Ok, So say I want to measure the real impact of some marketing on an isolated day/week.
Now because marketing is switched on and off it makes this harder, so what I was thinking is looking at a year on year difference and a versus the previous 8 weeks average difference, as a %.
This week we were say 85% versus the previous 8 week average of this year.
But that 8 week average was up 9% versus the same 8 weeks of last year.
YoY this week was up 75% versus last year, but the same week last year was 15% up versus its prior 8 weekly average.
I'm going round in circles here, thinking one of these numbers minus the other is the real impact. I'm just trying to make everything as comparable as possible to get a real idea.