I also think the 90% figure is misleading - but I think it's too optimistic. The number of people that quit their job and use their savings and/or family-friends-fools money to start working on a startup and never get to a point where anyone knows they even exist is huge. These people are never counted in any of the official statistics, and don't talk much about their venture because it's never nice to talk about your failures.

If you include this very large shadow group I'd argue that less than 10% even make it to ramen profitability. You never hear about the failures - only the successes. This skews the picture.

I read the other day (HBR?) that 85% of venture-backed startups are dead within 3 years. Say what you will about VC firms, but the companies they fund tend to be a notch above the mean. And they are certainly the types who get (and maybe even ask for) advice. Creating a business (the kind that supports 2+ founders) is HARD.

Thinking your chances are better because you're (by your OWN assessment) smarter, more open to advice, and more hard working than the mean is a dangerous attitude. If you walk into a room and say, "raise your hand if you think you're an above-average driver", pretty much every hand goes up. 50% of them are wrong.

Normally I love Gabriel's posts, but I think this one smacks of Mr. Rogers ("you're special because of who you are").

all the data on this is fuzzy, and may always be fuzzy. what type of startup it is is important, as well as the quality of the founders, none of which is easily quantifiable. I think the reality is that you make your own odds. a brilliant/determined/flexible founder can and will make something work.

I've always felt that what kills a startup are commitment/flexibility issues based on factors outside of the business itself - your spouse doesn't support you, you're only willing to live on peanuts for so long, children on the way, etc.

but I don't look at startup success as determined by Idea X, sink or swim. If you stay in the game long enough, and find a way to feed yourself in the meantime, you'll find a way.

also, as in life, you alone determine what is and isn't success. for many people, a lifestyle business that just allows you to work on your own terms (not too far past ramen profitability) is enough.

I see this discussed now and then, and I never really see the point. I don't care what a startup's chance to succeed is, I only care what MY startup's chance to succeed is. And since these estimates vary so widely even when they're completely general, they're not even that useful as a baseline.

Intuitively I would tend to agree, but being a physicist I would like to see some more quantitative data on this.

Both the links from the article address different sets of companies than what he is actually talking about, the one on tomuse.com is about the companies on that famous graphic, which of course had already gained a certain amount of traction, otherwise they wouldn't be on there. The other one covers a vast superset of that, and also covers a much larger timescale than what I imagine when I hear the 90% figure.

how do you define success?

I'd say even "breaking even + $30K/yr" does not apply to 90% of websites/web businesses.

"You're doing a software/Internet startup with low marginal costs."

I know this is where a lot of the media focus is on software and internet. But I think there is still a good number of service or product related start-ups, although I don't have actual numbers.

Same kind of misleading advice old folks tell the young to study hard, get good grades, work for a good company, and blah blah blah ...