One cannot eat the cake and have it too.

With any experiment there needs to be a control group which are part of the experiment but not receiving any of the incentives/treatments/etc

In software development it is called A/B testing, also in marketing

Graphs

  • just a time plot is not enough to find out if there is causation