This week I conclude the allocation series by discussion how data at the intersection of cost and target data causes an explosion in data volumes in allocation processes.
Last week we discussed this diagram in talking about allocations:
One can see how detailed data needs to be accumulated, both costs and activities, to created the basis upon which the allocation is done. And then using that basis, detailed activity data is again passed in order to create the detailed allocation results.
If one desires to create different cost “pools” (the amounts to be allocated), and allocate those costs over multiple sets of drivers, it is easy to imagine how data volumes can become large quickly. It is a multiplying effect for each dimension of the allocation to be performed.
Aggregation can be used somewhat, but the more attributes that are aggregated out of the activity data, the fewer dimensions are available for reporting purposes.
The concepts I’ve talked about for a metric engine can be useful in managing these types of volumes in allocation processes.
This is the last video in the allocation series. The previous episode is Allocation Processing Steps
Watch all episodes in order at the Conversation with Kip Playlist.