[chandika]

The key is to find customers first. If you have a hunch that something should work, there is always a possibility that you can uncover some pain points that you can build a business around.

You need to speak to enough customers and iterate on what you offer to figure out the validity.

Coding stuff up is always a lot more comfortable, and a no-touch approach like adwords based validation is 'easy', but really, you need to be in front of people to understand if there is any validity to your thinking.

Open ended conversations around the problem area usually results in great insight and can lead to some valid ideas you can work off of. Also, speak to more than one person in a given problem area and it's not too hard to ask 'do you know anyone else that has this kind of problem?' to find a common theme and get an idea of a market size thats worth your time.