Facebook for SEO

Are you spending valuable time thinking about FaceBook and SEO? If so, then read on my friend and we'll have a look-see if that time has been well spent.
First things first, I don't really like to think of FB as a search engine per se. At least not a web search as we're traditionally accustomed to. This is not a cat fight between Google and FaceBook. A traditional search engine crawls the web, locates, indexes and retrieves information. FaceBook is more of a 'site search' application than a (web) search engine. Sure, splitting hairs, but worth noting. So please, let's stop with this 'Google Killer' hype/crap, m'kay?
Ever since they announced the open graph and search enhancements we've been hearing more and more about the world of FB search and of course, optimization for same.
What we know about OpenGraph
Ok, so from what we do know the results are being partially ranked/scored based on 'likes'. These are considered a 'vote' not unlike links are to Google. We can then infer that 'like-bait' will be the currency in FB SEO as 'link bait' is with (web) search engines. This also means that 'like spam' will be the black hat currency as well. Is this a functional scoring system? Not in the least, they'll need some better tricks up their sleeves if it is ever to be truly useful.
We also have a bit of semantic code at play here, not dissimilar to what we know already with RDFa/micro-formats. Some of the supported OpenGraph types are;
  • activity
  • sport
  • bar
  • company
  • cafe
  • hotel
  • restaurant
  • cause
  • sports_league
  • sports_team
  • band
  • government
  • non_profit
  • school
  • university
  • actor
  • athlete
  • author
  • director
  • musician
  • politician
  • public_figure
  • city
  • country
  • landmark
  • state_province
Products and Entertainment
  • album
  • book
  • drink
  • food
  • game
  • movie
  • product
  • song
  • tv_show
These are used in the form of (the ever bloating) meta data in the elements of your code. This means there are a wide variety of targeting abilities available. Problem being of course that many common CMS out there aren't really equipped to fully max this out on a page level approach. Not ideal at all.
It is interesting to note that IF this were to be a 'Google Killer' it seems to avoid the simple fact that Google already parses RDFa/Microformats and as such parsing this code, wouldn't be all that hard. Thus Google can readily USE this data....not just FB.

What's the Value of FaceBook SEO?

That my friends, is the real question here. So far, from what I can tell, this really isn't about crawling actual pages and ranking them based on relevance or any of the hundreds of signals the web search cousins provide. It seems heavily skewed towards the meta data provided. This does limit it not only in terms of optimizing, but also in terms of the efficacy of the value to the end user – the holy grail of all search.
On the other hand, there is a massive user base which we may not want to dismiss quite yet.
Next I looked at a few actual search tasks...
FaceBook Search

First, we find that it may never get to an actual SERP because of the drop downs.... and seriously, an 'exact match' search engine isn't revolutionary by any means. This leaves it open to poor results and easier spam-ability.
And the actual SERPs? Not impressive either. We have some exact match, then 'posts by friends' which wasn't enlightening and of course, 'web results' which are powered by Bing... all in all, it doesn't seem like this is a Google Killer and surely seems to have little to do with meta data as it does straight match and 'like' spamming
FaceBook SERP 1
FaceBook SERP 3
Now, since this is more of a transactional query space, we'll refine with the 'pages' results set.
FaceBook SERP 2

Man, this looks like straight exact match and some 'votes'. Not exactly the most full bodied search result that I've seen. In fact it kinda reminds me of search engines some 10 years back... this would likely be VERY easy to game. What is the damned use if the results to the end user are going to suck ass?
I spoke to someone in the 'dark arts' about spamming FB and asked what safe guards were in place and only got the reply, “what safe guards?”. Kinda says it all.

Going Beyond the basics

Now, all of that being said, it does seem to get a little more interesting as you drill down into the various elements of the search including;
  • People – does a reasonable job of returning results, though exact match is too heavily weighted. Also skewed towards 'friends and 'mutual' friends, which makes discovery problematic
  • Pages – as noted, seems to have some issues that would make topics outside of your social circle somewhat easy to game.
  • Groups – again, outside of one's social circle, the results can be downright useless and relevance signals? Hardly any to be found
  • Applications – doesn't seem to be enough of them to be valuable. But from a marketing perspective, this might actually be a good thing (lots of virgin query spaces)
  • Events – see 'applications' lol... ghost town.
There really does seem to be the most potential here in that there are some ways of searching that we'd not find in most traditional search engines. What is a little more problematic is the lack of quality results (often due to folks not capitalizing) and a somewhat clunky interface. There is a glimmer of hope here.
We shall have to see how it evolves, at this point, it really isn't something I'd find myself using, if I weren't a search geek that likes to play with search engines

Should you be worrying about Facebook with your SEO?

Well, I am a huge fan of 'task ROI' in ones efforts, in short, what will the returns be on one's SEO efforts. First we must consider that FB only accounts for some 2-3% of all searches. That's even less than Bing-Hoo. So let me ask you; how much time to you spend optimizing for Bing/Yahoo? I suggest that you spend 1/3 of that time frame on FaceBook SEO. Oh and for the record, I spend exactly 0 time on that. Sooooooo...
One of the only potential benefits is that people tend to trust their social circles more than an open search. A small consolation at best. One has to consider the poor quality of the results as a serious hindrance to it's adoption at this point.
At the end of the day, if you're already using FaceBook for marketing, then it really can't hurt to keep one eye on the targeting you're doing from an SEO perspective. Furthermore, we know Google is interested in annotations (semantic mark up) and it may well be worth considering the OpenGraph elements where technologically feasible.
And there is my final advice; enhance existing social campaigns. Don't start targeting FaceBook as a standalone SEO strategy. That time is still likely best spent on Google... cause this puppy ain't gonna kill 'em just yet.

What do you think? Is it worth worrying about at this point? Sound off in the comments.

Posts of interest;