Answers to Your Forecasting Questions Chaman L. Jain, Editor St. John’s University | Jainc@Stjohns.edu [ Q ] I am a Director of Sales and Operations Planning. I have a demand planner overseeing about 1,000 SKUs. At an aggregate level, our forecast bias is within ±15%, but the devil is in the details. Some SKUs have an extreme bias. Is there any tool that systematically looks at the error and reduces the bias? [ A ] There are tools that can flag the extreme errors, but not reduce them. The best thing is to find the source of bias. I am not sure whether you have a formal forecasting process in place. It could be that the computer program is biasing the numbers, or someone with an authority is doing it. If this is the case, it can be taken care of by putting a consensus forecasting process in place. Also, bear in mind, forecasts generated by the computer are nothing more than baseline forecasts. Some forecasts may require judgmental overrides. It is not uncommon for a forecaster to find certain numbers don’t make sense, either too high or too low, and thus require intuitive adjustment. Some forecasts are affected by certain elements that cannot be quantified. There are also ...

From Issue: A New Approach to Effective Demand Planning
(Summer 2014)

image
Answers to Your Forecasting Questions