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.
One thing that isn't often mention is this: the average time to liquidity for a venture backed startup is just north of 7 years [edit: make that 8.7 years].
The first "class" was 6/2005:
Reddit (made it pretty big)
Infogami (joined reddit, "made it big")
Kiko (sold for ~$250k, started Justin.tv - likely to make it big)
Loopt (well-funded, great traction, seems likely to make it pretty big)
ClickFacts
Firecrawl (one founder joined TextPayMe, another YC company that sold to Amazon)
Simmery (one founder joined Reddit and presumably made it sorta big)
Memamp
--> That's 50-75% seeming to have a pretty good shot at making it somewhat big.
I would expect the exact opposite. In 2005 many were skeptical of the whole YC model, and consequently would have had a much more difficult time attracting investors and already-successful startups. The first demo day was twelve investors, mostly PG's friends; the S10 demo day was a three-day affair with over 350 investors and members of the press. The current batch also was able to attract some startups which already had launched and had good traction before YC even started; these startups probably would not have applied to the first couple YC batches because it was not yet clear just how advantageous YC was even for startups which could raise at higher valuations.
YC's star is rising still-- the most recent "class" was pretty epic IMO. It's not a talent show-- it's a hacker-optimized machine that's always getting better at picking, mentoring, promoting, and funding teams/startups.
"Lisa Lambert, Vice President at Intel Capital, said:
It just takes a long time to go from startup idea to a liquidity event. [...] During the boom days, it was like 2.6 years to get liquidity. And today, the latest average from NVCA, the National Venture Capital Association is 8.7 years."
It's too soon to say for most startups in the recent classes. From earlier rounds, there have been successful exits like AppJet, ReMail, Reddit, Omnisio, Zenter, as well as startups that appear to be thriving like Loopt, Justin.tv, Weebly, Scribd, AirBnb, Xobni...too many to count. YC started in 2005, and it can take 5 years sometimes to see if a startup is successful, so we'll have to wait and see who makes it big over the next few years.
Judging by the success we've seen so far, it looks like YC is getting better and better each year. We'll see a few solid exits per class in the years to come, and I'll bet it won't be long before we see a massive acquisition on the scale of YouTube which will really catapult YC into the spotlight (what's 6% of $1.65 billion?)
YC won't get 6% of such a deal. A massive acquisition like that usually happens after a round or two of funding diluting the original ownership interest.
Very true...in the case of YouTube, there were two rounds of investment at $3.5 and 8 million I believe. But still, any percent of $1.65 billion is a nice chunk of change.
True, hence the qualification "judging by the success we've seen so far" at the beginning of the paragraph. Based on what we've been seeing so far, and since the classes appear to be getting both stronger and larger, it's reasonable to expect this success to continue in the future. Even more so as the economy gets better.
Do you really think that out of 25-35 per class we won't see at least 2-3 who successfully exit or become profitable within 5+ years? If the YC team didn't expect as much, they wouldn't doing this anymore.
I'd be interested to know stats about the impact of Y Combinator on companies with a nice control setup. Revealed preference: people who get in, but choose not to go to Y Combinator, vs. those who do... or something like that.
Throwaways are the HN equivalent of an AC. The difference with 'elsewhere' is that they usually contribute and are only used for that one article rather than to hide in a sea of anonymous users all wearing the same identity.
Incidentally, I wouldn't want it to be much higher. If it was 100%, it would mean we were picking too conservatively.