So lets say you tested a bunch of offers, landing pages, and banners. After a while you finally get a banner/LP/offer combo that seems to be about breaking even over a good sample size, what is your next test?
Assuming you can't get a pay bump, what would you do here to push that ROI up?
At this point I would normally try to test new banner images, but I think this is wrong. There is a good chance the funnel is working because of the image I have chosen, same with the angle.
So what would you test next?
1. LP Headline/Images?
2. Offer Landing Page?
3. Something Else?
I know there are probably a million things you could test, but just curious in this specific situation what would you test next?
Thanks 
I would test the funnel on different traffic source first, not sure whether you are testing on just one, so my advice might be off 
Thanks Matuloo. Lets say you are on a legit traffic source where the pricing is reasonable, I guess what would you test then?
I would test the bid first.
Then I'd cut anything that isn't profitable - placements, exchanges, OSs, browsers, whatever else. (Certain things you cut, will increase the ROI for the rest, so cut the worst first, wait and see, then cut some more.)
So then I'd have something that's running and generating profits - next I would improve and scale.
I'd seek out the same offer on multiple networks, and also similar offers, and test them to see which one performs the best.
I'd take the best banners, and make different combinations of their text and images. i.e. banner 1's image with banner 2's text and so on. I'd test new banner angles. I'd test lander variations.
Once I have a better funnel doing good ROI, I'd scale to as many traffic networks as I can find.
I would then also expand to more banner sizes and more exchanges.
Hope this helps!
Amy