The Technology of Connections

The other day, amidst a lot of excitement around Microsoft and Yahoo’s potential future, Google released a little project with a decidedly small amount of fanfare. In fact, the project’s home page looks like it was tossed together in an afternoon and features a somewhat goofy YouTube video of the engineer explaining the technology in front of a whiteboard. The pages look more like an undergrad project than a release from a multibillion-dollar company. For Google, I supposed that’s on brand.

The project is called the Social Graph API. Essentially, Google’s Web crawler has been paying special attention to public-facing profile pages on social networks, blogs, and services such as Twitter — basically any service where you have a profile and connect with other people who also have profiles.

These social sites have been slowly adopting a couple of standards in how they mark up these profiles and friend connections: XFN (define) and FOAF (define). Both of these are sets of XML tags, which are bits of code, strategically placed on Web pages that explain the type of information being displayed. XML is extremely helpful in organizing the Internet. It’s the method by which a Web page author can communicate what sort of information is on her site quickly and automatically.

The Technology of Connections

The XFN standard introduces a new bit of code to the Web: relationship. If I’m a blogger and want to insert a link to my friend Ed’s blog, I can tag that link with the code REL= “friend.” More interesting, I can even specify whether I’ve actually met Ed out in the real world. If I’ve met Ed, I’d change the code to REL= “friend met.”

FOAF begins to unlock the power of the Social Graph API. FOAF introduces a new bit of code: “knows.” You can literally place this bit of information on a Web page about a person (such as a MySpace page), which communicates who a person knows. Imagine you have two of these profile pages: On one, Gary knows Ed. On another, Ed knows George. I merge these two together, as Google has done, and I can determine that Gary is one connection away from George. George has a FOAF relationship with Gary.

Structuring the Social Web

This sort of information fuels many powerful sites. If you use LinkedIn, the box that shows you a list of people you may know is built on this same principle. But as Google begins to pull this together and make it public, it represents an amazing opportunity not just for people but also for brands doing business on the Web.

The first phase of marketing and doing business on the Web was about broadcasting messages over the new medium. The second phase was about building functionality that tied back to your brand. The third phase is about participating in the human communities make up the Web.

But businesses need to operate on structure. It’s impossible to be effective on an ongoing basis when you don’t have a real sense of what environment you’re in. Google has taken an amazing step toward giving us that structure.

The Big Deals

Let’s take a step back. There are lots of strange bits of technology flying around the Internet. Of all of these, why is this particular one interesting to marketers? There are a few reasons. Google’s move is important in its own right, and it signals two things that online marketers need to consider.

First, connections equal content, which equal ranking. Google’s business is about capturing and cataloging content. In particular, it uses links as votes of relevance. If more people link to your page about Newton’s Second Law of Thermodynamics than to any other page about Newton’s Second Law, you win. Your page is the most relevant for that topic.

Now Google’s looking closely at declarations of friendship as a link. Do you have a MySpace page for your brand or company? Better make sure you have some friends. Potentially, the crawler could view a friend-free page much the same way as a Web site no one links to, sending it way down the relevance scale.

Second is knowing your friends. Say you have that MySpace page and you have a bunch of friends. If all you have is a list of people who linked to you or, worse, just a raw number of “friends,” all you have is a bunch of unqualified traffic. The Google system could potentially allow you to send a query out for everyone who connects to you and immediately learn who else they’re connected to.

Of course, you still won’t know if these people will actually buy from you or not. But you could get a very good sense of what the community — the environment — you’re operating in looks like. You can begin to grasp the structure from within which you need to operate.

Making This a Reality

Currently, this scenario is a bit imaginary. But the code is now available for in-house technologists to start playing around with. Marketers’ call to arms is to get a basic handle on what’s going on under the hood of social networks and begin requesting their programmers find ways to provide that structure.