It's hard, damn hard. For a quite a while I was lucky enough to have a very boring, mundane, easy $DAYJOB where I could go in, slog through fixing a couple of bugs, and go home, retaining most of my creative, emotional, and intellectual energy for the startup. By my current $DAYJOB is a lot different... more challenging, but also more fun. But it really makes it hard to balance things.

I wish I could say I had some real solid advice for you, but all I can really do is share some anecdotes in the hope that you can glean something useful from that.

Basically, I try to compartmentalize my thinking as much as I can. During the day, at the 9-5, I mostly try not to let myself get drawn into thinking about "startup stuff", except during lunch. I treat lunch time as "startup time" and I often leave the office, go to a restaurant or coffee shop, and do startup work for an hour there. The delineation isn't perfect though, as I do keep a browser tab open all the time, with Gmail for my startup email address, and I do answer and send a few emails or whatever, throughout the day.

I might also, rarely, pop into the wiki for the startup and make a few notes about an idea that occurred to me, or a link I found or something.

Then, after work at the $DAYJOB, I immediately go back into "startup mode" most evenings. I leave the office, drive to a Barnes & Noble or something, sit in the cafe with my laptop and get back to work. Since I know I have "startup time" built into the evenings, it reduces some of the temptation to think startup stuff while I'm supposed to be working.

Now the fortunate thing is, my $DAYJOB employer knows about the startup, is an entrepreneur himself and is generally supportive. As long as I'm getting my work done and not letting things slip, he doesn't mind if I leave work early on occasion, or take a long lunch now and then, to have a meeting related to the startup, or whatever. I will also sometimes just flat out take a vacation day, to accommodate the need to do something for the startup.

I also allocate almost 100% of my weekend time to working on the startup. So the toughest day to stay focused can be Friday, since I'm mentally preparing to shift gears into startup mode for the weekend, and as Friday winds down, I might start drifting that way a bit. Luckily, Friday afternoons here are usually pretty relaxed.

The toughest part, really, is dealing with meetings during the day. Writing code, market research, writing, etc. I can do anytime. But meeting potential customers, investors, partners, etc., is a challenge since my daytime availability is somewhat limited. Basically I just try to schedule meetings for either A. over lunch (and just take a slightly long lunch), B. as early as possible in the morning (so that, at worst, I come in a little late), or B. as late in the evening as possible.

It's tough to balance both, but it can be done...

I did this for three months with Geddit, but mostly on creating the MVP with my cofounder rather than sales.

Unlike mindcrime, my employer did not know what else I was working on, and I did no work on my startup during 9-5 business hours. I had to keep everything absolutely separate - not so much as IMing my cofounder about the startup during working hours. I did keep a personal journal a work where I would write down ideas and other startup related things, as well as general thoughts, but that was it.

The 9-5 job was mind numbingly boring, and occasionally pretty stressful. I developed some tactics to make sure I kept going with the startup while still doing my 9-5 well. Firstly, I developed a routine together with my cofounder, and stuck to it religiously. 9 to 5, we were at our day jobs. 6 to 7 we were at the gym. 8pm we had had dinner and were ready to work on a weeknight. Work had a midnight curfew, no matter what. This helped to establish compartmentalised parts of the day, as well as an effective transition between one and the other. At my 9-5, I could focus on my job since I knew no startup business would enter. Weekends were devoted entirely to startup work, or leisure activities about twice a month.

Secondly, we established a goal where we could switch to working on the startup full time. For us this was saving the money needed to move to Berlin plus 6 months runway. Having this in our sights made the harder parts of the three months more bearable because we knew we were working toward a determined fixed goal. We were able to execute on our MVP through this sort of discipline and successfully switched to working full time on Geddit.

This stuff was very tough, but I had the support of my awesome cofounder and we had each other's backs. If you're going at this alone your mileage may vary, but we managed to succeed on the 'nights and weekends' route, at least for some time.

Eventually you'll have to switch to working on your idea full time - probably sooner than you feel comfortable - but it'll be worth it.

At the time I was working for a .NET consultancy firm, every morning around 6 am I was learning and coding in RoR. 1.5 h uninterrupted time every week day for 2 months makes a huge difference. Plus the weekends.

Don't forget to embrace a healthy down time. You body and mind need to relax.

I generally like my job, so it has not been too bad. But there are moments when I'm way more excited to work on my startup idea as opposed to my full time job. I think it depends on how much creativity and engagement your full time job requires. I have had to push myself away from my startup for a few weeks so I can catch up on sleep as well as focus on my creative energies on my job.

I wish I had the time mindcrime has - I have wife and kids so my job is pretty important. I find time late at night as well as on the weekends when I can.

It isn't easy, although paradoxically I prefer the 9-5 to be interesting work in its own right. It is surprising how many ideas can flow from your day job into your own projects.

I don't think it's possible. Just put 30 hours into the startup on the weekend and try not to work on it too much during the week.

If you're passionate about your startup, it's impossible.