Currently within Analysis Workspace, there is the ability to use the "cumulative" function to get a running total of a metric, such as unique visitors, over a period of time and divide within that period of time, such as by month. However, this does not take into account that several of those visitors are the same visitors, returning from one month to the next.
Request is to have an option to deduplicate the grouped results over time so that it would be possible to have a running total of values that takes into account de-duplicated data, so that rather than adding up to the individual totals from each month, the total across the time period with this metric would potentially be less than the total by month.
Unique Visitors (Base Metric)
Unique Visitors (Cumulative function as it exists today)
Unique Visitors (Cumulative Distinct proposed metric across entire time period)
230 (20 of these visitors overlap with ones we've already recognized from the January row)
400 (30 of the March visitors have previously been recognized from previous months)
There is currently no easy/direct way to produce this sort of reporting, and it would go a long way toward seeing true over-time accumulation of visitors (or other items viewed in a similar manner.)