Randall Bennett suggests, among other things, that you should not feel bad about working for what he calls a "no-shot startup" (one where inexperience meets enthusiasm and results in some kind of startup disaster), because you will still learn from those, and:
Crucially, the biggest advantage of working lower down the spectrum is that mistakes don't stick with you. In general, mistakes don't typically stick with you, but the further up the spectrum you go, the tighter knit the community. Make a mistake at the bottom of the spectrum, and there's enough people making mistakes that it's unlikely your mistakes will give you a bad reputation. On the other hand, screw up a company with $41mm in funding, and those mistakes are more likely to follow you.
That's a fair point. Conversely, I expect that most investors with $41m to swing around won't invest in a team that hasn't cut their teeth on previous ventures. And in fact, they didn't, since the colors.com team, to take the example Randall presents, is actually pretty solid and experienced.
Randall adds that after starting at the bottom, once your first hopeless venture dies out, you should work at moving up the ladder, into more and more successful startups.
I think there's a very valuable further point to make.
Once upon a time, MBAs used to be designed for people who had 5, 10, or more years of business experience, to enable them to formalise and structure their knowledge of what makes a business tick. This was before the trend became to do an MBA 2-3 years out of university, or, god forbid, right afterwards.
The key point there is that until you have some of your own experience to drawn on, most of the things taught in an MBA won't stick, because deeply, viscerally, you won't understand why they're important.
The same is true for startups, but in reverse. Until you've worked (either as a founder or as a very early employee) in a broken startup, you won't know, deeply and viscerally, why the things that successful startups do matter. There are many lessons that you can only understand by contrasting them with the failure case. That's when the insights happen... "Aha! That's how you're supposed to do that."
In short, breaking your teeth on a "no-shot startup" before joining a successful one will help you make the most out of your time at the latter.
If you read this far, you should follow me on twitter here.