At least you tried. That is a good start. It's normal that startups/side-projects fail fast, the successful ones are only few of them, so keep trying and think more about what you will be building.
Here is what I did:
1) build something useful, at least, to myself. Even no one uses it, it's ok.
2) try not to spend too much (time, money...) on the project unless it's needed. ex-prj.com is powered by tumblr. I couldn't believe that I can achieve that without coding.
there are some apps by which you can build some mock apps. You can show your mock apps to friends to get some ideas. If enough people say it's not that great, probably you can stop.