http://en.wikipedia.org/wiki/List_of_social_networking_websi...
Facebook has the most users in one place, but the dust is nowhere near settled on where everyone will end up.
This bodes very well for a decentralized solution. I doubt Diaspora will be the first (I think Appleseed (http://opensource.appleseedproject.org) is much more likely), but ultimately, it will be an easily implemented protocol with mature cross-platform libraries that'll really take it down.
This is a good PR move on Facebook's part, because it willfully misinterprets people's frustrations with Facebook. It's not that you didn't have a local copy of your data, it's that you had no control over your data when it was out in the wild. This will take some wind out of the sails of the decentralized solutions, but the armada is still coming.
It seems like the majority of the non-Facebook figure comes from Chinese language sites. Since social networking sites will quite likely be segmented by language for the foreseeable future, I don't think this really counts. In the English-language world, Facebook has serious network effect and you would be hard-pressed to combine all of FB's competitors into anything.
I agree dispersed social networking is coming... I'm not sure if "control" of data is the way to phrase the "problem"... Still, creating a dispersed network where each contributor controls where their data goes is a very hard problem.
Habbo + MySpace + Bebo + Orkut + Friendster + Hi5 = 679 million. Assuming 50% cross-over (which is a very high estimate), that means 340 million unique users, for just six social networks.
And that is nothing to sneeze at.
The sites I hadn't heard of aren't Chinese language sites but failed English language sites. But they are still failed sites. I don't go to my two Myspace pages, ever.
How much of that overlaps with Facebook? Approximately 500 million? Also, they're measuring total registrations: FB has probably close to a billion registrations on their 500 million actives. All that statistic tells you is that FB has room to grow.
As for active versus registrations, we can't be sure of that from anyone, but rest assured, a decentralized network has much greater potential than Facebook does. Especially in the rest of the world where Facebook isn't nearly as popular.
I'm not thinking 'OMG now I can convert users!', I'm saying, 'Cool, I can possibly evolve what a social network can do for me and a group of friends, while working with the FB ecosystem.'
Facebook will have to build features for the masses - we can build social networks for the subsets of users. We can convince users to download their piece of the social graph by offering a unique niche experience, and getting creative about what it means to interact socially online for the subsets of the masses.
Today's announcement was more than a 'high-road' PR stunt. To extend the linux analogy, FB just made it easy for each user to create a virtual machine where we can run a niche social network side by side with the gigantic ecosystem that is facebook.
Alternatively, there may be a parallel to IBM giving away the PC specs, inadvertently pushing the monopoly down to the OS and weakening their market position (though creating a tech boon for everyone that made the market so much bigger).
All the more reason to use Diaspora or (X).
The biggest advantage Facebook has is that it has your data, being able to export it changes the game.
The biggest advantage Facebook has is that 500 million people have an account. Facebook would be utterly useless if nobody was using it. Network effects make Facebook hard to beat, not your data. They need your data only to make money but they get your data because you and your mom and everyone else have an account.
When everyone moved from MySpace to Facebook they had to re-do their profile in the new system. Hopefully when FooThingy replaces Facebook importing will be just a click away.
Everyone didn't move from MySpace to Facebook. Facebook pulled in many, many new people.
But if something does replace them having an API to export the existing data will be critical to the adoption of the new system.
Even if all your friends use Facebook now, having an export API is the difference between them clicking on one link (and "approve") to move your data to the new system, and having to start all over again.
Lowering the barrier of entry will do a lot to increase competition in this area. Hopefully Facebook will institute meaningful data export policies like Google has done.
I think the benefactors of stuff like this are sites like flickr, blogger or tumblr, since it seems you can export out your photos/notes and start a new photostream/blog.
Somebody has always said something like this about some predecessor.
Myspace was most popular among high school kids, indie music culture, and people looking for dates. Many people in their 20s and 30s started using it to keep in touch with old friends and make new contacts but it never grew beyond that. By the time they implemented the news feed, it was too late. Facebook had already gotten the momentum.
Now, people (and organizations) of all sorts have Facebook profiles and pages. Families, adults, colleagues, friends, new acquaintances, all use Facebook to stay connected. Sure, the kids will almost certainly find something new to play with soon, but that won't matter anymore.
Facebook is huge, and now pretty clearly shows enormous potential to continue growing.
Users still aren't going to leave Facebook. Facebook realizes this, which is why they can do this now. Your mother's cousin is on Facebook. She won't be on Diaspora or (X). That said, this is pretty cool for its own reasons. Like waterlesscloud said, it looks like they "want to be the source of all social interactions on the web without having monolithic control."
This is definitely a smart move for Facebook -- be willing to give a little on data portability to position themselves as open.
Facebook has a gigantic ecosystem. Thousands of sites use Facebook Connect to provide social context and are predicated on everyone interesting have a Facebook account. A social network is a very complicated webapp -- probably the most complex many of its users interact with on a daily basis -- and literally hundreds of millions of users don't want to have to re-learn that. Remember that we're talking about here: take, for example, the http://www.readwriteweb.com/archives/facebook_wants_to_be_yo... incident. They have such tremendous inertia that even being able to one-click import your data to another app isn't likely to damage them.
This move is kind of like Jobs saying, "all smartphones have problems. We're not perfect. But we're trying to do the best of anyone." At least for the moment, it's changed the discourse from "wow Facebook supports developers terribly" to "hey, Facebook's not entirely a walled garden!": what Scott Adam's called the "high ground maneuver".