I am currently working on project that requires forecast at an SKU / Item level.
I have past 12 months sales as well as distributor inventory opening for current month.
I would like to forecast demand on an SKU level (I have the SKUs listed by EAN Code for each market), taking into account P12M of sales & current inventory opening.
The data is presented in the image attached.
Since the EAN codes are NUMEROUS, around 181 EAN codes per market, I am struggling with forecasting for each of these EAN Codes separately. Is a loop required? Can it be done through knime to predict the forecast?
Forecast should present a logic of: If Inventory opening divided by P12M of sales * 30 days, is less than 120, increase forecast to reach 120 days. If it is more than 120 days, reduce forecast for the next month to reach the 120 days.
The target is to ensure distributor is landing at 120 days of coverage for inventory to avoid over or understocks.