This is a little of a spinoff from the question I previously posted on my thread and thought this deserved it's own.
I hear its pretty standard to test 5-10 offers/geos in a traffic source, find the ones that bite and scale it to the rest of the sources you have.
What I do is that I'm currently taking a geo, find similar offers across different networks, and launch it on 5-10 sources from the get-go and find green spots. Starting to learn so many new things and realise this might not be the right way.
With that comes a few questions:
1. For the former, if an offer that's not working well in traffic source A, might be working in B,C,D,E... Wouldn't this cause the issue of a single point of failure?
2. For the former, how do you determine which traffic source to use to start off? I've had the same exact lander/offer fail miserably on traffic source 1, and did well on 2, so this gives a huge dilemma.
3. I'm guessing there's no hard and fast rule in AM, but from the 2 different approaches above, is the former the better approach?
Interesting question!
First things I want to say, is that based on personal experience, it's rare for a campaign (offer+creatives) to be profitable on one source and not even convert on another.
So, one approach may be to use 2 traffic sources that you've had consistent success with, do the initial "test-the-waters" camps there, and only if a camp shows promise on at least one traffic source then continue optimizing.
In AM we're forever faced with having to choose a balance between 2 extremes. Testing broader will give us a better chance of finding gold, but will also cost more. So we find an acceptable balance that feels right to each of us.
Would love to hear what the more experienced folks have to say about this. 
Amy
First of all do some reading. You're paying 99$ a month to being on here so I'd read through everything. Then pick 2-3 verticals you want to work in and test them out on 2-3 traffic sources. Test at least a 100$ per campaign in order to get some valuable data.