[marban]

I think first and foremost is bringing yourself into a position that gets you enough runway (time and money) to be able to take things seriously — even if that sounds like the opposite of a side project. For me it was skipping university and having kids in favour of starting a web agency in the dotcom days. Also, I don't believe in working on something you love — you just need to hate it a little less than the other ideas and be positive that someone will find it useful. Like I mentioned below, luck and timing play an important role and these days I'd say you will also need one unfair advantage (contacts, cash, distribution, etc) to increase your chances.

[mediaman]

I just wanted to say that I love your phrasing of working on something that "you hate a little less".

In general I believe the idea of "work on what you love" to be awful advice; I have found that I get deep satisfaction out of working on something that someone is willing to pay for, even if that means that sometimes it is a headache and I hate doing some of the things that need to be done. Once you accept that those things will continue to exist and continue to need to be done, and it's not going to be bliss, those things seem to matter less.

[jjeaff]

I prefer to try to learn to love something that you believe can bring you success. It is very difficult to do something day in and day out that you don't love some part of. You just have to figure out your angle. I don't love my current business, but I love business in general and I love building things that people like to use.

[cies]

The "love requirement" makes it hard to continue once you wake up one day and find you no longer love it. The "hate it a little less" is much more useful is you dont want to give up on stuff because you have a bad day/week/month with it.

[bonestamp2]

Was there a general goal for each one or any sort of decision tree you used to decide on what to build next... for example, did you only work on things that could potentially be big, or huge, etc? Did you analyze the market size of each idea before starting, etc?

[marban]

I've outlined a few things in an answer below but another aspect I use more and more these days is a negative inbound filter for ideas. Ranging from macro problems like 'Will it require an app' down to even 'do I need user accounts'. Basically, try to put together a few things that you need or want to absolutely avoid due to certain constraints or knowledge and go from there.