is immediately getting more complicated—but it does not have to! Do you know the Greek Alphabet of Supply Chain? Below, we have summarized what the alpha, beta, and gamma values represent in a forecasting model:
Alpha- Alpha is used for single exponential smoothing, which represents smoothing for level demand. It is referred to as the smoothing factor or smoothing coefficient. The value can range from 0 to 1. If you have a higher value, the more recent months of history in your product impact the model more. The smaller the value, the more history that is considered to produce the forecast.
Beta- The beta value is used in double exponential smoothing. Beta represents smoothing for trend and works in tandem with alpha in a double exponential smoothing model. Commonly, you will hear the Holt’s model being referred to when using a double exponential smoothing model. Like alpha, the beta values ranges from 0-1. An important note about alpha and beta is that they are independent of each other. This means that the sum of the two values does not need to equal 1. You could have an alpha and beta of .6 if you want or an alpha of .3 and a beta of .8 for example.
Gamma- Gamma is a value used to smooth for seasonality. Charles Holts and Peter Winters designed this method of triple exponential smoothing often referred to as the Holts-Winters model. The gamma value can also range from 0 to 1 and is independent of alpha and beta.
Now that you know the Greek Alphabet of Supply Chain, how will you use it?