When you need to create demand forecasts for lots of products, one of the challenges is being able to organize the sheer volume
of information so that you can focus on what really matters.
SkuBrain’s hierarchical forecasts help to some extent by letting you group and categorize your forecasts into the various different
product classifications that are used within your organization. However even when products and demand forecasts are organized
hierarchically, it’s still hard to work out where you should focus most of your attention.
As such, hierarchical forecasts in SkuBrain now have a “split view” that lets you see the individual items contributing to a forecast and their relative importance.
The screenshot above shows two new buttons that let you switch between forecast view and split view.
Forecast view will look familar to anyone who’s used SkuBrain before. This shows forecast details for the selected item in the hierarchy tree.
The second button (the one that looks like a pie chart) displays the new split view, which shows a breakdown of the currently
When in split view, the chart no longer shows forecasts vs. actuals and also no longer shows the baseline statistical forecast.
Instead, it displays only final forecasts. However, instead of a single series showing the final forecast for the selected
item in the hierarchy, when in split view you see a stacked chart that consists of each of the separate components that contributes
to the selected forecast for each period (as determined by your forecast hierarchy).
So, for example, the hierarchy for the above forecast is
Category -> Subcategory -> Sku, so when viewing a split view of all
sales you see a stacked chart that shows how much demand each of the individual categories contributes to total demand forecast.
Hovering your mouse over any of the series on the chart will display the cateogry name and clicking on it will open the child
forecast (also in split view) so that you can easily drill down through your forecasts, focusing on the elements that you’re
Similarly, when viewing the forecast details for a particular category, the split view shows a stacked chart illustrating the
relative contributions of each of the individual subcategories in that category.
Below the chart, in the table area, things also change in split view. Rather than seeing statistical model details, baseline
forecasts, prediction intervals and final forecasts, when in split view you only see final forecasts. However you see
the final forecasts of all of the child forecasts along with the percentage that each contributes to the total demand forecast
for the selected item in the hierarchy (in the right hand column).
Again, you can click on any of the child forecasts to drill down into this.
And that sums up the new split reports! They’re shinny new and I hope you like them, but please yell out if
you’ve got any ideas on how you think these could be even better.