Executives and managers often ask themselves this very question, wondering why some forecasts that they followed closely to make critical decisions turn out to be so very wrong. Knowing what it takes to develop an effective forecasting process, a key element of business analytics, will allow for more accurate forecasts and better decisions.
Forecasts can be wrong due to several reasons:
- Software problems: Either the software is flawed and contains mathematical errors or is misapplied by an untrained user.
- “Fiddling”: Such untrained forecasters may constantly adjust or fiddle with the forecasts based on new information.
- Biases: Personal agendas of the process participants influence (consciously or not) the forecast, regardless of what external factors point to. Are you familiar with the study about subconscious human bias in NCAA basketball tournament selection?
- Impossible to accurately forecast some behaviors: Some behaviors determine forecast accuracy by their nature. Think about a coin toss. In spite of your wanting to achieve 60 or 90 percent accuracy, you will only be right about half the time. It’s the nature of the activity or behavior. The same applies to forecasting products or services.
How can you make a forecast right?
Companies from a variety of industries, including Intel, Cisco, AstraZeneca, Tempur-Pedic and Yokohama Tire Canada, have used the method of forecast value added (FVA) analysis to identify and eliminate activities that fail to improve the forecast – or may even be making it worse. FVA exposes the politics and biases that contaminate the typical business forecasting process, and allows organizations to achieve better forecasts with less cost and effort.
For example, a consumer packaged goods company used FVA analysis and found that its statistical forecast reduced forecast error by 11 percent. The company also found that manual overrides to the statistical forecast actually made it worse 60 percent of the time, allowing it to eliminate time-consuming inputs from the sales force.
Make more-informed decisions by applying FVA analysis as detailed in the book The Business Forecasting Deal: Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions by Mike Gilliland, SAS Product Marketing Manager for Forecasting. The International Journal of Forecasting said in a recent review that “Gilliland supplies lots of simple and easy explanations and examples, which will be very helpful when (forecasters deal) with non-technical colleagues or management and their possibly unrealistic expectations about what a forecast can do.”
Set realistic expectations – and get a better forecast – by learning more. Explore ways to improve forecasting in this white paper: What Management Must Know About Forecasting.