What Happens At Y Combinator
ycombinator.com
ycombinator.com
There must have been many mistakes along the way, while this article talks mostly about the successes. I'm curious what mistakes you made and avoided... maybe that's a topic for another article though.
It's painful to consider whether it would have been best to do YC in SV year round from the start. I loved alternating between the two places. I still miss living in Cambridge. And yet the founders in the Boston batches probably were getting short shrift. The startup community in Boston is so much smaller. We could never have organized something like Angel Day; there just aren't enough good angels. We wouldn't even have had enough alumni in Boston, because most of the successful ones used to move to the Valley.
You can figure out what we'd do differently by what we've changed. We give startups more money on average, but we don't give them any more for having more than 3 founders. Jessica now strongly encourages vesting, not to protect us, but to give startups a predefined way to deal with founders leaving. Interviews and Demo Day presentations are now much shorter, and I'd keep both short even if it weren't necessary for scale. We look more at the people and less at the idea when making funding decisions. We've added a bunch of protections against good groups falling through the cracks in applications. But we haven't changed that much structurally. We've made a lot of mistakes, but they've mostly been one-offs, like funding startups we shouldn't have, or not funding startups we should.
Examples?
(Sorry it's so long. I was surprised myself how long it turned out, but we do a lot.)
I am curious as to what you have in mind. (I tried to find an email address for you and couldn't. Sorry to junk up this thread. You can email me -- from one of the websites in my profile -- if you don't want to post it here.)
Part of it may be that you need medical connections to do the much of the work. We work directly with a ton of different doctors/hospitals (for a team of only ~15 devs), and still have issues with building the right things, 510k submissions, etc...
But yeah, medical applications are getting neglected by the startup community at the moment.
Many developing countries have had none or next to nothing health insurance till recently. But the market penetration of the healthy insurance is increasing (but still very low now). Because of that more people who could not afford advanced diagnosis or treatments earlier are able to afford them now.
Imagine the potential size of the industry when 1 billion people in India and 1.4 people in China (not to mention a billion or more in other developing countries) can afford as good health care as the 300 million in US, the 450 million in Europe, etc
23andme is a great medical startup, but it is a capital-intensive business that has to spend time and money on Washington.
We (Epocrates) are working on an Electronic Health Record system as an additional product. Our core team is small, but we have the resources and brand of a midsize company. We have 40% of US physicians using our existing reference software and we are going to leverage our brand to provide a usable and moden system that increases the level of patient care and effectiveness of Doctors. We're a mixed RoR/Java/Mobile shop. We're currently hiring RoR developers, so please feel free to contact me if you are interested in this space and want to make a real impact on US healthcare.
The medical device/software field is certainly interesting, exciting even at times, but it progresses at an absolutely glacial pace for the most part. You're dealing with slow, conservative institutions (hospitals, etc) and government agencies. Unless you hire people specifically to deal with those soul-crushing interactions, you'll probably want to shoot yourself.
The way a lot of record labels work is that they will give a band a cash advance. The band uses that advance to pay for living expenses. They are also forced to use it for various costs such as studio time, recording, producing a music video, etc. The band earns a percentage of the gross revenue for album sales (say 7%), the label earns a larger percentage. The band pays back their advance using their share of the earnings. Stop me if it seems like there's something a bit off about this arrangement.
Sending a fledgling startup company to work amidst that hell is not a smart move. Even billionaire multi-platinum bands have trouble.
One question though, what about startups from outside the US? Can you offer advice there visa-wise and such? I know it doesn't really matter for the three month period, but what about after that. It would really suck for a startup to do well on Demo Day, only to be unable to go to any meetings because they can't get back to the US for a few months (I think the wait period is another three months)
Once you raise enough to pay yourself market salary, the situation transitions from "impossible" to "annoying and long".
So focus on the product, then on funding and/or revenue: the rest will follow.
It is possible to bootstrap a company in such a way as to plateau out at, say, 3 people, each of whom gets a decent market salary. However, that company may not have further room to climb for various structural and/or marketing reasons, lack of appropriate connections, funding opportunities, etc. In principle, the purpose of YC is to address that specific gap irrespectively of how much short-term revenue the startup may be generating.
On the other hand, I agree that the amount of funding they provide per se is not particularly useful if you are already generating enough revenue to pay yourself(ves) a market salary.
As pbiggar points out above, there's no problem (at least for EU citizens) getting into the US for three months of YC - it's staying there long term that's hard.
If you mean do a program like YC, it's fine. The reason it works in this situation is because you aren't making any money. The INS does not care if you are "working" if you aren't actually making any taxable income.
If you mean draw a regular salary, it's harder to do that legally but I have a friend in France who runs a 1-man consulting firm that does short term contracts in the USA and he's figured it out. I don't think it was easy in terms of "no work" but it was just filling out a bunch of paperwork.
Doing the job you normally do isn't allowed, but it seems like it's pretty obvious that people do that.
Has anyone here on HN got into YC from London?
We have to recruit the largest founders in each batch to act like Tokyo subway pushers and herd all the investors back to their seats.
The founders of Cloudkick are sysadmin gods who within the alumni network play the same role for servers that The Wolf in Pulp Fiction played for the back seats of cars.
If you're not familiar: http://www.youtube.com/watch?v=wWmRTjLRMfU
My one suggestion is if you want to expand your age/experience bracket and get these people to apply, you may want to address issues directly relevant to them. One suggestion is to relabel the investment to something more like a 'stipend' and explain clearly that it's not expected to cover salary and life expenses, but only to help defer company expenses. That way, people know that they need to set aside enough savings to cover their family.
The other is to spell out clearly the policy of working on outside projects (to bring in some money) during the YC 3-month period. Leaving it ambiguous keeps out people who need a part-time gig to bring in some money but could spend the rest of their waking time working on a cool product.
I know people who work as consultants and make $10K-$25K in a month. To these people, the YC 'investment' is more of a deterrent. The main attraction would be to get wired into the network and tap into the experience YC and its alumni can offer. It would be good if you could clarify whether it makes sense for them to apply or not.
Overall though, a good write-up with a decent level of detail. Thanks for putting it up.
Additionally if someone is wary of losing their great consulting/contracting fees to start the startup, maybe starting a startup is not for them. Or maybe they just need to plan to minimize the risk that they will be taking. Either way, this is not YC's problem to fix, imo.
I know people who work as consultants and make $10K-$25K in a month. To these people, the YC 'investment' is more of a deterrent. The main attraction would be to get wired into the network and tap into the experience YC and its alumni can offer. It would be good if you could clarify whether it makes sense for them to apply or not.
Pretty sure the official answer is that they wouldn't be interested in taking someone who could only work part-time. I'm sure there are edge cases here and there, and YC is all for creative ways to stay alive financially (AirBnB sold cereal to stay alive - http://money.cnn.com/galleries/2010/smallbusiness/1003/galle...), but the YC school of thought is that a startup should occupy 100% of your work-related effort.
Of course, plenty of us worked as consultants for $10K-$25K/month before YC (er, closer to $10K) in order to build up some savings to live on.
It's funny that in my field (economics/finance) we treat startups as black boxes, and we have no clue about how they work. We argue that they are "crucial to the economy and to growth", but yet we haven't made any progress at understanding [1] them.
Do you guys think that we (academy) have any chance at finally understanding/modelling what you do?
[1] "Understanding" = Either writing a mathematical model, or test some hypothesis with a few regressions.
- 2 founders is best - founders should have known each other for a while before starting the company - team is much more important then idea
I do have one question: is the goal of all the companies to raise addition funding rounds? Or, do some startups still want to work without raising the extra capital, but rather grow organically (37 signals style). I guess another way to ask is: do you force (or even strongly expect) companies to get extra funding on demo day, do you let them choose (and they always choose to get more funding), or do some startups actually not get extra funding?
I ask because, from your article, it seems like most of the latter part of the cycle is geared towards raising capital, and I had always assumed that not raising capital was also a viable option for startups.
Though I don't mention this explicitly in the article, Demo Day isn't just for raising money. It's a good way to get mindshare too, because the people in the audience are so influential. So even the startups for whom raising money is not a top priority usually try hard to be impressive on DDay.
What signaling does YC look for in a company whose idea was birthed a week before applying, versus one who applies while their beta is in progress? Is it as simple as "these guys are smart" in both cases?
I've met lots of guys way smarter than me with less motivation, so there's got to be more to it.
For us the most important thing about the company having been running for a while is that it gives us more information about the founders. E.g. we can ask them "what have you learned from your users?"
On the other hand, I'm sure there are a handful of red flags and heuristics that they emphatically won't share, simply because they're effective filters. Their disclosure could let people better obscure true problems or falsely inflate perceived value. Reviewing applications after that would be a lot less fun.
Minor typos I noticed (since you seem like you want the feedback):
missing 'the': "to get the company into best shape possible"
spelling of predict: "know the investors but can preditct their reaction to each specific startup"
see instead of seem: "no one except the current batch of startups gets to seem them"
Maybe it seems that way because this talks about all the stuff that's common to all the startups, and the things to do with the product are usually more startup-specific than those to do with investors.
I'm the last person to give you writing advice, but maybe some examples would help.
I'd also say that even the fundraising stuff isn't just about raising money. Having a smart angel investor give you feedback on your startup isn't just about raising money; it's also about building a product, a team, and a company.
Like many people I sometimes feel ripped off when I pay 50% more for that box of Kellogs cereal or whatever. I imagine the same feeling might be the motivation behind AngelGate.
From an angel's point of view, YC is a deal refinery. Raw deals come in. High-quality, refined deals come out. Many investors will gladly pay a premium for those, because much of the early risk has been taken out: do the founders get along? Do they crack under pressure? Can they ship something? Can they adapt to changing circonstances?
YC figured out a repeatable, streamlined formula for a process which was previously a black art. They industrialized the gut feeling. And they deserve every bit of recognition they got for it.
That's not my understanding. I don't think they have any formula other than "We recognize great entrepreneurs when we see them" - It's not something they can write down in a book and sell to other people - so, in that sense, not industrialized. Still a gut feeling.
I could be wrong though - happy to hear otherwise.
Anybody who's talked to PG will tell you he is a walking database of startup patterns. "I've seen X and Y do this and they failed. Don't do it". I'm sure as YC scales, that database will be externalized to a medium other than PG's brain :)
So, yes, there is a formula: it's called data.
Funding 36 startups at a time would have seemed impossible back when we were doing 8. And in fact it would have been impossible to do 36 the way we operated then. So how many will we be able to fund at once, after we've taken advantage of whatever new techniques we discover for scaling? We can't say, because we haven't discovered them yet.
Interesting. If you know beforehand that those investors' expected value is 0, why invite them?
Also, zero is probably an overstatement. It would have been better to say epsilon. Because most of these investors I would take money from if the alternative would be to shut down the company.
Just a speculation: Different investors may also have different values to different companies.
I guess my question is, do you see single-founder companies as missing something (a cofounder), or does it depend on the founder?
One question (maybe a bit sensitive): all the glory of successful YC startups aside, what are the "failure rate" for YC funded startups? What situation would you consider a "total fail", if that ever happened?
And another related question is based on which criterion would you consider a YC graduate successful? user base? the amount of money raise? financially break even? profitability?
thanks.
"The general answer is to pick something where the product of how fast it can be built and how excited users would be about it is high."
x = speed-of-implementation(idea)
y = usefulness(idea)
rank(x,y) = x*y
Scary stuff starts happening in any business when you take the limit as x approaches zero. (This came up in client work today.)
I hope you do not think I am stupid for not immediately seeing the point of many seembols.
1. What do you - at YCombinator consider to be your best example of a successful start-up? If you can name that company, that would be awesome and appreciated.
2. On the other hand, what do you consider to be a flat out failure? An example of an idea that totally did not work? If you could name this venture, it would be appreciated too.
One good example of a bad idea that I can quote without hurting anyone's feelings is Artix: http://paulgraham.com/bronze.html
Have you noticed any particular combination that seems to work better than the others across the board, or does it depend on the startup?
People start to show up for dinners around 6 pm.
This seems wrong, unless the YC dinners conclude at 6 AM.
Huh. That never occurred to me as an interpretation.
I wonder if this is a British vs. American vs. Canadian English issue, or if I'm just slower than usual today.
This may be another of those "organ-eye-zation" versus "organ-eh-zation" things that pop up form time to time.
Oh, and it's "organ-eh-zation" through and through. :)
Do you think it would be possible to start YCombinator prime elsewhere in the country? If we put together many of the same factors, angels, infrastructure, good mentors, could this be replicated, albeit on a smaller scale?
It comes down to people, money, and time. Exporting Silicon Valley might be the wrong approach. Another place must have its own culture and value system in order to survive and those qualities can't be borrowed.
Boulder, Boston, Austin, and to some extent NYC and a few other east coast (of the US) cities (my heart belongs to Baltimore) have burgeoning scenes, but each is different than SV. Where they don't try to simply mimic, I think they're more likely to find success.
You need "enough" of all of the critical factors and then you have to overcome whatever penalty you incur from missing some of the "nice to haves". Considering the odds against success even with all of the stars aligned ....
Seriously - unless your customers care where you are, why are you so resistant to going to where the odds of success are highest?
It's okay to value things other than success, but you're competing with people who are focussed on success. How are you going to overcome that advantage?
seriously?
How about anonymizing it by removing the names of the startups/founders before publishing ?
Anonymising information while leaving it usable is hard to the point of (near?)impossibility
Approximately what percentage of startups do you accept that apply with no particular idea?
I quit my job three months ago and have been doing consulting work in order to generate enough runway to finally dedicate ourselves fully to our own product.
I think we will be done with all of our current gigs by november, so the Winter YC cycle looks VERY attractive right now. We have a couple of ideas, and it seems like we might have enough time to get started on them before YC, but still be flexible enough when the cycle starts.