Fundamental Forecasting Metrics

Part of the Platform Imperative Webcast Series

Traditional forecasting performance metrics, such as mean absolute percent error (MAPE), reveal the magnitude of your forecasting errors. However, MAPE does not tell you what degree of accuracy you should be able to achieve given the nature of your demand patterns. Similarly, MAPE alone gives no indication of whether your forecasting process is adding value by making the forecast better.

Join Mike Gilliland, Forecasting Product Manager at SAS, to learn more about the alternatives to traditional forecasting performance metrics, as well as how these alternatives can provide new insight into both the “forecastability” of your demand patterns and what the current process is achieving. During this on-demand Webcast, he discusses how forecast value added (FVA) analysis can help you measure the impact of each step and participant in your forecasting process. Webcast topics include:

  • Traditional metrics (including MAPE and bias).
  • Alternative metrics.
  • Relationship of demand volatility to forecast accuracy.
  • FVA analysis.

Gilliland is joined by Brenda Wolfe, Product Manager at SAS, who will demonstrate the performance metrics currently available, the stored process for forecast accuracy and forecast value added analysis, along with the graph of volatility, which will cover coefficient of variation versus forecast accuracy.

If you have any questions, please contact Natasha George at Natasha.George@sas.com or (919) 531-0493.