We don't know ourselves yet, because it takes so long for startups to exit. But we have a decent idea how many are successful at the next stage after YC-- the number who are either able to raise more after YC, or don't need to because they're profitable (or in very rare cases acquired). I think for the winter 2010 batch the percentage was in the low to mid 80s.
Incidentally, I wouldn't want it to be much higher. If it was 100%, it would mean we were picking too conservatively.
Isn't 80% already a warning signal? Do you think you turn teams down because their ideas are too ambitious? Or that you tend to talk them down from crazily ambitious concepts?
(Don't get me wrong, the number of business you get on their feet is one of the things I admire about YC.)
I don't think that stat is too worrying. Remember, we're not talking about actual success yet, just that people manage to raise more money. In general the startups working on crazy-ambitious ideas do manage to. Maybe not from VCs, but at least from angels.
We definitely don't turn down people because their ideas seem over-ambitious. Nor do I think we ever talk people out of them. We like ambitious ideas. At most we get founders to think about what to launch first. But they should keep the big idea in mind-- not just as something to aim at, because investors want to hear about it.
The sort of risk I don't want to stop taking is not a risk on ideas, but on people. E.g. I don't want to stop accepting young founders, who tend to have a sharply bimodal distribution of outcomes.
In my experience, we were encouraged to be more ambitious with our ideas than we had been before starting YC, while working to launch the first version as quickly as possible.
Do you think you turn teams down because their ideas are too ambitious?
I think YC has been shifting focus to evaluating teams rather than ideas. If there is a solid team with a wacky idea, they might be accepted based on the strength of the team alone.
Tech startups are a special breed of company in that they are very inexpensive to start (sometimes all you need is a laptop, a server, and an internet connection), they can run leaner than other types of businesses (ramen, hosting, and lots of Mountain Dew), and they can scale very easily from a handful to hundreds of thousands or millions of customers (open source and Amazon AWS).
Contrast that with other business types included in those SBA figures, like restaurants and liquor stores. Those businesses start out in debt because of leasing costs, equipment purchase, payroll, licensing, insurance, etc. It is also much harder to scale a restaurant than a website or app (because of geography).
The ready availability of funding plays a role as well. The 80% figure isn't for startups that actually achieved profitability, but for startups that achieved more funding. VCs aren't as willing to fund unprofitable restaurants.
Whether that's due to real promise of future profitability, versus a funding bubble, is something to judge in hindsight. ;-)
Incidentally, I wouldn't want it to be much higher. If it was 100%, it would mean we were picking too conservatively.
You've mentioned that the classes are getting larger and there are more companies who are applying to YC when they're already profitable. Won't there be a time when this number has to hit 100% because there are more profitable companies coming in than you can accommodate in a class?
While there are always extreme exceptions (asshole founders), would YC really turn down a profitable company that had an upstanding & scalable business model (i.e. not a service business, local business, or something porn/vice related)?
A VC may balk because the company can't absorb a million or more in capital, but YC is only invests roughly 20K.
EDIT: To answer my own question, I suppose if there were enough other applicants that were for whatever reason more promising, this could easily happen.
would YC really turn down a profitable company that had an upstanding & scalable business model
I think there's your answer...there are likely quite a few applicants with profitable companies (ramen or otherwise) but that aren't scalable, likely to attract outside investment, or suitable for acquisition.
In very early stages, profitability is not the best predictor of a future large liquidity event (i.e. the thing YC wants to optimize for). Facebook and Google would been thusly rejected.
I'm curious what the months-to-profitability was at Google. They were incorporated in September 1998, and turned a $7m annual profit in 2001. Were they profitable in 2000? 1999? They don't give data going back before 2001 in their own reports (http://investor.google.com/financial/2003/tables.html has 2001-2003).
I remember a gap between when I started using Google and when ads appeared on the site. Also, it's somewhat relevant that Brin started working towards what would become Google in 1996.
I was referring to PG's post above. He said that YC has a good idea of who is successful at the next stage after YC (the number who are either able to raise more after YC, or don't need to because they're profitable). If more and more YC applicants are profitable on the way in, the number who are profitable as they exit YC (one of PG's definitions of success as described above) will necessarily trend up.
Incidentally, I wouldn't want it to be much higher. If it was 100%, it would mean we were picking too conservatively.