SkuBrain now incorporates two new features that should help when you’re building hierarchical forecasts, particularly if you
plan to use these forecasts to create Stock Plans (for inventory planning decisions).
- You can now create a Stock Plan from any forecast that has the SKU field in its hierarchy
- When building your hierarchy in the forecast wizard, SkuBrain automatically corrects crazy hierarchies
Forecasting by Outlet
Previously, if you wanted to use a forecast for inventory planning in SkuBrain, the last item in the forecast hierarchy had
to be the
Sku field. So although you could create a forecast with a hierarchy like
Department | Sku | Outlet, you
wouldn’t be able to use that forecast as the basis for a SkuBrain Stock Plan… which is a pity, since it would be nice to
aggregate the sales of a particular SKU across all of your Outlets for forecasting purposes but still make Outlet specific
inventory replenishment decisions.
Now you can!
Of course you don’t have to… hierarchies like
Outlet | Department | Sku or
Department | Outlet | Sku will still work
Automatic hierarchy correction
Some hierarchies simply don’t make much sense. I’ll give a detailed example to show why.
Imagine you have sales for the following SKUs:
When you create a hierarchy like
Department | Sku against the above data, SkuBrain prepares 3 separate groups of
forecasts, resulting in 6 forecasts in total:
Level 1 – Root forecast
- Root forecasts (using the aggregate sales of all SKUs – 29 units)
Level 2 – Department forecasts
These are the forecasts for the sales in each separate department within the root level forecast.
- Forecast for Footwear (using aggregate sales of all SKUs in the footwear department – 22 units)
- Forecast for Clothing (using aggregate sales of all SKUs in the clothing department – 7 units)
Level 3 – SKU forecasts
This group includes forecasts for each individual SKU in each department.
- Forecast for Footwear –> blue-shoes (12 units sold for this specific SKU in the footwear department)
- Forecast for Footwear –> black-shoes (10 units units sold for this specific SKU in the footwear department)
- Forecast for Clothing –> blue-shirt (7 units units sold for this specific SKU in the clothing department)
On the other hand, if you created an upside down hierarchy like
Sku | Department, SkuBrain would generate
the following groups of forecasts:
Level 1 – Root forecast
- Root level forecasts (using the aggregate sales of all SKUs – 29 units)
Level 2 – SKU forecasts
These are the forecasts for each individual SKU within the root level forecast.
- Forecast for blue-shoes (12 units sold for this specific SKU)
- Forecast for black-shoes (10 units sold for this specific SKU)
- Forecast for blue-shirt (7 units sold for this specific SKU)
Level 3 – Department forecasts
These are the forecasts for the sales in each separate department for each of the SKU level forecasts.
- Forecast for blue-shoes –> Footwear (12 units sold for all units of this SKU in the Footwear department)
- Forecast for black-shoes –> Footwear (10 units sold for all units of this SKU in the Footwear department)
- Forecast for blue-shirt –> Clothing (7 units sold for all units of this SKU in the Clothing department)
See the problem?
The last three of these forecasts don’t make any sense at all. Any given SKU can only ever belong to one Department,
so grouping first by SKU and then by Department simply creates a bunch of redundant department forecasts which will be
duplicates of the parent SKU level forecasts.
As such, if you try to add an attribute such as “Department” below the SKU in your forecast hierarchy, SkuBrain now
automatically shifts these attributes further up in the hierarchy so that you have a hierarchy that makes sense.
In fact, the only attribute which SkuBrain will let you add after the SKU in your forecasting hierarchies is
this being the sales channel or physical outlet that you sell the SKU through (maybe you refer to this as a location,
depot, store, warehouse or something else in your organization).
Lie down hierarchy
OK, enough about hierarchies – the next post will be about something completely different!