Y Combinator Gets an Historian: Inside YC’s Summer 2011 Session

The best way to find out what it’s like inside Y Combinator is to found a startup and get admitted to the program. The second-best way is to read Randall Stross’s new book, The Launch Pad.

Y Combinator, of course, is the exclusive school for startups in Mountain View, CA. It’s been described as a seed fund, a venture incubator, a startup accelerator, and (inaccurately, in Stross’s view) a startup “boot camp.” But at its core, it’s an investing operation. The goal is to suck in as many promising young hackers as possible, invest enough money in their startups so that they don’t have to worry about the rent and pizza bills for a few months, shepherd them toward product ideas that people might pay for, and hope that at least one of the teams in each batch grows into the next Dropbox, Airbnb, or Heroku.

It’s a model that has revolutionized the startup world. Y Combinator started from humble beginnings in 2005 (there were just eight companies in the first batch in Cambridge, MA) but has spawned scores of imitators across the country, as we’ve documented in our 2012 Guide to Venture Incubators. YC’s semiannual Demo Days attract a fleet of Tesla Roadsters and Fisker Karmas to Mountain View’s Computer History Museum, each with a checkbook-toting Sand Hill Road venture partner behind the wheel.

But getting into YC isn’t a golden ticket by any means. Even after soaking up three months of advice from the YC partners and the parade of prominent entrepreneurs and investors they invite in, many of the companies eventually sputter out. YC Founder Paul Graham makes it pretty clear to the admitted startups that most of them are unworthy of the money that YC, Start Fund, and SV Angel invest in them on such incredibly lax terms. (Starting in the winter of 2011, the two outside funds began putting $150,000 into each admitted startup in the form of uncapped convertible notes, meaning the equity shares they get back aren’t determined until the companies raise more money.) In the end, the startups are just variables in a giant risk calculation. “They’ve got to offer you all terms that sixty-two out of sixty-three of you don’t deserve, to make sure that they get the Dropbox,” Graham told the Summer 2011 batch, in a talk that Stross reports in his book.

YC’s job, however, is to make sure every team with the requisite talent, grit, courage, and nimbleness gets a shot at the jackpot. (Graham calls this black swan farming in a widely discussed essay published this month.) Stross, a PhD historian from Stanford who writes the weekly Digital Domain column for the New York Times’ Sunday business section and teaches business at San Jose State University, decided back in March 2011 that he wanted to understand how this process works. Luckily for him (and for readers), Y Combinator gave him unprecedented access to the program, letting him observe the Summer 2011 batch all the way from the application process in April to Demo Day in August.

Photo of Randall Stross

Historian, journalist, and author Randall Stross

The resulting book is not a tell-all or an exposé. You won’t find out which coders downed the most Red Bull, which ones broke up with their girlfriends, or which venture investors have the most colorful vocabularies. Stross’s approach is that of an historian with an anthropological bent, and the book is as close to a straight narrative of the summer session as one could get, considering the challenges of following 63 startups for three months.

Dialogue drives the story forward, but it’s not banter—it’s serious discussion between the YC partners and the startups about the challenges they face building compelling new products and services and getting customers interested in them. One nine-page passage focuses in unbelievable detail on the pluses and minuses of YC startup Tagstand’s plan to build a business around infrastructure services for near-field communications vendors. “Journalists shouldn’t get to be the only ones who do the first draft of history,” Stross says—and you definitely can’t call this draft sensationalistic.

The Launch Pad, published by Portfolio/Penguin ($26.95 hardcover), hits bookstores on September 27. You can read an excerpt now in this month’s issue of Vanity Fair. I spoke with Stross yesterday, and I’ve typed up the following edited version of our conversation.

Xconomy: As you point out in the book, the Y Combinator partners are always telling startup founders that they should be ready to answer the “why you” question—in other words, why are you the person who should be starting this company? So I’ll start by asking you a similar question: Why were you the person to write this book about Y Combinator?

Randall Stross: I’ve been paying attention to startups from the time I wrote a book about a startup that didn’t do very well, Steve Jobs’ NeXT Computer. I had a friend in grad school who went off to work at NeXT, and he urged me to write a book about what he thought of as the most interesting startup in Silicon Valley, and he was right, but not for the reasons we all expected.

A few years later I wrote a book about startups that were funded by Benchmark Capital and spent two years in the office at Benchmark, embedded in the firm and watching how they went about making investment decisions and following as many of the portfolio companies as I could. This was from 1998 to 2000. The book that came out of that experience was titled eBoys. And although I didn’t know it when I knocked on Paul Graham’s door, he had read and liked that book. So he was familiar with my work. I gather that was very helpful in his and Jessica’s [Livingston, YC Partner and Paul Graham’s wife] deciding to let me watch the program.

X: What was your pitch to them?

RS: The Benchmark book used a fly-on-the-wall approach that tried to give the reader a vivid sense of what happened. I reused the pitch from eBoys with Paul Graham and the other partners. Which was: this is a historically interesting moment, and journalists shouldn’t get to be the only ones who do the first draft of history. Historians should get to too.

X: So you think of yourself primarily as a historian? You’ve published in a lot of journalistic outlets, such as the New York Times.

RS: I was trained as a historian. I think I have a historian’s sensibility. So I practice journalism, but I think at heart I remain a historian.

X: Did the partners buy your pitch, or did it take some persuading?

RS: I don’t know what went on—it was a pretty short exchange of e-mails. The only objection initially was that the timing was not right, because they had already given permission to a group of filmmakers to film during the Summer 2011 batch. Paul had initially said “I can’t have you both here at the same time, so it’s not going to work.” And I thought that was the end of it. Then he came back a few weeks later and said he had given it more thought, and there was going to be a large batch, so there would be lots of companies for each of us to follow.

So I started with the expectation that I would have the filmmakers there too. Which would not have been ideal, because I think the presence of cameras would have made everyone self-conscious—more so than just a writer with a tape recorder. As it turns out, the filmmakers apparently had difficulty getting the funding they wanted, so it was just me.

By its nature, startup culture is always a collection of beginnings. And what is more hopeful than beginnings? It's like hanging out in a nursery with a nothing but cute newborns.X: Speaking of self-consciousness, do you think you got a good look at how Y Combinator really works? The presence of an outsider always changes things a little bit. Do you think the YC partners and the founders behaved around you the way they naturally would?

RS: It’s hard for me to judge. Certainly, as time went on, I think everyone became more comfortable with my presence. Because most of the situations I was sitting in on involved three, four, five, or more people, the numbers were in my favor. I don’t think I stuck out as much as I would have in one-on-one conversations.

X: Also, I’m assuming the startups all knew that you were working on a book, and that the details wouldn’t come out for a year or more.

RS: Yes, that was something that was important to the partners. They understood that this was a different sort of thing from a feature story. There would be that delay, and that meant that things that obviously were very sensitive at the time, were they to be made public, would not be so sensitive a year or a year and a half later.

X: Did you have full editorial independence?

RS: Yes. The only restriction that was imposed on me—and I didn’t realize it when we started, but I understood the necessity for it—was that I could not record or use the Tuesday guest appearances. Those are off the record permanently. They have a tradition of enjoying complete blanket protection of what is said, and Paul Graham didn’t want to impinge on the speakers’ comfort in any way.

X: Did you have a picture of Y Combinator and what it’s like to be a YC startup before you started the project, and if so, how did that picture change as you went on?

RS: Because I took an anthropological approach, I did not have a model in my head that I needed to adjust and edit and update. I was ready to observe whatever unfolded. Also, I was so busy trying to follow as many of the companies as I could that maintaining some mental model wasn’t really the thing I was worried about.

I had this challenge of following what started out as 64 companies, and I knew I couldn’t follow them all. So how would I know which ones would produce the most interesting stories, the stories that would have the most representative characteristics? That’s what I struggled with.

X: How did you resolve that?

RS: It was partly affected by who was willing to work with me. Probably four-fifths of the companies actively reached out to me saying they wanted to help. That still left a lot. So one of the things I tried to do was focus on the companies whose products or services would be ones general readers might encounter. There were some very specialized companies, serving narrow, vertical markets, where it would be very hard to describe what they were doing. There were a number of companies in the batch who are what I call D2D, developer-to-developer. Some of them were included in the portraits in the book but there were many that I could not include.

But I felt myself fortunate that the class had expanded in size. It effectively doubled between 2010 and 2011, which gave me that many more possibilities.

X: Can you contrast your experiences with the Benchmark Capital book and your experiences inside YC? What struck you most about the ways startups have changed since 1998?

RS: One of the things that struck me, of course, was how young the founders were, and the absence of a requirement that they put in time at one of the technology powerhouses before they would get investors to back them. Back in the late 1990s, you would see engineers leaving Microsoft, HP, Oracle, or Cisco to start their own companies, but they had professional experience. And when they hired, they looked to other software engineers with impressive professional resumes.

At Y Combinator, you have founders who are coming straight out of college and have spent time not at recognized technology companies but at other startups. You have other founders who haven’t finished college—one of the two founders of Codecademy had not finished college, and they would end up being one of the companies that was most successful at raising funds at the end of the summer batch. This is something that Paul Graham has advocated for years—that the age boundary for founders could be pushed lower than it had been, and that what matters is what you can do, not where you went to school or where you worked.

X: Alongside that, you have the fact that the technologies and the skills that young people need to become good programmers are more within reach today.

RS: I think you’ve put your finger on one of the differences, which is the dramatic drop in the cost of very sophisticated computer resources. It allows teenagers and adults to use state-of-the-art technology. Back in the late 1990s, you needed a $20 million Series A round to cover the $5 million that you were going to pay Oracle and the $5 million that you were going to pay Cisco. There were four big checks you were going to write just to get started. Obviously that’s not the case today. You just sign up with Heroku and you are off and running.

There is another characteristic of Y Combinator, and I’m not sure to what degree other accelerators across the country share this, but that is the hacker culture, where many of the teams are composed entirely of hackers. If there is going to be anyone who doesn’t fit that, it’s a designer. But certainly there are no junior VPs of marketing on these founding teams.

Back in the late 1990s you needed a sales guy, or at least it was felt that you needed business skills on the founding team. And what you see among the YC founders is a belief that whatever they don’t already know, whatever would be needed on the business side, they can pick up as they go. The very concept of an MBA is so foreign to them that I didn’t hear a single discussion of whether there would be any need to hire someone with an MBA.

X: Even if there aren’t any MBAs, isn’t there usually at least one founder on each team who has the “sales guy” personality—the fast talker, the extrovert?

RS: I would say what’s surprising is the absence of that person on so many of these teams. I was fortunate to be able to listen to not just the interviews [of the companies applying to the summer 2011 batch] but the discussions that followed as the YC partners would decide how they were going to rate the finalist teams. And one of the things that would put off the partners was if they saw one personality dominating. That set off alarms. The YC model is of coequals working very closely together without any hint of a hierarchy. So if there was one who was more articulate, that would be acceptable, but there couldn’t be any suggestion that that person had more authority than the others.

There is a passage in the book where I talk about the two members of the Snapjoy team. They did an interview at TechStars where they were asked “Who is the CEO?” They hated that question. The answer was that they were both writing code. That is a question that would never have come up at YC.

X: What else jumped out at you as major differences between the startup world of 1998 and the startup world of 2012?

RS: Ideas are going to be tested in a much more accelerated way now. The Internet allows ideas to be cast out and put to the test and then discarded much more quickly than was the case. Of course, many of the investments back in the 1990s involved hardware, and that’s for the most part not something YC has invested in. But also, beginning with the winter 2011 batch, the first batch that enjoyed the $150,000 in debt financing that every member of the batch received, the companies know that they’ve got more time beyond Demo Day to come up with an idea that’s going to work. So not only do they in many cases discard the idea that they started out with during the session, but they are also trying out new ideas even after Demo Day. They have the means now to experiment in a way that would be unimaginable back in 1999 if they’d been given $20 million and spent it all, only to discover that the original idea isn’t going to fly.

X: You mentioned a minute ago that there’s a general disdain for MBAs at Y Combinator. Do you think that’s entirely a good thing? Even if business schools aren’t teaching it, there really is a set of theories and ideas about building startups that people like Steve Blank and Eric Ries are trying to spread. It seems to me that a lot of these accelerator-stage startups make mistakes they could have avoided if somebody on the team had had a little more business training.

RS: There is this YC ethos of learning from trying. There is no substitute for putting the idea out there. I suppose there is room for a kind of hybrid. I don’t know—maybe it would be studying [business theory] by day and doing side projects by night or on the weekends. But in the groups I followed, the primary source of advice seemed to be the YC partners and the fellow batch-mates.

X: So, in the big picture, having spent a year on this project and getting to know more than 120 startup founders in this batch, did you come away feeling optimistic about the state of innovation? Do you think tech startups can help to pull the country out of its economic doldrums?

RS: I would hate to put the hopes for an entire country on the shoulders of this one group. But there is no question that part of the appeal of startup culture is that it’s a bright spot in a landscape greatly in need of some encouraging stories. And by its nature, startup culture is perennially regenerating. So it’s always a collection of beginnings. And what is more hopeful than beginnings? It’s like hanging out in a nursery with nothing but cute newborns.

Wade Roush is a freelance science and technology journalist and the producer and host of the podcast Soonish. Follow @soonishpodcast

Trending on Xconomy