Sets the threshold for the identification of seasonal demand. “High” will result in more seasonal forecasts than “Low” but may come at the cost of accuracy.

For businesses that operate in a very flat forecast industry such as hardware or groceries, apply a setting of “Low”.

For businesses with loads of seasonal items, such as those selling outdoor products or seasonal sports items, apply a setting of “High”.

For the average business, a setting of “Medium” would do.

Pro tip: Leave these settings as the default, have demand planners forecast as usual, see where their forecasts are falling short, and adjust these parameters based on the finding.

The Advanced forecasting module offers the ability to detect and apply seasonality for groups of items, which can result in significantly more seasonal forecasts that aren't subject to the random variability found in individual product sales histories.

This parameter is used to specify how biased the forecast should be to assuming seasonality. A setting of “High” means the forecast will be quick to assume seasonality while a setting of “Low” will be more likely to assume non-seasonality. Below is a matrix illustrating the resultant forecast when applying the “High” and “Low” settings to two sets of sales history.

"Seasonal"-looking forecast, "Low" seasonality setting

"Seasonal"-looking forecast, "High" seasonality setting

Flat-looking forecast, "Low" seasonality setting

Flat-looking forecast, "High" seasonality setting

In simplest terms, a setting of “High” will assume seasonality, even if it’s not there, whereas a setting of “Low” will ignore obvious signs of seasonality.

Obviously it’s not this drastic in the app. The above is just for demonstration purposes.