So lets say you launch a set of 5 banners, same design/angle but just different images.
After a while you find 2 winners. Banner A is +30% ROI and Banner B is +20% ROI
So at this point you pause Banner B and just run Banner A.
Eventually when Banner A starts dying, which is the better strategy:
Strategy 1: Pause Banner A, and start running Banner B?
Strategy 2: Pause Banner A, and start with a new set of banners, and maybe a couple weeks later when those banners die, Launch Banner B?
I guess Im wondering if people will experience "banner/angle/design" fatigue if you run 2 winning banners from the same set back to back?
What do you think?
Thanks in advance 
In this case it`s important to have plan B and C
I would run banner A, in the same time test another set of banners, when banner A die, I would run banner B and ant this point I will already have results from new test
I would do either:
A) Keep both banners running, since they are both in profit. Especially if it's in a GEO where it's harder to reach higher ROI.
OR
B) Keep banner A and test more images with the same layout/copy to see if I can beat the performance of A.
Should I decide to just run A and leave it up until the point where it's gonna start losing steam, I would definitely retest B later on, since it showed promise in the initial test.