Intro from Rand - this post comes from Dr. Matt Peters, SEOmoz's data scientist. He came on board in February of this year after stints at Harvard (working on climate science models), Washington Mutual, JP Morgan and Fannie Mae (analyzing mortgage securities) and more (including some research into Google Places rankings last November). Matt's particularly passionate about bringing the best practices of scientific and quantitative analysis to the world of inbound marketing, and I'm very excited to welcome him to the Moz community.
One of the most interesting findings from our 2011 Ranking Factors analysis was the high correlation between Facebook shares and Google US search position. In fact, Facebook shares was the highest correlated single factor among the 100+ factors we examined:
This blog post presents some modeling results that look at whether Google may be using Facebook shares directly in it's relevance calculation, or whether the correlation between Facebook shares and search position is coincidental, aka the byproduct of other causal factors.
Correlation and Causation
As we have said time and time again on this blog, in our presentations and when speaking, correlation is not causation. However, this post will discuss issues of both correlation and causation, so for the purposes of this discussion it's important to understand the relationship between them on a deeper level. Correlation does not, in general, imply causation. However, two things that are causally related will often be well correlated. Correlation can therefore only be used to support or deny causation, but not to show it directly. Put another way, if we have some prior, a priori, reason to believe that two things are related, correlation can be used as a tool (with rigorous statistical underpinnings) to support the relationship. In turn, weak correlation can be used to weaken the argument that two things are related.
Before we started our work on the 2011 Ranking Factors, we had some reasons to believe that Facebook data may be used by Google. There was an interview with Google/Bing in December 2010 where they disclosed that they were using social media signals in to rank search results. We also began seeing Facebook share information in our search results, so we knew that Google had access to at least some Facebook data.
Even having this public comment from Google and seeing the Facebook data in search results (you can also observe them in Google realtime, e.g. here), we were still surprised at the size of the correlation in our ranking factor result and we wondered whether it was causal or the result of other factors like links. As a simple check, we ran some hasty partial correlations to control for links and concluded that they accounted for some of the correlation, but not all. This appeared to be another data point to support causation, or so we thought...
SMX Advanced
At SMX Advanced in Seattle last week, Rand Fishkin presented the main results from the ranking factors work, including the partial correlations controlling for links. We were still not confident that Facebook shares were being used directly, so Rand was very careful to add several caveats saying that these might not be causally related. You can see his presentation embedded below:
That evening, Matt Cutts, the head of Google's web spam team said that they do not crawl Facebook "wall" pages and implied that they don't use Facebook shares for ranking. His language was somewhat vague (leaving room open as to whether some forms of Facebook data are used or come via a special feed), however we and many others felt that Matt had implied that Facebook shares, specifically, are not part of their web ranking algorithm. Rand pointed out that Google does have some access to Facebook data overall and set up a small-scale test to determine if Google would index content that was solely shared on Facebook. To date, that page has not been indexed, despite having quite a few shares (64 according to the OpenGraph).
Both Rand and Matt's talks and the subsequent discussion with Danny Sullivan on Twitter was well covered over at Search Engine Land.
Sitting in the audience, I began to think about the implications of this new information. If Google wasn't using Facebook shares, then the high correlation must be explainable with things that they are using. I made a short list of different factors that Google might be using to determine relevance that would also be correlated with FB shares:
- Links. Pages that are heavily shared on Facebook tend to also be heavily linked to.
- Other social media signals. Pages that are shared on Facebook also tend to be tweeted about and shared through Google Buzz.
- Quality content. People share pages that they find interesting and have high quality content. This drives positive usage signals (time spent on page, bounce rate, etc) that might be used.
- Association with well known, quality brands. There is likely more interaction with well known brands then lesser known brands, and this might drive deeper engagement with them on Facebook.
Building a better model for Facebook shares
I thought back to the partial correlations I had run a few months prior. At the time, I was mainly interested in a first look that could be done in a few hours, so I chose partial correlations using a limited set of four metrics from Linkscape as the control variables. Partial correlations use a linear regression model to predict the correlation variables (in this case Facebook shares and search position), the simplest type of regression model. It has the benefit be being well established and easy to use, but falls short when the underlying relationship is more complicated or non-linear. In addition, I didn't try to control for other social media signals since at the time we were interested in links.
I began to wonder if the results would change if I tried a more complicated model using Twitter/Google Buzz and all the available link metrics from Linkscape so I set out to build the model. Before describing the model, it's important to write down our modeling assumptions. They are:
- Google uses link metrics for ranking, similar to those available in our Linkscape API.
- Google uses other social media data, in particular Tweets and Google Buzz shares to rank.
We are testing whether Facebook shares provide any increase in predictive power beyond these factors.
To build the model, I took a subset of the full dataset used in the ranking factors report (scrubbed for data quality, but otherwise unchanged). The baseline mean Spearman correlation between search position and Facebook shares in this data is 0.30. Then I took all 61 keyword agnostic link metrics used in the ranking factors and (1) ran them through a generic filter to transform them to something close to Gaussian and (2) did a principal component analysis and kept the first 19 principal components that explain 99% of the variability in the original data. This allows me to use a complicated non-linear model without worrying about collinearity issues. I augmented the 19 principal components with three social media metrics, namely the number of Tweets and Topsy Influential metrics from Topsy and the number of Google Buzz shares.
I used a two step process to fit the actual model. First, about 33% of the URLs don't have any Facebook shares and the rest have at least one share. Fitting a regression model to a distribution with a big spike at 0 is generally not a good idea, so I first fit a binary classifier to the cases for 0 / 1 or more shares. Then, I fit a non-linear regression model to the remaining data with at least one share. Over fitting was controlled through cross-validation. The total predicted number of shares can be computed from the output of these two pieces.
The final model performs moderately well, with the correlation coefficient between actual and predicted shares at 0.73. However, the mean Spearman correlation between the predicted number of shares and search position is 0.27, nearly as large as the baseline 0.30. This strongly suggests that our interpretation of Matt Cutts' statement is correct and Google is not using Facebook share data directly to rank.
Takeaways
- Facebook shares, at least as related to Google searches, act as a sort of "super-metric", encompassing a variety of different factors (similar to SEOmoz's Page Authority and Domain Authority).
- Don't stop sharing and generating brand engagement through Facebook! Driving deeper engagement through social media can only help your brand and likely has other positive benefits (by generating tweets or links, for example).
- Earning Facebook shares (probably) will not directly boost your Google rankings (though it may have positive effects that indirectly promote links, tweets and other signals Google may use).
- The process of doing this type of correlation work and sharing the results openly started a great dialogue in the search community, and through that we all learned more about how search works. We plan to do more work like this, and are planning a project for the Fall to compare Google and Bing results.
Finally, I'd like to thank Danny Sullivan and Matt Cutts for sparking this work through their discussion at SMX Advanced.
Hey Matt - really appreciate the extra work on this, and the contribution of the blog post. It's great to have someone much more technical and mathematically savvy writing these :-)
I have to admit, I'm really surprised. I know we always say "don't interpret strong correlations as being sure signs of causation," but this one was so suspiciously high, it made me wonder. Knowing that there's another explanation (that can be modeled and shown with data) is awesome, and I think it will serve as a great example going forward when people ask about both the difference between the correlation numbers and actual cause as well as the value of producing these reports (without your initial work, we couldn't have learned so much more about how Google works).
One thing I do wish is that they'd been more clear in their interview with Danny Sullivan on SearchEngineLand late last year. Parsing the statements, I can see the "wiggle room," but the implication, even reading it over again, is that they do directly consume and use that data for rankings.
I agree completely - I learnt more reading this post than anything I can remember for quite a while. One thing this all tells us is that it seems that FB shares would be a great ranking factor. If Google really can't use them (perhaps because of contract issues) it could be a competitive advantage to Bing.
That's exactly what I thought to myself: Google not using, or not being able to use Shares as a ranking signal is a big "social" disadvantage. For me this could be one of the neckbreakers for google. Not getting onto the social train soon enough and scaring off people with those scammy search results, where in the social world you have to keep your west clean otherwise you'll just be blocked away.
The alliance with Bing and Facebook is beginning to influence people's beharior already. Interestingly, Microsoft could push people towrads Bing and facebook. In return, Facebook could also start pushing people towards Bing. But I like Google! Alas, it's not up to me. Thanks for your hard work Matt!
I remember people saying that the $240 millions Microsoft invested in facebook was too much money for a little share. Now we can say that was a brilliant business move by Microsoft.
Indeed.
Impleting Bing on Facebook was very idiotic! Bing is about 7 years behind Google. I tried to use Bing but JEEZ - if I have to search for something for more than 10 minutes and can't find a relevant result on the first 3 pages - Thanks for your service, but it sucks :)
The problem is Microsoft doesn't listen to its customers or the users, but Google does and shows it every 6 months :)
And Facebook using Bing for external searches - Man, this is so comunistic. The best way to provide an external search in such a large site will be to give the users a choice - to choose the search engine for themselves. It's just a single option... with 3-5 checkboxes... or radio buttons.
Am I the only one suspicious that Google may be manually affecting your test url in this case study? It should certainly have been indexed when Danny Sullivan linked to it in his SEL post, right? That was a very popular post that Google indexed for your unique search term, yet suspiciously your URL wasn't indexed still despite that quality link? Seems suspicious to me that they may be trying to throw off your test, eh?!
I think you are looking at things the wrong way round. You seem to be assuming that people share things first on Facebook and Twitter and then it appears in the SERPs.
It's actually the other way round - people have to find the page first, and then they share. Typically it goes like this: someone gets into a discussion on a wall - then they Google to find the answer, or to back up their argument, and if the result they find is helpful, then they share it. BTW the same process happens in forum discussions.
There is a correlation - but it's more accurately described as "Your Google SERPs ranking influences your Social Graph", instead of "Your Social Graph influences your Google Ranking". Because people have to FIND things first before they share them, and what is the primary method of finding stuff online? Google.
OR- The entity who creates the content shares it via social media (ex: @mashable or @ChrisBrogan) it is then shared and boosted up the rankings. For instance, Twitter is essentially a conent/ link sharing machine many people could share their content before Google gets a chance to crawl it ... it's naive to think it happens exclusively one way or the other.
At the end of the day, it's what works to get results. Whether Google uses Facebook directly or indirectly doesn't much matter. What matters is that using Facebook to promote content is good for SEO. Why? Perhaps it's direct. Perhaps it's secondary or tertiary factors, but it does make a difference. Perhaps it is the user data that Google uses. Perhaps it is the network of user's friends that shared the content that then makes it rank higher for that individual who then clicks on the link and that helps it to rank higher for the friends in their network and so on and so forth.
Yes, it's technical. But let's strip that away and in the end we know that it is good to promote content via Facebook and that content that is promoted well on Facebook does, indeed, help the content creator to get a larger audience. If that audience is properly targetted, then that will help the client to achieve their goals.
Social and SEO will become more and more integrated. Almost to the point of lacking distinction. An SEO that feels they should not promote via Facebook, Twitter or any of the other 300+ social media sites is simply sticking their head in the sand and doing their client an injustice.
But what do I know . . . I'm just a BaldSEO. ;-)
I wish I was there. I'm 100% sure they use facebook "likes"
Which came first, the chicken or the egg?
Do the high page ranks bring more visitors and so the page get's more Facebook Shares, or do the FB shares give you higher page rank and then spiral upward to bring more traffic?
:-) I'm sure the answer is something of both.
Joe Shaw
Hey there,
in my opinion its important to keep some technical details in mind. When talking about "Facebook Shares" people might end up talking about different things. This social signal can be seperated into three "dimensions":
- Time: Recently Shared Items might have high relevance for the search results (for instance: Iranian riots)
- Social: What a friend of mine shared might be interesting for me
- Statistics: The amount of total shares and likes are an indicator for an item's popularity
Ok - nothing new so far - but heres my point:
Google only has access to the "Statistics"-Dimension (via Facebook Graph API / FQL - this data can be aggregated without any authentication!). But to the other dimensions ("Time" and "Social") Google doesn't have any access, as they would require authentication by every single user (= app privacy settings!).
An exception is Bing - Microsoft has an exclusive Deal which allows them to access some Data to improve their SERPs. You can see a list of pages which have an exclusive Data-Sharing with Facebook at:
https://www.facebook.com/settings/?tab=privacy§ion=applications&field=instant_personalization&h=0ed0b3c542103d7b1a77fe0adb07fc03
So maybe Matt Cutts was just referring to these Social+Time Dimensions which are quite attractive for a search engine - but solely the "Statistics" Dimension gives search engines a nice and easily accessible info on websites which can be used to verify a ranking of a page.
Best Regards,Merlin // Super-Trainee @ Tandler.Doerje.Partner
Nope you are NOT exactly right about "Statistics" part. Google cannot see any personal shares (and nothing connected to them) without authentication, it can see ONLY PAGES and what`s on them (shares, likes and comments).
Google cannot see "personal" shares but they CAN get the raw number of likes/shares/comments for any URL they want via the OpenGraph. Want to know how? It's not just Google, ANYONE has access to this information. There are several ways to access it, but here's a quick one for you. Just type in any URL into the form at this link as if you wanted to create a Like button for it, and the preview window will show the total number for that URL:
https://developers.facebook.com/docs/reference/plugins/like/
Edit: Afte re-reading these posts again, maybe we're all saying the same thing. Cheers...
Yeah right, the raw numbers, sharedcound.com is the easiest way yet to see social data for url as far as I know , but since they are really easy to manipulate (fake fans, fake likes fake whatever) don`t think google will ever be interested in raw numbers only, without profile authority analysis and so on...
Greetings.
common sense makes me believe in correlation much more than causation, because most of the things i share and i see shared by friends come from authoritative websites...
and i think this happens simply because, trusting more an authoritative website, we are less worried about how sharing something would affect our own authority... for example, if there's a news i want to share about SEO, i'm more willing to share something from seomoz than something from an unknown website...
however, if someone tells me there's surely causation between ranking and shares, i'd think it's the ranking causing the shares, not the opposite. and this would be explained simply with the fact an authoritative website has more traffic (also from search engines), meaning it has more exposure, making it more likely to be shared. any thought on this?
It is somewhat what I think too. The fact that there is corellation does not mean causation. the type of ages being shared can truly be at the core of their ranking. however, if a test was run with mediocre pages that had been shared a lot on facebook then the results could really help in seeing if facebook shares are always what drives a site up in the rankings.
FB "Share" can't be that important if SeoMoz is only using "Like" on their blog, right?
I feel like I should have to pay extra for this... This is brilliant. Welcome Matt, keep these posts coming!!!
Of course this data and Google's commentary doesn't necesarily suggest that the Goog will never use Facebook shares as a ranking signal. In fact, I think it would be naive to assume it would never be. Facebook is highly personal, and most of us wouldn't share things on Facebook to our closest friends and family unless we really meant it. It's a pretty legit signal with tons of potential. (Obviously, or these kind folks wouldn't be testing it.) Though the Goog did create +1, perhaps to get their own share metrics so they don't have to rely on the more inaccessible and proprietary Facebook data... so who knows? Anyhow, thought this article was great and I really appreciated the detailed explanation of how the data was acquired.
I'd like to make a great comment, but I am just doing this for the points.*
For those of us who are more intrinsic than scientific, all of this continues to support the generalist approach to web popularity. One, people come to interesting and unique information ( a share indicates that). Two, people respond to incentives (all of economics summed up in 4 words). 3.) Google is probably applying many elements of the Bayes theorem** to it's algorithm and therefore correlation is not to be under estimated and causation is never to be over valued. Thank you for a great post. This information helps us steer the ship in stormy SEO seas! * partial fib ** Likes to mention my own last name in context of a great thinker
The individual impact and collective weighting of individual social signals sources upon web rankings is sure going to keep us on our toes for a while. So many sites are gaining high volumes of 'Likes' and retweets through offering incentives (I've just witnessed a competitor grow FB likes by 1780% by offering an iPad2) that their value as votes for the quality of a website are already diminishing. Google's task of utilising social signals as a predictor of a website's authority and merit is not only hampered by Facebook's walled-in world, it's also hampered by the ease of manipulating their worth. As ever, the creation of genuinely link-worthy content will prove a healthy long-term policy.
Personally, I think that link signals are as easy if not easier to manipulate than social signals.
Sean, I agree completely, and Google have built a very high profile web spam team working to identify and devalue manipulated links.
Incentivising likes and retweets for the purposes of increasing the apparent social buzz around a website is also very simple and very time effective in it's own right. But... as the search engines increasingly incorporate social signals into their ranking algorythms they will surely put a strong focus upon identifying and devaluing social manipulation.
And the techniques I mention above will be easy to spot and devalue, should Google wish to do so. Just looking for the inclusion of phrases like "retweet to win" will be a piece of cake.
Mmmm, cake.
Sorry, where was I? Oh yes...
Hence my assertion that building truly link-worthy (and by association social-share-worthy) content is a safer long-term strategy than 'buying' retweets and likes. Just a bit of food for thought as we all run hell for leather into the glittering glades of social SEO :)
Noooo !
I was so hoping the the FB shares would be directly effective, since I started my tactic to tap into these shares on one particular website last month I've now got roughly 680 shares.
I'm going to change tactics and see if there is any obvious benefit and attempt to get the same number of +1's instead and see how that goes.
Thanks guys for your hard work, your webinars, and posts like this are what I feel I get out of my Seomoz subscription the most.
+1 affects rankings directly, and will be very interesting to see some specific results.
Indeed it seems it affects rankings... but it is still a "small signal"...
I think you're missing the point.
This post does not say to stop Sharing things... simply to not think that Shares by themselves are going to make you rank better.
Are all the conseguences of sharing on FB that can keep your site ranking better.
So, abandoning what you are doing to jump on +1 train hoping it is (again) a factor causing better rankings... wrong.
@gfiorelli1, I absolutely agree with you, for the majority of cases, but in my particular case, the audience who are sharing my home page via facebook are not the audience who are likely to convert to customers. It's really been an experiment to see if the shares would have an effect. So I have room to modify the experiment and see if the +1 approach has any effect.
"Pages that are heavily shared on Facebook tend to also be heavily linked to."
I think that is a key insight. Great content is going to be shared across a variety of channels. People share in dozens of places, not just Facebook. Even if Google can't directly pull information from Facebook, they have the rest of the Internet to use as a data source.
First of all: Welcome on board! :)
Second: the greatness of "scientists" is directly correlated to their ability in making clear for everybodies things that are not so easily understandable at first. You have an high correlation metric (let say, an 0,41) :D
Even though I was and I am a strong believer of the importance of the social signals as ranking factor, everything related to Facebook was and is a big interrogation point to me due to the walled garden Facebook is behind. That is why I have my personal radars paying attention to any possible adquisition move Google can commit in order to break the wall and win the siege, as the PostRank adquisition can mean (something believed by Danny Sullivan, in a laconic but clear answer to my question about this, during that same conversation he had with Rand)
My take, being far from considering myself an SEO Scientist, is surely empiric but "good-sense-based", and substantially reflects what is the main message of this post IMHO: people shares what really likes (sic)... and someone like 1 thing for different reasons:
If you think it, they are the same reason someone could link to something, or tweet or stumble or +1 something.
So, does our (optimized) contents provoke those three reactions to the user? Yes? Great, that means they will have more changes of ranking better because of the buzz they generate, that delivers in links, citations and any other external signal taken into account by Search Engines.
They don't provoke anything but indifference? Bad, you're wasting time and probably you will have to suffer one between this two consequences:
wow, I'm still a bit confused about everything that has been said in this post but I understood the concept and the message.
Shares contribute to our rankings, but not in a direct manner, it's like a bait, you throw it into a wide mass of people that are on facebook and someone is bound to link to your post or share it even further sometimes making it go viral. of course that all depends on the quality of the content that is being shared.
Exactly, if FB Shares drive links, then the links are really the ranking factor, in theory, but the shares themselves were still essential. Take away the shares, and you'll have no links and no ranking. Unfortunately, there's no direct way to measure that.
And that's it in a nutshell really.
I had a feeling something like this was going to come about sometime.
A organization I work with recently released a mini-site that went viral. In less than 72 hours, we had 20k+ FB shares on the home page. But despite all of the shares, they didn't even rank for their brand term. It wasn't until a day later when a few links with brand anchor text came in did we finally show up #1.
Great follow-up, Matt. While I don't think it implies we should stop going after FB Shares, pulling apart the data really helps illustrate how complex the SEO game can be and how on single factor is ever again going to account for rankings. We have to be able to see the broad picture of how factors relate to each other.
Aside from the obvious takeaways of this post I would like to point out that I love posts like this. I never was great at statistics, so I am extremely pleased to have people like Dr. Matt in this world. It's awesome being able to have access to quantitative data like this without actually having to do the analyses myself.
Nothing against quant guys, but I'd probably rip my hair out if I had to do that all day long. I know you guys love it though, and my hat goes off to you for that.
Thanks Dr. Matt. Awesome insights in this one!
@Matt: great post!I was going through the presentation from SMX advanced and noticed something interesting on slide #23:"number of external links on the page is valuable in SEO ranking".I have a functional question on that, since I was unable to find a good comment on that anywhere on the web.Do I read that statement as: if you place links to valuable external pages (e.g. pass on linkjuice to that non-owned website), that actually helps the ranking of my OWN website?I ask, because right now on my site https://www.wunderwedding.com on a profilepage of a business, I dont directly link to that company's website, but do it via a redirect. Would it be better to directly link to that site? An example can be found here: https://www.wunderwedding.com/weddingvenues/venues/763/the-westin-richmond under the 'visit website' link at the right side of the page.Hope to hear your advice!Thanks!
Facebook is too big to ignore for any search engine. Kudos to Bing for getting in their, but their is obviously some connection between Facebook and Google.
Just a quick quick question here: Google has no access to facebook profile ( facebook pages exculded), then how Google gathers data about shares and likes???
Thanks for this articles, I always gived lot of effort for conntents to be embed in Facebook... your article prove me that I didnt wasted my time ;)
The question isn't so much: does Google use facebook shares in search results calculation? But: What weight is attributed to it?
In my mind any data Google grabs will be taken in account.
Denis - Google Expert
Denis Decroq is currently working on the following sites: https://www.consultant-web.org/ and https://www.formation-pizzaiolo.org/
So what would happen if you directly pinged the page of your facebook share and told google something was their? I know it says they won't index it, but has anyone tried to ping it? From what I read it says that it doesn't directly help, but in the end someone will link to it and then it will? This makes no sense really. It sounds like google just needs a way for the spider to get to the page.
I am going to conduct a test using googles ping servers and see if they index it without ever linking to it in any way. If they index it then that means google just needs a way their.
Hello,
After making a facebook that had a website url in it. I pinged it with many google ping servers and I was indexed in 7 days.
I also made another facebook with a website url in it. I didn't ping this facebook, but linked to it with a hub page that had the ability for DoFollow links and was indexed in 3 days.
I haven't tested this, but I am willing to bet that if you make a facebook. Comment 1 time in huffingtonpost in a comment section under an article, that it would be indexed in 1 day. I have never had anything indexed faster than a huffingtonpost comment. They are obviously relevant comments and not spam, I am in no way telling you to spam huffington full of links.
just want to say thanks matt for serving up this slice from the big rankings report, the FB shares factor was a big question on everyones' mind. What you've expanded on here is great and the bullet-points and take-aways really help digest it all. And sparking up a topic like this on here always attracts interesting POVs and comments. I just love that I can always google a seo question and I'll always get a solid answer, as long as I include the magic keyword of "seomoz" in my search!
This was engrossing. Thanks for clearing it up, definitely, once for all. It's still puzzling, though, to get at the "why" behind the correlation, other than (as others suggested) an FB share results in other concurrent activity where Google does crawl and index resulting links and activity.
The "mystery" I think is untangled when you realize that FB is doing it's thing beautifully: recording human activity, which activity results in a flurry of other "viral" activities that are detected by G-bot. So the correlation would be there (unless the FB activity was outright spam) across the board, we humans being so human and all...
What I can't help wondering, though, is why I'm the first to actually "Like" this particular article? Ironic, or did you just disembowel FB activity with your analysis? That's OK. I got your back, Doc. ;)
hhhmmmm
this is a great thing... but the problem is the Old FB Share mechanism is not supported anymore (at least they are no more available in fb api page , it is replaced by https://developers.facebook.com/docs/reference/plugins/like/ )
and in here, we saw that google uses Share more then Like or any other FB Signal... hmmm interesting but confusing indeed :|
I'm perplexed by Google's apparant adoptation of Facebook shares as a ranking symbol. It seems to give creedence to Facebook's dominance. I wonder where Google +1 fits in?
Great Post. I was hoping that someone would do more rigorous statistical work on this very subject, I'm just glad I didn't have to. The post also explains the lack of movement on some our domains that have agregated 400+ shares.
As with all the moz articles well written and analysed, also the use of constant images sets a standard for SEO or any article writing. However I felt you got a firm no from Matt Cutts so there are other factors at play, not FB shares. In other words it has no relevance to Google organic search, though that suggestion still seems to be strong in the article itself, otherwise known as "barking up the wrong tree".
It's an easy enough mistake, a typical cause/effect/correlation near-miss. My wife makes an excellent quiche. I'm overweight. It's her fault.
Well, she says it's not...
I agree - a big player like Facebook can't just be ignored... and you raise some interesting points about the "other factors" for results.
I'm interested to see if the recent 2-month trend of declining US users of Facebook will continue; and further, how we'll react as developers especially since we tend to push social media for additional weight in the search engines.
Thanks for explaining and sharing the technical side of it. I know then that in the long run, facebook would affect rankings.
Great post Matt! It's awesome to see what bright minds can do with data sets like these. Keep up the good work!
It's interesting but maybe you are looking for the wrong thing here. What Google has admitted to is assessing the authority of the person sharing the information. Having a lot of shares increases the probability of people with influence sharing the page.
It might be worth using a buzz monitoring tool like sysomos to assess the people who have passed the content on and then run a correllation on some clever maths around the combined influence of the crowd.
It might find the causation you're looking for
+1 to this post and more like it.
This deep analysis is why SEOMoz is successful. Really valuable work that helps clarify what the search engines are really doinng, rather than relying on pseudo-metrics that aren't correctly tested.
Great post!
This whole Facebook - Google thing is EPIC, and it`s getting more and more intense and interesting.
Thank you for this post, Matt! Like Will, I learned a lot from it!
Although everyone at SEOmoz always tell people that the data is about correlation and not necessarily causation, people seem to ignore this point. I guess it's because causation data is somewhat easier to understand, and a lot more actionable. SEOs might find it hard to explain bosses and clients that the things we are doing might not work, as the data we are acting on is correlational.
This means at least two things: we'll always have these debates where some people misunderstand the data and get somewhat unsatisfied. It also highlights the need to conduct experiments, as this is the only way to get more certain knowledge of causation.
Great but quite technical article. I believe at some point of time Facebook will become more open to Google and then Facebook stats will influence Search rankings on Google.
I followed with huge interest the discussions caused by Rands and Matt Cutts statements.This post is really a highlight and seems to me a solid basement for our explanations regarding our customers.
I am really looking forward to your next announced blog post regarding the comparison of Google and Bing results. Seems to be very thrilling, too!
This is a great write up, Matt. Thanks for taking the time to add the explanation, the big picture makes a lot of sense, and as Rand said above - it is always good to remember that correlation does not equal to causation.
Looking forward to more data posts in the future!
Just take a look at how much content is indexed withing Google from the site:facebook.com domain, it is evident they know alot more then they claim to.
This research only emphasises the fact that FB shares are even more and more crucial in SEO, along with a host of other social signals.
Interesting times ahead.
I just love the way Facebook republishes Wikipedia in it's entirety, and slaps a nofollow & target="_blank" on the attribution link. Google ate up all that content.
Interesting...I prefer using Twitter so if Facebook Shares do not at this time directly affect rankings, all the better (I know, I know, I have to kick Facebook into high gear)...Google Buzz affecting rankings is a drag though, it doesn't seem remotely as popular as Twitter or Facebook.
Correlation is not causation" but if correlation[Correlation(r) = NΣab - (Σa)(Σb) / Sqrt([NΣa2 - (Σa)2][NΣb2 - (Σb)2]]is positive it's maximum probability for causation. Facebook share helps for real time search on Google rather any other social networks data.So we can make a guess what Google applies for its ranking algorithm!
Thank's for a nice coverage!
Right now the word Viral make the difference in SEO and for viral the best resource is Facebook and other social media sites BTW matt you pointed out some good points :)
Wow! Great technical analisys Matt. Thanks for your work!
Well Done Matt! I really appreciate this because there was a lot of speculation and it does a fantastic job of detailing the correlation/causation dillema. Frickin great work guys!
Matt: Great post because I LOVE articles based on data not on some expert opinion or rumor. Especially in the Google environment where everything's still binary (you are #1 or #2, not 1,5 (bad analogy I know)) it is important to jump through those math-hoops in order to understand ranking. This article is one of the reasons I pay for seomoz: hard facts.
Just curious. Wasn't the analysis based on shares, as distinct from likes?
Previously shares were better than likes because likes were just a thumbs up, and shares actually included the link to the page. Then FB decided likes were the future, removed share from the developer documentation, and moved the share functionality into likes. So isn't this now out-of-date?
Google may have said they weren't grabbing data from wall posts, but grabbing the number of likes displayed on a website would have to be tempting and valuable.
I was actually going through all the comments to see if anyone did address that...
Facebook Shares are no longer supported!!
so what are we talking about here? Facebook Shares, that no longer exist? Facebook Likes? Or maybe Facebook Send?
WOW! Thank you so much for taking the time to research and clarify the importance of Facebook likes. I am have been wondering how this fits in to our SEO "recipe" and you have answered many of my questions. Keep up the great work!
excellent post, can you share the data you used?
This is a great post.
Will take some time for me to digest!
Bizarre how they seem to correlate well, yet Matt Cutts said "Google can see Fan pages but that’s it."
Its still useful as a social metric even if claimed not to be a ranking factor.
Did Rand or anyone from the 'Moz team ever actually say "FB shares cause higher rankings?" If they did I must have missed that. ;-)
If FB shares drive any links at all, any traffic at all, that will help with ranking, but when you're measuring what SEOMoz measured, it seems like "shares cause ranking" has it backwards.
Higher rankings will undoubtedly cause FB shares. Rank higher -> get more traffic -> get more shares. Greater traffic will also cause.... more links. More revenue. More likes. More comments. More tweets... More of lots of things.
Greater traffic combined with a solid social marketing strategy will cause more traffic, more likes, more shares, more comments, more... oh yeah, it will also cause more search engine traffic from the "socially active" crowd that's also the most likely to give links, likes, shares, tweets, etc.
The conclusion to all of this hubbub is... "marketing is a good idea." Who knew?
Google is using sharing data from Facebook, Twitter, Google profiles and additional more. Google must focus on this section. Because, too many websites still not hire SEO expert to expand it's view in front of giant viewers.
But, tricky and long term keyword search are giving us such a good website. It suppose to share by users on Facebook or Twitter. Google can understand that: This is charity marketing stuff from Facebook or Twitter users. So, why not give some ranking benefits to those websites.
Dispaly of Twitter or Facebook profile over Google search can make some changes over CTR. What you think about it?