Demand Forecasting - Part 2

There are certain characteristics of demand forecasting that supply chain managers should keep in mind:

- Forecasts are always wrong. Forecasts should include both the expected value and a measure of forecast error. If we forecast sales of a certain SKU to be between 100 and 1900 units, and another forecaster estimates sales to be between 900 and 1100, both forecasts expect average sales of 1000, but the sourcing strategies for the two scenarios would be very different.

- Long term forecasts are usually much less accurate than short term forecasts. This is why like to "tweak" longer term forecasts with monthly and/or weekly planning as we develop sourcing and production schedules.

- Aggregate forecasts are usually much more accurate than disaggregated forecasts. It is usually much easier to forecast the annual sales of a company than it is to forecast demand for a given product, or even annual demand at the SKU level. The greater degree of aggregation, the more accurate the forecasts.

Tomorrow we will review how to deal with forecast error, components of a good forecast, and the benefits/limitations of various forecasting methods.

Synchronous LLC is committed to maintaining a continuing dialogue on operational excellence best practices. We are no longer providing on-site consultations, and are unable to take on new clients. To pose a question, post a best practice, or otherwise contribute to the dialogue send a note to webmaster@SynchronousLLC.com . To subscribe to our weekly newsletter send your preferred email contact address to Sales@SynchronousLLC.com with SUBSCRIBE in the subject line.