The “Open” Graph

Face­book, at this week’s F8 devel­oper con­fer­ence, unveiled sub­stan­tial changes to its “open” graph API. Today, we look at the win­ners and losers in this announcement.

The Open Graph

Before we look at the win­ners and losers, let’s recap where Face­book is going. A cou­ple of years ago, the com­pany unveiled the “Open Graph API”, which gave exter­nal sites the abil­ity to add “like” but­tons to their sites, and in the process allow end-users to things they like into their face­book stream, where they would be shared with their friends. This mean that a user could implic­itly pub­lish infor­ma­tion to Face­book by just click­ing a but­ton. At that point, two things really happened:

• First, Face­book cre­ated a sys­tem that allowed it to bet­ter under­stand what peo­ple were read­ing because every time a like but­ton is pulled down, Face­book can keep track of it and asso­ciate the infor­ma­tion with the account of a Face­book user.
• Sec­ondly, Face­book offered pub­lish­ers a way to gen­er­ate more traf­fic to their site by pub­lish­ing con­tent on Face­book that would return users back to their site.

This was a lit­tle creepy because it gave Face­book a tremen­dous amount of power but, in exchange, the com­pany gave pub­lish­ers access to more traf­fic so the agree­ment seemed bal­anced to every­one but the end users (which doesn’t mat­ter, as we all know that “if you’re not pay­ing for it, you’re not the cus­tomer; you’re the prod­uct being sold.”)

But it seems hav­ing access to that data is not enough for Face­book; now it wants more data so it can build bet­ter elec­tronic pro­files of its users. So this year, the com­pany has decided that “Like” was not enough and they wanted to get users to give them access to any­thing they read on a par­tic­u­lar site. So the com­pany unveiled actions like “read, watch, lis­ten” which allow devel­op­ers to share all action data from a user after a one-time sign-up to the Open Graph API.

Min­ing the web… but not giv­ing back

This time, the com­pany is min­ing the web but the main ben­e­fi­ciary of those actions seems to be Face­book, which gets a bet­ter under­stand­ing of where users are when they are not on Face­book and what they are look­ing at. If we assume (and I am will­ing to take the bet that this will be the case) that most main­stream sites will start offer­ing the new verbs, then Face­book will have one of the most com­plete under­stand­ing of a user’s demo­graphic and psy­cho­graphic pro­file. In other words, the data it will get access to is the holy grail of online mar­ket­ing: users that can be tai­lored to based on extremely gran­u­lar preferences.

That data can then be used to send ads that res­onate with the user in the chan­nels the user accesses most. Think this is crazy? Well, wait ’til next year’s F8, when Face­book unveils a tool to help exter­nal sites mon­e­tize their traf­fic bet­ter by tar­get­ing adver­tis­ing based on users’ pref­er­ences. If this sounds sus­pi­ciously like AdSense from Google, it’s because it is part of the end-game. Face­book is no longer happy to have the largest site in the world, now it wants to have access to peo­ple when they are not on Facebook.

What’s fas­ci­nat­ing here is that the data goes into Face­book but there is pre­cious lit­tle infor­ma­tion as to how to get it back out, mak­ing the word “open” a head­scratcher as it is unclear how Face­book defines open­ness. To the Palo Alto com­pany, it appears that open­ness is a one-way street: you open up your data to Face­book and in return Face­book “sim­pli­fies” the online expe­ri­ence by keep­ing your app on its plat­form. This is some­what sim­i­lar to the app store model offered by Apple and Google, where it’s OK to play as long as it is within their rules. Face­book is doing to the open web what Apple and Google have often been accused of doing, stick­ing another knife into its immi­nent demise.

Also of note is the fact that Facebook’s approach to get­ting all this data makes it impos­si­ble for any­one to cre­ate valid HTML5 pages as the Face­book code does not val­i­date under this frame­work. So Face­book is also ham­per­ing the future of the web by mak­ing it nearly impos­si­ble to live by the ideal of the new web stan­dards if you want to play in the face­book arena. This seems to shape up another fight between Face­book and the open web.

Another com­pany has had sim­i­lar ambi­tion and it staked its approach first on offer­ing supe­rior search prod­ucts and then on using the search data to tar­get adver­tis­ing on part­ner sites and even­tu­ally offer such capa­bil­i­ties to any­one who was will­ing to give them a per­cent­age cut of ad rev­enue. That com­pany, Google, has real­ized the lim­i­ta­tions of its model and is busy try­ing to ensure it can get more data by build­ing up offer­ings for ways in which peo­ple access the inter­net: so they’re push­ing Android for mobile phones and Google Chrome as a bet­ter web browser because they want to be able to access data relat­ing to where peo­ple are on the web, data that can then be used to cre­ate more cus­tomized ads.

Face­book at $150 billion?

This week, Google’s val­u­a­tion sits around $150 and the high­est Face­book has ever been rumored to be worth is $100 bil­lion. I’d ven­ture that peo­ple are sell­ing the com­pany short and that it is worth some­thing on par with Google. It has mas­ter­fully played fears from online pub­lish­ers and other sites and par­layed that in a poten­tial posi­tion of power in the online mar­ket­ing world.

The only thing that could make it more pow­er­ful than it is about to be is if it were to pair up its data with Google’s. A merger of the two of them would cre­ate an unpar­al­leled data­base of inter­net users, con­tain­ing just about any­thing about people’s intents (from Google’s search), their inter­ests (from Facebook’s data), the amount of time they spend on cer­tain prop­er­ties over oth­ers (from either Face­book or Google’s data), and what they liked enough that they would share it with peo­ple their know (from Facebook’s data).