While still a teenager Mark Zuckerberg coded and launched Facebook—a digital version of the traditional college facebook—in his Harvard dorm room in February 2004. Within weeks virtually the entire campus had signed up. He and his roommates, Chris Hughes and Dustin Moscowitz, then ported the social networking site to other Ivy League campuses, and the rest, as they say, is history. Today Facebook continues to grow at a frantic rate, with hundreds of millions of users across the globe.
Penenberg: How did you get into coding? What was your experience with it from a young age?
Zuckerberg: Well I got my first computer in the 6th grade or so. As soon as I got it, I was interested in finding out how it worked and how the programs worked and then figuring out how to write programs at just deeper and deeper levels within the system.
Penenberg: Did you take things apart and put them back?
Zuckerberg: I did that more with software than with hardware. You hear these stories about people who take apart their radio and put it back. Or just learn a lot from taking it apart. But I wasn't as into that stuff as I was just into how computer programs work..
Penenberg: Did you teach yourself coding or what?
Zuckerberg: Yea, I mean I bought a book. Kinda messed around myself. Asked a lot of people questions.
Penenberg: When you first created Facebook, were you surprised that it took off so fast on campus?
Zuckerberg: Yea. I wasn't expecting that it would grow as quickly as it did. We had to keep building new infrastructure to deal with the scale that it was operating at.
Penenberg: What was driving you?
Zuckerberg: There are a few other things that I built when I was at Harvard that were kind of smaller versions of Facebook. One such program was this program called Match. People could enter the different courses that they were taking, and see what other courses would be correlated with the courses they are taking. And over my time at Harvard I built programs like that. On a small scale.
Penenberg: When you opened Facebook to other Ivy League schools and you saw the same kind of growth were you surprised at that?
Zuckerberg: There were a couple of things going on. One was that we had a sense of the type of dynamics we were tapping into were pretty universal and could apply in all places. The thing that was surprising was that our implementation, specifically, was so efficient at doing it. We didn't realize the magnitude of what we were onto with our version. After launching at Harvard, a bunch of schools wrote to us and asked if we could launch it at those schools. The first 3 schools we launched at after Harvard, were Yale, Stanford and Columbia and we selected them because they all had some kind of other school community Web site which almost everyone was using. We wanted to make sure that we had an implementation that was so efficient that even though everyone already had something they were using, they would just switch and start using ours. After we launched at those three and that happened within the first couple of weeks and there was a widespread switch from whatever the local one was to Facebook, then we knew that it would be something that could expand really far. So we launched at about 25 more schools that semester and it just picked up really quickly because there was no local community thing at those schools.
Penenberg: Did you notice powerlaw curve. How early was it that you saw that the rate of user acceptance was like that?
Zuckerberg: It was pretty quick. I remember we had these graphs up on our wall about how each school is adopting. They were normal S curves. The bigger the school, the longer it took to get into full inflection. At schools like Cornell, that were a bit bigger, it took about a couple of weeks before it really started taking off. But the nature of it was that in these schools everyone was on it.
Penenberg: Were you aware of the network effects and viral growth from the very beginning?
Zuckerberg: Yeah. I think the concept of network effects is pretty intuitive in something like this. Its basically that the value that people get is tied to how much information everyone is sharing.
Penenberg: How early was the invitation system phased into the program?
Zuckerberg: The ability for people to easily import their contacts from an email program, is I think what you're talking about, was added in late 2006.
Penenberg: Does growth continue at the same rate?
Zuckerberg: At first, we focused on getting everyone at a specific college to sign up. I think that now the growth curve that we have is very similar to that, except instead of colleges we have countries. Where there are some countries like Canada where almost 40 percent of Internet users are on Facebook, its still growing but its not growing at that exponential rate because too many people are using it. Then there's a lot of countries that are just starting off now that we are translating into other languages. So the growth rate question is a more complex question. Last year we saw growth rates of 3 percent a week for almost the whole year. And I mean is the aggregate user base growing by 3 percent everyday? No, it's really a combination of markets that are more saturated with other new markets that are just starting to grow exponentially.
Penenberg: How were you able to scale without a lot of major dramas, whereas other sites have had tremendous problems scaling?
Zuckerberg: I think there were a few phases that were really important. The first was early on we weren't structured as a company, and we didn't have a lot of money, and we were running ads to make money to buy more servers to launch at more schools. But because we knew we were going to be constrained, we built into the system that not everyone at the system could sign on at once, so in a lot of ways we had slowed down our growth. But the flipside of that was that as we were growing, we were able to not fall over.
Penenberg: Marketing to one's social graph offers a number of huge advantages. Why is that?
Zuckerberg: I think the basic idea here is that there is a phenomenon in peoples' interaction. The message that you get, in a lot of ways, is actually less important than who you get it from. If you get it from someone that you trust a lot more then you'll really listen to it. Whereas if you get it from someone you don't trust you might actually believe the opposite of what they said because you don't trust them. I think that's the basis of the value that people get on the site. I go to someone's profile and see that they like this band. That means more to me than if I just saw a billboard for that band. We figured that in the really organic way to make money and sustain the company, that these interests would be aligned.
[Photo By Jason McElweenie]