Attempting to import historical supply forecast data into D365 supply planning module but getting errors about invalid forecast periods. Our legacy ERP used weekly buckets while D365 is configured for monthly periods in the master planning calendar. The import validation shows:
Error: Forecast period '2024-W23' does not exist in planning calendar
Entity: SupplyForecastEntity
Record: SKU-10234, Period: 2024-W23
We have about 18 months of historical weekly forecast data (roughly 75 weeks × 2,500 SKUs = 187,500 records) that we need for demand planning baseline analysis. The planning calendar in D365 is set up with monthly buckets starting from fiscal year boundaries. Do we need to aggregate the weekly data to monthly before import, or can we modify the planning calendar to support weekly periods without breaking MRP runs?
One gotcha with forecast period alignment - if your legacy system’s week definitions don’t align with D365’s fiscal calendar, you’ll need logic to split weeks that span two months. For example, if week 23 runs from May 28 to June 3, you need to proportionally allocate the forecast quantity to May and June based on the number of days in each month. This gets complex quickly, so document your allocation rules clearly.
You can maintain weekly forecast data in a custom table for reporting and analytics purposes, while importing aggregated monthly data into the standard SupplyForecastEntity for MRP. Many organizations do this - keep detailed forecast history in a separate data store (even Power BI dataflows work well) and only push the MRP-compatible monthly aggregates to D365. This gives you the best of both worlds: detailed analysis capability and MRP compatibility.
The planning calendar configuration is fundamental to MRP execution, so you can’t just switch to weekly periods without significant impact. D365 Master Planning expects forecast data to align with the calendar bucket definition. You’ll need to aggregate your weekly data to monthly buckets before import. Create a transformation that sums weekly forecasts into the corresponding month based on your fiscal calendar start dates.
That makes sense, but we’re concerned about losing forecast granularity. Our demand planners rely on weekly patterns for seasonal products. If we aggregate to monthly, won’t that smooth out important week-to-week variations? Is there a way to maintain weekly detail for reporting while having monthly aggregates for MRP?