Starting isn't the hard bit, I've started 50+ projects, I've finished hardly any of them, not including client work, that's my job, so it's easier to finish.
What's different with a project that's mine is I go all gusto for a while, get a half MVP up and kinda lose the will to carry on. The problem is I have too many ideas and don't stick with my own things to see them through.
I'd probably be more successful if I hired someone else to build it, and get client work myself to pay for it.
Maybe you could work on your own projects as if it is a client, i.e. take 2 hours out of your 40 hour billable work week and deliver as if it is a client.
To make it even more real you can take some savings, draft a contract with someone you trust and tell them to only pay the money back as you deliver the 'client work' to yourself. If you fail to deliver, the money goes to charity.
Also, once you have some revenue from your own app, divide by your hourly rate to know how many hours you will spend on it per week.
Although I'm not in exactly the same boat (i.e. I do finish projects but I'm slow), I'd love to be a member of such a group: a mutual accountability forum or weekly lab-meeting where everyone declares what they've done this week.
I know that the Micropreneur Academy has such a forum as one of the attractions, but I think it can be open and free for anyone who's not too worried about their ideas getting pinched (speaking for myself, I wouldn't be worried).
Hey Jon, yeah i have a kind of support group, its a small IRC room i run that a few regulars hang around in, you're welcome to come in. www.chatwebdev.com
Try releasing early. When I work on personal projects and drag them out to perfection or until they have every feature I originally envisioned, they never get finished. I redesign them half a dozen times throughout their development, recode bit and pieces, etc. However, when I hack together a project over a weekend or week, toss it online and it gets a little traction, it's good motivation to continue. You have users behind you providing feedback and ideas, and it's fun to push out little updates every few days when you have people excited to see them.
At the same time, it prevents you from becoming too involved in a failed concept. Release quickly, if there's an audience, keep polishing the project and grow it over the coming years. If the interest just isn't there, well you wasted a week, learned a few things, and move on.
What's different with a project that's mine is I go all gusto for a while, get a half MVP up and kinda lose the will to carry on. The problem is I have too many ideas and don't stick with my own things to see them through.
I'd probably be more successful if I hired someone else to build it, and get client work myself to pay for it.