-
Notifications
You must be signed in to change notification settings - Fork 10
Visualization
Spencer Buja edited this page Jul 6, 2019
·
3 revisions
An assumption for this analysis is that the "key performance indicators" (KPIs) should have trended the same, if the users did not start using your product. One way to check this assumption is by observing historical data between the two groups over time. There can still be a delta between the two products, but the lines must be parallel.
Is there a measurable improvement in the KPI that I'm intending to drive?
This visualization shows that the KPI improved for users of my product after they were exposed to the product in comparison to the control/comparison group. We should take a look at the printed out statistical estimates to test for the signficance of this improvement.