What would be proper way (if there is one) of using date drill in combination with semi-additive measures? Our use case considers daily user counts, for which we’d like to display the week- and month-end counts when changing the granularity to week and month respectively.
Thanks for the suggestion, it looks like it could help.
I solved it for now by creating the daily absolute change in user count, which is additive over time. By calculating the running sum of the absolute change you get the total count and allows you to properly adjust the time grain.