Forecasting for offline machines is based on historic service data, which can be enabled and adjusted in your VMS from Company Settings > Route Planning > Number of Historic Services for Forecast (Offline). The number that you set will be the number of past services the Forecasting uses to predict the machines' inventory needs.
Offline Forecasting is calculated by the amount of product a driver indicated as Stocked averaged over the number of days between services. The Stocked product can be viewed in the Machine View > Sales tab, seen below.
As an example, if 8 Jalapeno Cheetos were stocked, and it has been 4 days since it was last serviced, then the VMS predicts the machine sells approximately 2 units per day. This consumption rate is then averaged over every service that falls under the number of historic services used (see above in Company Settings).
When you plan a route for the offline machine, the system will predict the number of units needed to fill to parlevel based on how many days since it was last serviced (DSLS) x the product's daily consumption rate (DCR).
Please note that offline forecasting is a prediction based on past servicing data. It is not an accurate substitute for live DEX data. Irregular sales will not be accounted for using historic data.