The forecast of demand forms the basis for all strategic and planning decisions in a supply chain. Throughout the supply chain, all push processes are performed in anticipation of customer demand whereas all pull processes are performed in response to customer demand. In either case, the first step is to forecasting what customer demand will be.

With the importance of demand forecasting in mind, we now move to the most relevant characteristic of forecasts:

• Forecasts are always wrong. :-)

That being said, how does one measure the accuracy of a forecast given that the following error measurements exist?

• Mean Square Error (MSE): This can be related to the variance of the forecast error.
• Mean Absolute Deviation (MAD): The average of the absolute deviation over all periods.
• Mean Average Percent Error (MAPE): The average absolute error as a percentage of demand.

Selecting the right error measurement depends on what aspect of the forecast is most important to your organization and what your data looks like. For example, if your data consists of very small measurements, then even a small absolute error will show a large percentage which can be deceiving. On the other hand, for very large numbers, the percent error can be very small while the MAD could look very large. A lot of this comes down to a judgment of how people perceive numbers and whether they can put these values into perspective. Assisting people in understanding these values is as important as the calculations themselves.

I hope this edition of the Numerical Insights newsletter has provided a bit more information to assist you all in your quest for data-driven decisions. As always, I welcome input from my readers and requests for information.