- if you want to forecast variables using historical averages e.g. average your new users for the last 3 months of actuals/historicls.
- or for having a different formula for your forecast months vs. your ‘actual’ or historical months.
- e.g. for a Churn % variable, you might want to calculate the actual churn percentage in historical months, but for forecast months this might be an input based off of other variables.