My story. I used to use news API services for my side projects. One day, I realized that I can build my own service that will be both, best quality (like real multilanguage support), and cheap.

8 months into it, it is my full-time job and we are releasing our first paid version (post-beta). We have 600 subscribers (beta-testers) and some people claiming they will become paid clients.

It is a huge experience in both, business and tech sides.

So, I would say the most important things are:

* forget "ideas"

* resolve your personal problem that you understand well

* always make a (side-)project with which you can charge people (making it "free" is just an excuse for not making something of good quality - 99.9% of cases)

* to the previous bullet point, you either ship a final solution that has business value for someone, or you are wasting your time (because no one can tell if it is of value or not (by paying))

Product: newscatcherapi.com