(I’m live-blogging from Startup School, a daylong program from startup incubator YCombinator held at Berkeley today. Mark Zuckerberg is on-stage for a question-and-answer session and I’ve embedded a short clip from the talk below. This is paraphrased.)
Jessica Livingston: First I want to go way, way back. I know you built a few things in high school. What did you learn from those experiences?
Zuckerberg: I mostly built stuff I liked. I built games and I built AIs (artificial intelligence programs) to play the games against me. My friends would build AIs to play against my AIs. In college, I built a bunch of programs that required other people to use them. In the reading period before finals of my sophomore year, there was a lit and arts class I took. I didn’t go to class all semester. I was supposed to be studying. But instead, I was building Facebook. Before the exam, I was screwed. So I went to the course web site and downloaded all the images that we had to learn the significance of. I made a site that would randomly take you to the image, and then you could write a caption in a bubble. Then it would show you what other people did.
Later the professor told me that was the final where people had done the best ever.
It was my first social hack….
Livingston: Who were the first people using it?
Zuckerberg: At first it was very simple. Just 10,000 lines of code maybe. There were profiles. And you could poke people. That was important. The idea is launch early and iterate. Early on, I didn’t just start Facebook as a company. It was a project that I wanted to exist. It’s amazing how much stuff we messed up.
Livingston: Tell me about the dumb things you made the first year around.
Zuckerberg: We weren’t even set up as a company originally. I started the company with a few friends who I thought were really smart. We were at different levels of commitment. When I moved out to California, I had no idea about all the things you needed to do to build a company. A lot of the other founders didn’t want to move out. The early group was fractured like that. I didn’t want to be involved in building out the company stuff at all. We had this really smart guy Eduardo (Saverin). I thought he was very smart. Instead of setting up a normal C-corp structured. He set us up as a Florida LLC. I don’t even know that was wrong with that. When I moved out and I met a lawyer, they said that’s number one to unwind. In the beginning, we weren’t trying to make it as big as possible. We wanted to see if it resonated and provide a lot of value.
We launched it at universities where there were a lot of competing online services. We launched it at Columbia and Yale where they already had an online community. When we launched it and saw this huge reception.
But we always had this joke about people who wanted to build companies, but didn’t want to create something super-valuable. There are a lot of people like that in Silicon Valley. But we wanted to make sure this was a valuable thing before we invested in.
Livingston: A lot of companies face this chicken-and-egg problem.
Zuckerberg: Facebook, by definition, is a viral product. There are lots of sites now that try to include a contact importer. For a lot of them it doesn’t make sense. It wasn’t actually until a few years in, that we started building some tools that made it easier for people to import their contacts. There was a huge amount of organic spreading.
We then launched at schools that had the most people asking for it. We didn’t have enough money to launch at all these places. We just got more servers at $85 each, and we had to wait for ad revenue to get more.
Livingston: Who were your first investors and how did you pitch them?
Zuckerberg: I’m not good at pitching things. The cool thing was that we never really had to pitch anyone. Peter Thiel was one of our first investors. And then Sean Parker helped connect us with everyone. He helped set up our first outsourced accounting firms. He really talked to Peter the most. At that time, there were already hundreds of thousands of people using the service. It was clear that if we executed well and launched at more schools, it would continue to grow.
I was 19. I doubt I was at all impressive to him.
For the first two years, the only advice he gave me was don’t mess it up. The YCombinator motto is build something that people want. We were never that great at building a company. Now we’ve built this sort of hacker culture. Early on, we were clueless. All of this came from having users and serving them well and growing at a sustainable rate. We’re typically a little bit ahead in terms of the financing we did. All this stuff, when we did our investment with Thiel, we didn’t even have facebook.com. We just had thefacebook.com
Livingston: If you could do things differently, what would you do?
Zuckerberg: The iterate advice is the most important. We finally got facebook.com, for a lot of money, like tens of thousands of dollars. A lot of the management team has steadily improved. It was very hard to get people who were good at managing businesses to join early on. Now we have Sheryl. She’s excellent. She’s as good as it gets. But there’s no way she would’ve joined early on.
A lot of my friends who have either left Facebook and went into other things — it’s more likely to make the mistake of trying to be too perfect. It’s just better to just launch and have something cool that you can fix over time.
Livingston: So as you grew, what did you do differently at Facebook than at other companies?
Zuckerberg: I looked around at Valley companies. I really wanted to build a technology company. Some call themselves technical companies. But they’re not. We were always really committed to just kind of maintaining technical folks and entrepreneurs running the company. Even in non-technical roles, one of our marketing roles is someone who has a technical background. Chamath (Palihapitiya). Our new CFO was CFO of Genentech. Before, he ran product at Genentech.
As we’ve grown it’s become more apparent to us how we’re different from other companies. Google is a technology I admire. It’s become apparent the way we’re different. They have a really strong academic culture. They do a lot of research and a lot of Ph.D’s. What we pride ourselves on is having a lot of impact, having a strong hacker culture, giving people a chance to grow — those are things that we work on. We have more than 300 million users and less than 300 engineers. The ratio of users to engineers is far more than any other company.
One of the things we’re committed to is having a smaller team. We have frequent pushes for code. We encourage people to build a lot of their own stuff. We have a technology company that’s really a hacker company.
Livingston: How do you keep and attract people?
Zuckerberg: Our goal isn’t necessarily to keep people forever. There are companies that train people really well. A lot of Harvard people went on to McKinsey. A lot of people went to IBM because that was the best place to learn sales. A lot of people go to USC to learn how to play football. One of the things at Facebook is have a place where it’s one of the best places to learn how to build stuff. If you want to learn how to build really good products and practices and have a large impact, I would argue there’s no better place to do that than at Facebook.
If people want to come here for one, two, or three years. Steve Chen, who founded YouTube, was working on Facebook before. I’m not encouraging people to work at Facebook to leave. But I think you learn valuable skills. We’re not pretending we’re building a company that hackers are going to want to work at forever. I want to be a part of building some institutions to be a great hacker institution in the long-term.
Livingston: Tell me about the moments when you thought — Whoa, this is going to be big.
Zuckerberg: A lot of founders will tell you they had some huge vision. Facebook was more polarized at the beginning. My partners who I did problem sets with just talked about trends — like how there would be greater openness in the world and how that might impact global governance. And I thought that was really cool. But I didn’t think I would have any chance of influencing it.
On the side, I built this little project. Over time, it really grew. A few years in, we were sitting around one day. Maybe this is going to be one of the things that is going to make an impact. Of course, there are a lot of other really great companies working on this too. We kept seeing people using Facebook after college. We realized people in all different locations and ages would use this. Everyone has friends and family. Wanting to keeping in touch is a really universal thing.
Livingston: So scaling, was it really gradual?
Zuckerberg: We limited it for awhile. There was a period where we accelerated user growth ahead of revenue growth. But when the economy turned south, we wanted to get cash-flow positive.
Some of the stuff when we first rolled it out was pretty crazy. Photos — there was a two month period where people weren’t sleeping a lot inside of the company.
Livingston: What have you learned about running a big company?
Zuckerberg: Not clear that I’ve learned much. A lot of it is — I don’t know — you want to have the values and culture that you think are going to be unique and are going to be the best for what you’re going to be doing. You can move so much faster when you’re small. You get bigger and you get slower. We need to institutionally push moving really fast, or else it will just be inertia.
I actually think a lot of the best stuff we’ve done is try not to take all the advice from other people. What I said the last time at startup school and I got yelled at — is it’s a group of entrepreneurs who are mostly technical. All of these people would tell me you have no experience. You’re just this engineer. You can’t do this. My message the last time was No.
People interpret this as you don’t value experience. But you need to value what you bring to the table.
Audience: What are you going to do about storing all of this personal information for posterity — like historians 2,000 years in the future.
Zuckerberg: I pride myself on thinking long-term. Maybe not that long-term.
Marc Andreessen told me this a long time ago about his competition with Microsoft. One of the things that made them win was that they had a longer time horizon.
Two thousand years out? I don’t know. A lot of my approach and our approach. We feel like we’re part of this movement. We’re trying to push the world to be more open. The best companies have this strong sense of purpose. You have to understand that this thing that you’re trying to do isn’t necessarily going to be forever. Microsoft just released Windows 7. It’s really good. But a lot more is now about the web. There was a long period where OS was the most important thing in the world.
Now, pushing the world to being more open and connected is going to be one of the most transformative trends. But at some point we’ll be mature and hopefully we’ll get to that point. Some other trend will take over.
Audience: One of the things that showed me you’re in control was the backlash to newsfeed. And you said the most disruptive companies have to
Zuckerberg: The biggest risk you can take it is to take no risk. In a world that’s moving quickly, you know that if you don’t change you’ll lose. Not taking risk is the riskiest thing you can do. You have to do things that are kind of bold even if they’re not obvious.
We made this change yesterday that lets you see the most recent stuff or the most interesting updates. Regardless of what we do, if it changes the site. Change is really disruptive for people. Especially if you’re providing a web service that people aren’t opting into directly.
Values are worthless unless they’re controversial. We want to move quickly. We’re willing to give up a huge amount of stuff in order to move quickly.
Audience: You’re now a public figure. People can pounce on any statement you say. You sound different than you did in the beginning.
Zuckerberg: More interesting or less?
Zuckerberg: I feel like I’m pretty honest about how I think. Maybe we need more controversial questions. What I know about is what we’ve done.
Robert Scoble: I see you walking around your employees and managers. What management skills are you learning?
Zuckerberg: Management skills?
Scoble: Give the audience what your management philosophy is.
Zuckerberg: You need to focus on a few things you care about. Moving fast. You need to give up some things you need to do that. I think a really big part of a CEO’s job. You hire the team that’s going to go build everything. You set up the direction. What really comes off is what trade-offs you’re going to make.
We think eventually you get judged not by how you look but by what you build. I think there’s a lot of pressures inside of companies where people want to optimize for a launch to go smoothly. But really you want to be the best over time. Take more risk. If we do those, we have a good shot of succeeding.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.