SEO Implications of Facebook's Open Graph Search Engine

It was revealed on bnet recently that "external websites have begun to show up in Facebook's network search results" which the blog AllFacebook have referred to as "the first signs of an Open Graph search engine".

What's an Open Graph search engine?

Basically, it is a search engine based on a social graph (a map of all the social connections online) instead of a link graph (a map of all the website connections).  Most of us know that Google's algorithm is heavily based on determining the authority of web pages by treating links from one page to another (part of the link graph) as a vote of sorts.  A theoretical flaw in that is the number of people with the power to cast a vote is a lot less than the number of end-users (visitors).  And by looking at the social graph of the web instead of the link graph, Facebook figure they can hand over the vote to the average users who not only make up a much larger group but are also the target audience.

As such, you cannot for a second consider this to be just another "me-too" search engine.  This is something quite different, to which we really must pay attention.

It appears as though this is something Facebook have been working towards for a while now.  Earlier this year, the "Become a Fan" button on Fan Pages was changed to a "Like" button.  This change didn't make a lot of sense at the time and attracted a lot of complaints but it would seem that as well as fitting more closely with terminology elsewhere on the site this wording may have been chosen to encourage people to use the button more freely - it feels like a bigger commitment to "Become a Fan" than to simply "Like" something.

Not long after that, the "Like" button was released for use on any web page rather than just on Facebook's fan pages.  This was a key piece of the puzzle; pulling external web pages into their social graph allows them to rank those pages based on the votes from that social graph.

SEO Implications

While we are talking about a fairly unique search engine there are still some basic principles that are unchanged.  We still have to think about page authority and reputation, where reputation determines what queries a page will rank for and authority determines the likelihood of ranking.  So an Open Graph Search Engine is unique in how it gets those signals but certainly not in its need for them.

In the following table I've summarized what data Google and Facebook each use to determine the reputation and the authority of a page.  The differences are actually quite significant.
 

While Facebook may be giving the voting power to the average user rather than just site owners we can see there may be some quality issues with their algorithm as it stands now.  First of all, we see that Google's reputation drivers include both on-page and off-page data whereas Facebook determines the reputation of a page solely via data provided by the site owner.  Secondly, while Facebook's only driver of authority is the quantity of votes, Google recognizes that not all votes should be considered equal online and incorporates various signals to determine the quality of each vote. Without these quality controls, which make up a significant portion of Google's algorithm, I fear Facebook will be leaving their search engine wide open to site owners "gaming the system".

The other potential flaw I see right now is that search results will be seriously favoring the types of things people like to "Like" via Facebook.  For example, while I am clearly very interested in internet marketing I rarely (if ever) "Like" anything I read on the topic because I know that only a few of my Facebook friends are interested.  Instead I might tweet it and/or add it to my LinkedIn updates since I am connected with business associates through those channels.

Despite these possible problems I think this is a very exciting development from Facebook and it will be very interesting to see the quality of results they are able to produce from the world's first Open Graph Search Engine.  And let's not forget the enormous number of improvements that have been made to Google's algorithm over the years; most of which have been about fighting spam and other quality issues.  So let's give Facebook a chance and see where they can take this thing.

As someone interested in online marketing you should be excited at the prospect of another great channel through which you can reach your target audience.  The best part is that this one doesn't actually require too much extra effort but given the size of Facebook's social graph might just reap some big rewards.

Actions

Here's what I think you should take away from all this.  Start giving Facebook what they want immediately.  In other words, go and familiarize yourself with the Open Graph protocol and add the meta data they are looking for to your site.  While you're at it make sure to add "Like" buttons to your pages.  This can be achieved with the code provided on that same page or with plugins for many of the major content management systems.

Once you've done that you'll want to start getting creative in terms of how you can get more users to "Like" your pages.  Aside from making the button more prominent and asking people to use it you could consider incentivizing its use with competitions or even free downloadable gifts (not unlike an opt-in offer).

And finally, you will want to monitor your results through your web analytics.  Are you seeing more traffic from Facebook?  Is it converting?  What trends can you spot?  Are you able to monitor which Facebook searches are producing results for you?

Oh, and one more thing, please "Like" this post if you enjoyed it.  ;)