Paul Graham of Y Combinator Pulls Back the Curtain on What Goes On At His Start-Up Factory

When Michael Arrington reported last week on the secret summit of Silicon Valley investors that is at the center of what’s now being called “AngelGate,” one of the most specific details he offered was that in the course of their meeting, the angels had complained about the “growing power” of seed stage investment firm Y Combinator, and had discussed how to more effectively compete with them for start-up deals. Arrington’s report– in which he accused the angels of collusion and price fixing– has set off a ferocious controversy in the tech world, moving lots of people to write lots of angry emails and blog posts on the subject.

Yesterday, Y Combinator co-founder Paul Graham jumped into the fray. But unlike some of the others who have made their voices heard amid this controversy, though, Graham took an understated approach, writing out a lengthy and pleasingly detailed explanation of what exactly it is that Y Combinator does for start-ups. The closest Graham comes over the course of the 8,000 word memo is at the very top, where he explains his motivation by saying that he and the YC team “realized recently that a lot of people don’t understand very well” what they do.

If you’ve been dreaming of starting your own company but haven’t known where to begin, reading this document will probably give you butterflies. Graham really makes it sound like an idyllic, life-changing program that provides people who know nothing about the power structure or protocol in Silicon Valley with everything they need to leave the runway, including introductions to lawyers and investors.  

Graham begins with the basics: there are two three-month funding cycles per year, during which start-up founders who have been accepted into the program move to the Bay Area. While there, they work with YC’s team of advisers on building their projects and readying them for the eyes of investors. The first YC cycle was in summer of 2005, at which point they funded eight start-ups. They’ve since grown, funding fully 36 during their most recent session this past summer.

What actually happens during the sessions sounds a lot like summer camp. Once a week there is a group dinner featuring a speaker from the start-up world. Everyone comes at 6pm and shows off to one another what they’ve been working on before taking their seats around 7:15. “The general atmos is like a modernist version of an Oxford college dining hall, but without a high table,” Graham writes. The featured speaker, usually a successful founder, takes the stage during dessert and speaks frankly about the origin story of his or her company, leaving out little because the talks are strictly off the record. “Because YC has been around so long and we have personal relationships with most of the speakers,” Graham writes, “they trust that what they say won’t get out and tell us a lot of medium-secret stuff.”

Start-ups are advised to launch as quickly as possible– as soon as their product has a “quantum of utility”– because it’s only through feedback from real users that founders learn how they must change their product. They’re also encouraged to come up with a hypothetical business model.

When a start-up is ready to launch they’re asked to draw up their pitch on a whiteboard, photograph it, and send it to Graham so he can turn it into prose and put it before tech publications (“usually but not always TechCrunch”). The reason Graham is the one who writes the email, he writes, is that he knows how to explain what the companies are up to in “plain language instead of the marketing-speak even technical founders often revert to” and that journalists know he doesn’t “bullshit them.”

Once a company is ready to launch– Graham says that about a quarter of YC companies have launched by the time they arrive at the doorstep– its founders spend their time talking to the YC team about preemptive troubleshooting. “Most things that happen to newly launched startups are bad,” Graham writes. “But paradoxically, these disasters are precisely the reason to launch fast: they all represent problems you’re going to need to solve eventually, and the only way even to find out what they are is to launch. In practice they vary from technological bottlenecks to threats of lawsuits, but the most common problem is that users don’t like the product enough.”

The next stage of the program is dedicated to fundraising. Graham writes that “the most important tactical question is usually who to close first, and how to convince them.”

The week before what’s known as Demo Day– when the founders make presentations to a room that includes “most of the world’s top startup investors”– start-ups rehearse what they’re going to say and how they’re going to say it at all hours of the night. “Startups who come to YC a lot the week before Demo Day always have better presentations than the ones who for whatever reason don’t,” Graham writes, “even if their companies aren’t actually as good.” About 40% of startups have already raised some angel money by Demo Day.

At the end of the memo, Graham asserts, as he has before, that Silicon Valley is the best place in the world to found a startup. “There’s no question you can start a successful startup elsewhere,” he writes. “There are plenty of examples to prove that. But even the most ardent boosters of other cities wouldn’t claim they’re at parity with the Bay Area.”

Other interesting bits:

— “Because the alumni network is so large and tightly knit, investors or companies who try to maltreat a YC-funded startup can usually be made to stop. But because everyone in the Valley understands the power of this network, most of its value is in deterrence: in practice people rarely do try to maltreat YC-funded companies, because they know it wouldn’t be worth it. The kind of horror stories you hear about investors dicking over startups rarely happen to those we fund.”

— “Early in YC’s history we used to take each startup out to dinner at some point during the cycle. Now that we’ve grown past the point where we can do that, we invite several at a time over to our house for dinner instead.”

— “We host a reunion dinner for each batch at YC. Because the time scale is so fast in the startup world, we don’t wait a year: the reunion dinner is usually about 6 weeks after the end of the cycle, because that’s when startups running into headwinds start to get discouraged and need cheering up.”

— One “type of work we do is mediating disputes between founders. These are more common than anyone realizes, because the parties involved generally try to keep them quiet, but they occur in even the most successful startups. We’re now very good at dealing with them because we’re not only neutral, but have enough experience to tell which disputes are serious, and for those not serious enough to warrant a divorce, to figure out how to fix the underlying problem. Paul Graham of Y Combinator Pulls Back the Curtain on What Goes On At His Start-Up Factory