This forecast method is tailored for parts, including fast-moving parts, with intermittent or sporadic demand. It creates a smoothed forecast based on historical demand and the application of the smoothing factor α (alpha). It is a combination of the Crostons and Single Exponential Smoothing forecast methods.

When demand is intermittent, this method generates a modified Crostons forecast, and when demand is not intermittent, this method behaves much like Single Exponential Smoothing.

Use this method as the default forecast method for those streams where a lack of data does not allow for the use of Best Fit.