Forecasting Research Survey
The Centre for Marketing Analytics and Forecasting at Lancaster University is conducting a new study on how demand forecasters do forecasting. They invite all practitioners who are working with forecasts, such as demand planners and supply managers, to take part in the study. The survey is anonymous and will take 10 minutes.
Insights from Academic Research on Forecasting
A significant portion of academic research on forecasting deals with statistical models and methods for generating more accurate forecasts. I have referred to this as the "offensive" approach to forecasting, where you are trying to make good things happen -- more accurate forecasts. It is akin to playing offense in sports, where you are trying to make good things happen -- score more points.
However, the BFD blog is more about a "defensive" approach to forecasting. This is akin to playing defense in sports, where you are trying to prevent bad things from happening (your opponent scoring points). The defensive approach aims to eliminate the bad things -- ineffective and wasteful forecasting practices that fail to improve the forecast or even make it worse. (For a thorough discussion of the offensive vs. defensive approaches, see this 12-part blog series published in 2016.)
Previous academic research, out of Lancaster University and elsewhere, has examined some of these defensive issues -- such as the prevalence (and frequent lack of efficacy) of manual overrides to statistical forecasts.
As a practitioner, you can play an important role in the latest research by participating in the survey. Sharing the issues practitioners face will help researchers (and also forecasting software vendors like SAS) improve existing forecasting systems and process.
Free Practitioner Workshop in London (April 13)
Lancaster's Centre for Marketing Analytics and Forecasting is organizing a half-day event in London on April 13, 2018. The event brings together well-known practitioners and academics to present and discuss new insights and approaches to forecasting using hierarchies.