The practice of business forecasting falls well short of the potential exhibited in academic research and forecasting competitions.
Chris Chatfield* noted this in a 1986 editorial in the International Journal of Forecasting, where he called on statisticians to find a better way of communicating the better use of existing methods to practitioners. Fast forward to 2000, when Michael Lawrence** penned another editorial in the IJF, lamenting the lack of progress toward Chatfield's goal, and deploring the lack of research examining the causes. More recently, Morlidge's studies (cited frequently in this blog), show businesses in many cases still failing to achieve even the lowest benchmark of forecasting performance (the naive "no-change" model).
How has this problem persisted for over 30 years with little evidence of progress? Is now the time to solve it?
The UFO Project
This generally poor Usage of Forecasting in Organizations has lead a team of academics and practitioners (listed below), to embark on the UFO Project. Our objective is to improve the practice of business forecasting by encouraging adoption of systematic forecasting in organizations that now employ only ad hoc methods.
By "systematic forecasting," we mean the use of appropriate quantitative methods when suitable data are available, while allowing for judgmental inputs and adjustments that are supported by a documented and defensible rationale. Where little or no data are available, such as with new products, our definition encompasses structured management judgment including use of intention surveys, decision aids, Delphi procedures, and others.
Foresight: The International Journal of Applied Forecasting is providing free access to our new article describing the issues and challenges of the business forecasting environment. You can down download it here: The Benefits of Systematic Forecasting for Organizations: The UFO Project.
Our next step, to better understand why many organizations do not exploit the advances in forecasting knowledge and technology, will be a survey and interviews with business forecasting practitioners. Are you interested in participating?
If willing to take the survey or be interviewed, please contact me directly at mike.gilliland@sas.com. We'd love to get your insights.
The Project UFO Team
- Spyros Makridakis, University of Nicosia
- Ellen Bonnell, Consultant
- Simon Clarke, Crimson & Co.
- Robert Fildes, Lancaster University
- Michael Gilliland, SAS
- Jim Hoover, University of Florida
- Len Tashman, Foresight editor-in-chief
-------------
*Chatfield, C. (1986). Simple is best? International Journal of Forecasting, 2, 401-402.
**Lawrence, M. (2000). What does it take to achieve adoption in sales forecasting? International Journal of Forecasting, 16, 147-148.
