After doing more reading and experimenting I have found one thing, the egg (market need) does come first.
Just like in the scientific method everything starts with a problem, then you have a hypothesis, then you do an experiement, then you draw conclusions and repeat.
- Problem (ask the market if this is a problem)
- Hypotheis (test possible solutions with potential people in the market)
- Experiement (you build something and go private beta)
- Conclusions (you learn from experiment modify and repeat)
A great way to test the market to see if it has an idea is to run an adwords campaign to see if anyone is interested in your product/solution. You might need to run a number of campaings to see if you are correctly undertsanding the problem. You might also need to part with a free-giveaway to see if your idea has merit, people love contest and usually these "triggers".
Also I am not sure of other's success, but avoid the content network, I have my lowest scores here all <1% but the number of impressions there is 95% of the total.