There’s No Stopping Facebook

The phrase tour de force is frustratingly over-used by movie and restaurant critics, but that’s exactly what Facebook’s f8 conference on April 21 in San Francisco was. The company convincingly laid out an ambitious plan for web domination, pleased web developers with straightforward and far-reaching tools, and showed that it is a capable and well-run business. Competitor Twitter’s developer conference, held nearby a week earlier and fraught with tension over the company competing with developers, was amateur by comparison.

However, Facebook’s weak spot continues to be privacy. Personal information management is a necessary component of building a service that makes the web more like real life — where we have complicated relationships, secrets, and can be held accountable for our actions. But every time Facebook presses forward with new ways to share information online, it gets tripped up on privacy.

That’s for three reasons: 1) making personal relationships and access rights explicit is awkward; 2) people would rather use a service than fiddle with the settings (even with all its alerts and widely publicized changes, only half of Facebook’s users have ever changed a single privacy option); 3) Facebook is trying to get people to share information and activities in ways they’ve never done before. (Twitter, by comparison, has it incredibly easy; nearly everything is public by default, and that’s that.)

So here’s what happened last week:

Facebook’s CEO Mark Zuckerberg and director of product Bret Taylor announced three major initiatives: social plugins, what it calls the “open graph protocol” and a “graph API.” All are aimed at enhancing the relationship between Facebook and other web sites. Now, any web site creator can make their site instantly social by using the relationships and communication channels Facebook has already built with its 400 million users. Here’s what you need to know from the keynote:

First, social plugins are little widgets that bring Facebook to the rest of the web. They offer “instant personalization,” said Taylor, with the goal of increasing user engagement, using an iFrame and a cookie remembering the Facebook user. So when you visit a website, even if it’s new to you, you’ll see which friends have also logged in there, what their activity is and a set of recommendations based on their actions.

One action in particular will be closely tied back to Facebook: the like button. If you indicate you like an article, a band, a restaurant — anything, really — a site using Facebook’s open graph protocol can create a persistent relationship with you around that content. Sites give Facebook semantic information around the thing you liked — for instance, the title, type, genre and city for a band you like on Pandora. Then that band goes straight to the favorite music section of your profile. Same thing happens if you like a movie on IMDB, another launch partner.

The objects that you like are first-order citizens on Facebook, said Taylor. So if another user hovers over that movie you liked, they see information brought from IMDB. A click goes back to the source. If a user searches for restaurants on Facebook, the top things that show up in Facebook’s own search could be restaurants your friends liked on Yelp. And the sites can communicate back directly to that specific subset of users who have liked something. So when Stanford football star Toby Gerhart gets drafted tomorrow, Bret Taylor could automatically see that information in his feed.

Lastly, Facebook’s Graph API aims to make developing on its platform much simpler for the long haul. Every object on Facebook has now been given an easy-to-formulate, unique ID. The API will allow sites to search user updates and get real-time updates every time a user adds a connection or posts on a wall. Developers, with permission, will be able to hold onto user data for more than 24 hours. And Facebook will be adopting the open authentication protocol OAuth.

Reactions from industry observers immediately after the keynote were admiring but also wary of the power Facebook was awarding itself. Facebook employees told Om they weren’t stressing about any additional load on their infrastructure. We wondered: where were the location features Facebook has clearly been thinking about and working on?

Then we started delving deeper into the implications of what Facebook had announced. My two critiques were:

  1. Facebook, by inserting itself into the workings of 75 launch partner sites and many more since, has made itself a point of failure for the web.

    Organizing the world’s information by powering it is clearly a direct affront to Google. Where Google observes links and relationships between web sites from a distance, Facebook aims to be the glue that connects the web itself. The implications are thrilling, but also scary — what if Facebook goes down?

    That concern was realized by Friday, when Facebook went down for many users, including partner sites, for an extended period of time.

  2. One underemphasized feature, which Facebook calls “instant personalization,” gives carefully selected partner sites the ability to connect a new user with their public Facebook profile information, as well as those of their friends, before they have ever logged into the new service. The idea is to make unfamiliar sites magically personalized, but it makes the web much less anonymous than we’ve come to expect. I wrote on that at length here.

By Thursday and Friday, developers and users had processed the changes and some started taking action to combat them. A group of web developers including Hunch co-founder Chris Dixon launched an effort towards an “Open Like” standard to compete with Facebook’s proprietary “like” system. (Perhaps that proves just how right-on Facebook’s innovators are — that its competitors would feel a concept they’d never heard of until Facebook invented it was so important that it needed to be open source.)

And meanwhile, Mathew wrote a readable guide to protecting oneself from Facebook’s new features for those concerned about their privacy. Some users welcomed it with open arms — including a contingent of Google engineers who apparently deactivated their Facebook accounts using Mathew’s directions.

But even if Facebook is forced by public outcry to scale back on some features (and it’s not clear at this point that such a thing will even happen) the fact is, the company has the critical mass of users and the clear product vision to make the vast majority of its plans happen. And by pushing relentlessly forward with its ambition — even when it oversteps, like by giving advertisers user action data with Beacon nearly three years ago — Facebook gets more done than any other web company with the exception of Google. We journalists may nitpick where we see weakness (hence the GigaOM stories on open-source alternatives, privacy, and outages) but Facebook is pretty rock solid.

So that’s the week in Facebook. But the company’s convincing momentum didn’t come out of the blue. To help put it all in context, here are five pre-f8 stories we wrote about Facebook’s ambitions and the competitive landscape.

Please see the disclosure about Facebook in my bio.

Question of the week

Who or what could stop Facebook from its ascendancy to the top of the web?