Chris Peterson

Benjamin Mako Hill, an excellent activist and academic associated with the Center for Civic Media and Berkman Center (among others), has a widely-circulated blog post up entitled Why Facebook’s Network Effects are Overrated. This is something I wrote about in Losing Face.

Ben makes a lot of really good points, but there are a few with which I disagree, and I wanted to comment on them further here. He writes:

And the relationships between services aren’t always peaceful coexistence. Remember Friendster? Remember Orkut? Remember Tribe? Remember MySpace? MySpace, and all the others, are great examples of how social networks die. They very slowly fade away. MySpace users signed up for Facebook accounts and used both. They almost never just switched. Over time, as one platform became more attractive than the other, for many complicated reasons, attention and activity shifted. People logged in on MySpace less and Facebook more and, eventually, realized they were effectively no longer MySpace users. Anyone that has been on the Internet long enough to watch a few of these shifts from one platform to another knows that they’re not abrupt — even if they can be set in motion by a particular event or action. Users of social networking sites simply don’t have to choose in the way that a person choosing to boot Windows and GNU/Linux does.

First, Facebook is bigger than these other examples. A lot bigger. So much bigger, in fact, that I would argue the difference is of character, not of degree. Metcalfe’s Law tells us that the value of a telecommunications network is proportional to the square of the number of connected users of the system. If this is an accurate way to think about social networks – and I believe it is – then the fact that Facebook now contains almost a seventh of the world’s population is very meaningful. Remember: Ma Bell never died. She had to be killed. And her dismembered bits and pieces are slowly oozing back together.

I don’t think the appropriate comparison for Facebook is MySpace, or Friendster, or Tribe anymore. Those sites remind me, in hindsight, more like Path or Instagram or Pinterest: much smaller, segmented, affinity communities for particular populations. Facebook has become something almost more like electrical current or railway gauges, or even the TCP/IP stack: yeah, there may be global variations, but generally speaking everyone is using the same thing.

Ben was talking about network effects; I’d like to add in the concept of a natural monopoly. It seems to me that it makes sense for there to be one big social network site for everyone. It doesn’t need to be the only one (and Ben makes this point about Diaspora having its own valuable niche). But just in the same way that we’ve realized it doesn’t make sense to have multiple Internets (goodbye CompuServe, goodbye AOL), I’m not sure it makes sense to have multiple social network sites of global aspiration. Facebook is there already. And if Google can’t beat them – with more users worldwide than Facebook has through search, GMail, etc – then no one can in the foreseeable future.

Second, I don’t think the switching costs are lower for social network sites. The Data Liberation Front and Facebook’s archiving system are nice. But they’re also trivial. They’re trivial because the point of privately enclosed social intermediaries is not the things but the people enclosed within them.

Network effects + Metcalfe’s Law means the social network site which contains the largest subset of my set of friends is the most socially useful site to me. There may be design or affinity differences which add a weight to the equation but the general arithmetic holds true. I might have stopped using Facebook a long, long time ago, except that even if I wanted to I can’t; it’s where all of my friends are, where they plan events, where they post links that I find useful / informative / entertaining, and until they are elsewhere I must remain.

That’s a much harder decision to make than what OS to boot into. I can boot my Macbook into OSX / Windows / Ubuntu. For that matter, I can emulate any of them in VMWare. And, because of hard-fought interop battles (and the economics of software publishing houses) I can increasingly rely on some cross-platform compatibility in key software tools. Every day there is less vendor lock-in in the software space that actually limits me. Tableau is Windows only? Fine. I’ll install XP on VMWare and run it under Unity View. To me, the user, it’s transparent. I’m not trying to trivialize interop / OS wars / lock-in here. I just think that, for me, the decision to drop Facebook is about a billion times more difficult than what OS I boot into.

The real thing that is interesting to me, though, is something Ben mentions only in passing, presumably because it is such a complex and difficult question to grapple with:

Over time, as one platform became more attractive than the other, for many complicated reasons, attention and activity shifted.

What exactly did happen here? How the hell did Facebook get so big so fast? How did it penetrate so many markets so completely? And could anyone else do what Facebook did (which it would need to do in order to replace it)?

I don’t know all the answers to all these questions. But I think I’ve got a dim sense of one.

In June 2005, when I was graduating from high school, I remember standing in the high school cafeteria with my friend Erica Getto. We were practicing the graduation walk and chatting idly about life in college. Erica asked me if I had heard of TheFacebook. She had, from her elder brother, and she said that “all college kids are on it” and it was how they kept in touch.

When future historians go back and try to figure out what the key was to Facebook’s success I’d be willing to bet money that it was in their (completely accidental) market positioning. They went very, very deep into the college demographic, which is different from going deep into the hipster demographic or goth demographic or punk demographic because while college students are certainly demographically distinct from their age cohort a lot of different people go to college at one point in their lives. It’s as if rather than Facebook trying to capture one branch of a social tree, they instead caught one four-year long segment of the trunk. The thing is, as time goes on that segment graduates, and then goes out into the world, but still has a Facebook account with all of these ties, and a new group comes in.

In other words the thing that made Facebook a success (I think) is that it deeply penetrated a broad set of people in a narrow set of time / life experience. This meant that people who graduated could still use it, and people who were entering college would get indoctrinated in, and it just grew from them.

Incidentally, I think the same thing is about to happen to Apple. At MIT and at every other college where I have spent time Macs are used at a rate disproportionate to the adult professional population. Most of industry, enterprise, government, etc is still old PC boxes. But Apple has had such deep penetration into a generation of college students that I have to think they will demand (and eventually receive) Macs in the workplace going forward. Anecdotally this has already begun to happen.

Again, this isn’t the only explanation. It’s one explanation in a complex ecology of explanations. But I think there is something to this idea of deeply penetrating a broad-base, time-slice cohort, and then hoping that it grows and networks out as it moves in time. The takeaway is: build a tool that’s useful for a slice, and they, like Rick Astley, will never want to give it up.