Hi there, here's some unsolicited advice from someone who's not much unlike you -- I quit employment about a year and a half ago when I had a fair chunk of cash, and now I'm launching startups for myself.

I've built about 5 projects to "almost completed" status in the last year. And each project has been "re-built" at least twice. It sounds like I'm in a very sad state, no better off than you, but my friend, I have fantastic news! I'm running out of money! There's tons of pressure now! And because of the accumulated pressure I've been feeling in the last 2-3 months, I finally put up one of the projects I'm working on, and after a stint of work today, it should be ready to accept paying customers (even if it's not fully glossy and conversion-tuned yet).

My advice to you? Find ways to put yourself under pressure. If it's not financial strain (like I have no choice in), then make it peer pressure and social pressure; Put up a website for what you're working on, and start talking about it. Post it on your Facebook, to HN, and anywhere else you feel like talking about it. Pray to a god-you-may-not-believe-in that people give you negative feedback, because that's the best motivator. "You can't do it, you're gonna go broke, you're doomed to fail!" is what the little man in my mental coal-room is shouting as he whips the hamster that powers my typing fingers.

(Not Mike, but I can relate) - I suffered from a similar problem some time ago, that I have since overcome. I'm not sure if I can offer any words of advice, but I can give you a recount of my experience that helped me overcome it. Like you, I would often start projects, leaving them unfinished after the interesting part. After noticing this trend, I stopped and did a little introspection. I thought long and hard about what I want to do in an ideal world. Do I want to be a corporate citizen my entire life (no harm in that) or do I want to do my own thing? I determined that I eventually want to be my own boss and try to build things that people (and I) care about. Although apparent, I had to realize and implant in my head that nobody was going to do anything for me. If I want to be successful, I'm going to have to bust my ass and do it on my own. I was going to have to make sacrifices and modify my behavior. I had to exercise self-control and only focus on one project at a time (I would tend to bounce back and forth between ideas). I had to treat it like a job and set up expectations and deadlines. I had to set up tasks and goals and deadlines, because flying by the seat of my pants hadn't gotten me very far, and sure as hell hadn't allowed me to finish any projects. So, I guess, what it boils down to, is that I had to make a behavior modification and realize that I have to do it on my own, and that I can gripe and groan and read about other's success on Hacker News all I wanted, but if I wanted to be successful, it was all up to me.

I thought this was a REALLY great question. While I think all of the advice already given is important to consider, I have a slightly different perspective: http://nobulb.com/2011/08/just-do-it-and-actually-finish/

Thanks for taking the time to write that up Mike. It is a new perspective that I hadn't considered.

Reading it made me realize something: On most of my projects, I am more concerned with how others will view it. I see developers with several successful projects and think to myself 'I want that, I want to show people I can do that too'. That has enough force to get me to start ('Look at me, starting another cool project!') but not to see it through because I am only personally interested in a subset of the problem.

Maybe I need to re-order my process. `Start -> Finish -> Tell` instead of `Tell -> Start -> Finish`.

Best of luck with your online persona project and thanks again!

You're not finishing projects because your projects are too big. Choose smaller projects.

It's a lot faster to plan than it is to code. Before you start coding, plan. Try to figure out the Minimum Viable Product.

If you want your project to get used by others, try to verify that people actually need your project before coding.

You can plan and verify need much faster than you can code. After planning and verifying several projects, pick one that is very simple and do it. If you are unable to complete it in time, you messed up your planning.

+1 - I could use some advice in this area too ...