stats from the first traffic source - felt the numbers were big enough to be relevant. No other combination showed any kind of recognizable pattern. SO I felt pretty confident starting a new campaign at a different source already selecting the manufacturer/OS combiantion that were proven converters .
(Un) fortunately I messed up setting up the campaign - included all OS - and woke up the next morning only to find that here the results were the exact opposites - whatever converted on the first source didn't here and vice versa.
SO when it comes to scaling I am now ' gun shy' again - thinking that I will have to test all the parameters again. Oh well - all this ain't nothing but a thing ... but the lack of volume I find is getting to me ... ( I mean that in a encouraging .. getting mad and more determined way - I can't stand if I am not good at something )