ADOTAS – When late last year Bing announced its partnership with Facebook, the news sent the SEO world ablaze with questions and predictions on how this would affect SEO and rankings.
While interested in the direct correlation, we felt that that the larger story was really the potential of Facebook becoming an individual’s (and a corporation’s) go-to social network hub and search engine. With approximately 2.5 billion searches per day, most people were leaving Facebook to search with Google, but with the Bing integration, leaving Facebook and choosing another search provider would no longer be needed.
Now, several months later, we are asking another question: Do pages with a lot of shares and likes from Facebook really rank better in Google? To verify this, in April, Rand Fishkin of SEOmoz analyzed various data points that did show an existing correlation between Facebook Shares, Likes and Comments; leading him to the conclusion, that yes, there is an impact of Facebook on the Google ranking.
Using this finding, we checked it against our own Searchmetrics data from the U.S. and Germany. Using the first 30 URL results of 20,000 German keywords with the biggest search volume in the German Google index, we found some additional interesting correlations.
Let’s first take a look at the correlation between Facebook Activities and the Google Ranking data:
For the x-axis, we took the top keywords and their corresponding URLs to show keyword positions on Google. The y-axis shows the percentage of Facebook activities, including all likes, shares and comments, coming from those URLs with keywords in positions 1 through 30.
What is pretty obvious is the direct correlation. You can see that most of the keywords in top positions also have the highest numbers of shares or likes. Keywords in the top 10 positions actually have 50% of all of the shares, likes and comments and this significantly drops for keywords in lower Google positions.
If you drill deeper into the differences between the number of likes compared to shares and comments, you find even larger differentiations. Keywords in position 1 have 93% of shares and comments with approximately 76% of likes.
We ‘Like’ the Short-Tail
Most of the likes are based on short tail keywords, but this is not all. As you can see in the first chart, the graphs of the Top 100, top 1,000 and up to the top 20,000 keywords do differ.
Since we usually find in the top 100 keyword list short-tail keywords like “car” or “iphone,” usually brand homepages rank on the first position and they usually attract the most shares on Facebook. In the German index you also find the U.S. homepages of the brands, mostly a few pages lower than the German versions, but they are also included in the graph.
What is very obvious is the drop of the Facebook activities with less popular keywords. The more we move into the long-tail, the less shares, likes and comments can be found corresponding to those words on Facebook. In the short tail, you also find more homepages and less subpages. This could also lead to the assumption that people tend to share homepages rather than subpages.
Not Everyone ‘Likes’
We were surprised to find that although Facebook can be found on a lot of pages, there are still many website pages without Facebook Like buttons. For example, on Wikipedia you won’t find like or share buttons for Facebook. In addition, for a lot of major newspapers you can find the like button below the articles, but not on the homepage. Since the figures state that 93% of shares mostly include homepages, this means that users share the homepage on their own, without the need of a Facebook button on it.
It’s the Chicken! No, It’s the Egg!
But even a result close to 100% is not concrete evidence that Facebook impacts the Google ranking. It can still be a correlation, or it can mean that Google is already taking Facebook activities into their algorithm.
At SMX in Seattle, Matt Cutts answered this question of Facebook correlation with the argument that Facebook won’t let Google get their data; an answer that does not really say no, but leaves a lot of room for discussion.
For one, could Google get hold of Facebook data via the use of their Google Chrome browser or the Google toolbar? This is just one of many questions concerning the dynamic between Google and Facebook, what we know and what is possible.
So the question remains: Is Facebook influencing Google or vice versa? Does a keyword rank highly because it was shared, liked or commented on frequently or was it shared, liked and commented on frequently because it ranks highly on Google?
It’s clear that this will be an ongoing discussion and with so many variables, the answers will continuously change. We all know that sooner or later the “search machines” need to take the user behavior into consideration, and this has to include Facebook shares, likes and comments.
And of course, this will always be just one of the many ranking factors for Google.