasting Management. Sage Publications, Inc., 2004.
O'Connor, Marcus, William Remus, and Ken Griggs. "Going Up—Going Down: How Good are People at Forecasting Trends and Changes in Trends?" Journal of Forecasting. May 1997.
Sanders, Nada R., and Karl B. Manrodt. "The Efficacy of Using Judgmental versus Quantitative Forecasting Methods in Practice." Omega. December 2003.
Rasmussen, Rasmus. "On Time Series Data and Optimal Parameters." Omega. April 2004.