How Facebook’s Open Graph will own identity and threaten Google

Facebook today announced some potentially ‘net-changing features they are releasing under the moniker Open Graph.  Open Graph replaces Facebook connect, or perhaps deprecates it if you like, making it easier for people to utilize their Facebook data within the context of other websites.  Sounds fancy, eh?  Let’s break it down into understandable examples:

open-graph-stream1Most prominently, websites can embed “Like” buttons on their pages, just as Facebook has on its activity feed items and various other pages around their site.  Website creators will embed these Like buttons because it lets their users publish links they like back to their Facebook feed with a single click – they don’t even need to sign in to the creator’s site, as long as they are already signed in on Facebook - free marketing for the website.

In addition to the Like tool, Facebook offers a variety of other “social plugins” to help site creators make their sites more social and more integrated with Facebook.   The Activity Feed lets users see what their friends are doing on the creator’s site.  Login with Faces shows a user which of their friends are already members of a site and prompts them to sign up with that site to connect with them.  Comments lets users comment on individual items on the creator’s site, and gives them a seamless option to post that comment back to Facebook as well.  All this without having to create an account on the creator’s site.  You get the picture.

Facebook = identity

The most significant immediate implication of Facebooks Open Graph is that site creators may no longer bother having their own registration systems at all, as FriendFeed founder Bret Taylor (now with Facebook) explained.  My interpretation is that Facebook wants to own identity on the web, and site creators are likely to step in line because Facebook has made it in their best interest. If creators adopt the tools, they get free marketing tools and a seamless experience for their users.  All they sacrifice is having to share their user data with Facebook.

Open Graph is easier to implement than Facebook Connect, and people can start interacting with sites immediately, no login required, which is great.   It appears that Facebook will share users’ “basic” information once they connect on a site, so creators get names, email addresses, genders, etc. – all the basic things they would ask for anyway.  Easy for the user, easy for me, everyone wins.

Especially Facebook.  Once Open Graph plugins become widespread, Facebook will know exactly what users are doing…all the time.  They’ll know what sites you visit, and they’ll know what things you Like on those sites.  Who knows what evil lurks in the hearts of men?  Big Brother Zuckerberg knows.

Spam bait

Interesting note for you privacy fans: it appears that any data you make public on your profile (which is most of it by default), including things and sites you like, will now be available to other sites so they can tailor their content to your tastes.  Cool? Yes.  Spooky? Yes.  Ripe for abuse? Most definitely.  While I don’t have a problem with this personally because I am pretty careful and sparing about what I share on Facebook, a lot of people are going to get stung, and spammers and direct marketers will try to abuse the system to deliver unsolicited ads.  I wish Facebook well, but this is going to be a hornet’s nest.


The data

Now for what I think is the real meat of all this: the data.  When site creators implement these Like buttons and other plugins, Facebook is encouraging them to tag their pages with specific types of common metadata that may be relevant:  image, name, location, email address, phone number, and “type” (e.g. sport, activity, restaurant, athlete, city, product, book, blog, website, etc.)  If creators take the time to tag their pages like this, then when their users “like” something, Facebook will know exactly what it is and can present it nicely within the Facebook context.

Think about this for a minute. Suddenly, one organization on the web has the ability to know what pages are about without having to crawl every page (and its backlinks) to figure it out.  Site creators are telling Facebook exactly what their pages are about using structured data.  Here is the quote from their Open Graph page that jumped right out at me:

Based on the structured data you provide via the Open Graph protocol, your pages show up richly across Facebook: in user profiles, within search results and in News Feed.

Search results, eh?  Any page on my site that I tag with structured data can show up in Facebook search.  Facebook could presumably let their users filter the search so it’s for “actors” or “politicians” or “athletes” or whatever type of object.  They can search for activities, landmarks, restaurants near their current location…  This sounds an awful lot like Google, but with 1/100th of the effort that Google goes to when compiling their monstrous index of every page on the Internet.

Even better, all these links are ranked by humans.  Every “Like” button that we press makes this massive index of webpages and real-life offline things smarter.  This is getting impressively close to the holy grail of search: social search.  Not only is it vetted by humans, but it’s real-time – no need to wait for a crawler to poke around every corner of the web.  The best of Google search with the Best of Twitter search in one package.

Google is surely watching these development keenly, and probably wishing they had acquired Facebook back when they had the chance.  Microsoft is surely dancing a jig.  (Hey Stumbleupon: love you guys, but it’s time to pack up your bags and go home.)

If I was Google, I’d give an arm and a leg for all this data.  With Microsoft being a major investor in Facebook, don’t be surprised to see this data integrated into Bing in the not-too-distant future.



  1. I think it will help google. How hard is for google or any search engine to pick up these additional tags and use that in the index. So potentially it can complement the search engines for the better.

  2. Google can crawl those tags, and they probably should. What they lose out on is the real-time aspect of that data. Google may see the tags days or weeks after they go up, while Facebook has already seen them, figured out how important they are from user feedback, and integrated into their search accordingly.

  3. What bothers me is not just the clicking of the “Like” button, but the fact that the mere presence of it on the page allows FB to track user browsing habits completely without their consent (or even knowledge).

    I could be wrong, but I explain more here:

  4. Certainly. There are a lot of things to be concerned about with this program…hehe.