According to Matt Cutts, no….. and yes.
The Tl;Dr is they can’t rely on their robot army to determine social signals, yet. In 10 years, less now, they will have whatever A.I. he is talking about, but for now we have to deal with the pleab humans running things
Matt Cutts says that social signals are not considered within the ranking algorithm, as far as he knows. This was also in November, 2014.
Mr Cutts in that video says that Google doesn’t use social signals as a direct ranking signal, as one of their “200 ranking factors” they like to claim matter. He dances the question though, and I wanted to see what the history was. A couple of hours and a rabbit hole of searching lead me to a 2010 video referenced by a 2012 post by Danny Sullivan. It all basically said that Google uses the social signals, sure, but not directly for causation of rank.
That’s kinda…uh, bullshit, right? I mean of course they use social signals for ranking. There are entire industries built on the back of that fact. There is no denying that a strong social presence will help your ranking, but to what extent do things like Facebook likes or Twitter retweets really matter in the 2016 SEO landscape and to what correlation is that to quality vs quantity.
Since Twitter shut down its API endpoint in November, 2015, have things changed? On one hand you have Google saying they don’t directly affect SERPS, but then you have very smart people telling you they do, like http://www.searchmetrics.com/knowledge-base/ranking-factors/.
A snippet from that great post, which you should read:
The question remains open as to exactly how social media signals such as likes, tweets and +1s boost rankings. Google maintains that it does not directly use these signals as a ranking factor, but our data shows a positive correlation between the amount of social signals and rankings. Social signals remain important for brand awareness and help to drive organic traffic to top ranking sites.
The first search keyword we will check out is “Best Headphones”. I needed a new pair of headphones so why not kill two birds with one stone right?
Check the image and we’ll talk after the jump.
For this keyword, everything matches up, except for that stumbleupon purchase from PCmag, good job boys, that’s not obvious or anything. I was actually interested in that anomaly. If you have read this blog before, you know what means… another post has been created because of the research of this post, lol. The black hat bullshit just doesn’t end, does it?
Anyway, check out PC Mag Stumble Upon profile http://www.stumbleupon.com/stumbler/PCMag?_nospa=true. See anything as interesting as I did? How about their top liked post has 69 likes and their “best headphones” page isn’t even shared by them, but it has 16K+ likes.
Suspect activity exposed!
Ok, so that was an easy keyword and I actually wound up buying a sweet set of headphones, thank you very much. Lets look at something a little more abstract, a long tail that people really like to boost with social, or a deep link to an article shared often. I’ll go with something near and dear to my heart:
“psychology in SEO”
It gets a bit shakey in this keyword because the authority sites within the industry are ranking better, with less social signals. MOZ, which a lot think is the defacto SEO blog, is ranking better than searchenginewatch (SEW), which is also a well respected site. SEW has 326 total shares, almost all from LinkedIn, MOZ’s post has a total of 3 shares across 3 platforms, which is interesting as their posts usually have more social strength.
It makes sense that SEW is shared on LinkedIn though, there are quite a few authors who are quite active there, and they all know how to get a discussion going. This article was written by Jim Yu. I did some digging and found that Jim has an average of 103 shares on LinkedIn per post. Mr. Yu is a medium level influencer in the tech world, he is a contributing writer for multiple respected websites and he is active on LinkedIn to share his story, yet MOZ outranks that post with .009% social shares because, well MOZ is MOZ right?
The take away from this keyword is that sharing across platforms is key. Again, the most shared article is the best ranked in this keyword. Is it because its the best content the would naturally be shared anyway? Or is it the best content because its the most shared?
Welcome to the wonderful world of psychology in SEO!
Ok, fine so we have a clear correlation from a simple purchase keyword in “best headphones” as well as “psychology in SEO”. More shares = better ranking, but only if your website is trusted and has good quality content.
This goes directly in line with what Mr. Cutts said in the video about, that the correlation from social sharing and content comes from the content quality and natural ability for that content to be
Fine, fair enough, but what about when you get away from these guys? What about when you get into the underbelly of the SEO world? Well, let’s take a look at the people trying to tell you that social signals matter for SEO. So these guys are selling the social signals, so of course they are going to claim they work as a ranking system alone, but lets take a look at the data and decide for ourselves.
“Buying Social Signals”
These results are getting interesting, now. This SERP result is almost filled with black hat web design, I could post about all of their garbage, but let’s focus on their overall goal, to rank for keywords,install faith that its a good idea, then sell you social signals.
So, caveat 1: ranking for the keywords they want.
Obviously the most “shared” site on the SERPs ranks decently enough, but it’s being outranked by .004%, 0% and 7% compare social signals, respectively. This is interesting to me, because you have people touting how important their services are and how much you need social signals to rank, but even in their own keywords, it obviously is not an “important” factor in ranking.
Digging a little deeper I added each of the top 7 results for the keyword “Buying social signals” in to MOZ Open Site Explorer and the results were even more interesting. This is a huge image, so open it in a new tab if you want to take a closer look:
So to break that down, the average Domain Authority given by MOZ is 32.25 and that is boosted up quite a lot from Neil Patel’s QuickSprout landing a solid 79/100. Without that score the overall average drops to 21.28. Domain Authority can be gamed, sure, but it’s not here and these metrics are overall in my opinion accurate for this keyword.
So what does that mean? Simply that this is a low performing search keyword that is affected by a heavy hitter talking about it. Every single result on this page has increased because a large domain has entered the fray.
If that doesn’t tell you that their service is bullshit 101, I don’t know what else to say.
Let’s take a look at another social metric: “Buy FaceBook Likes” to round this out.
So here we have clear indication that the most shared articles are not the best value or highest quality. The most shared here are obviously bought. #1 result is an article explaining why you shouldn’t buy Facebook likes, and that is naturally shared across all platforms, it’s quality content that answers a question. With only 2,000 social shares, it is under-perfoming based on these SERPS, but it ranks #1 consistently for long tail variations of this keyword.
Because it answers the question the user is asking without selling something and it has a solid trust flow to it. Look at the people buying social signals here, I mean its obvious when you see it together that you can’t just buy a bunch of social signals to shit and expect it to work and rank, but for every single keyword, having social signals on quality content is a factor, all things being equal.
The final thought
Well, thanks for sticking with me, this got really long, but I think it’s needed to talk about. So what did we learn here? We learned that social signals ARE an important ranking factor in Google, no matter what they say. When all things are equal, if you have a DA of 100 and a trust flow of 100 and your links are great, Google is going to use social signals to determine the winner. The main point to remember though is they can identify the quality of account the share is coming from. You can see this from the over-performing site with low shares. That MOZ article? Shared by @Randfish, one of the most influential people on social media, his share = thousands and thousands of random bot accounts.
So should you buy social signals for your posts? Probably not, to be honest, because the correlation of your content to the people sharing it is real, if you don’t care about a bunch of random bot accounts sharing your stuff along with whatever else garbage they are being paid to share, it may help you in the short run to get to the first page of your results, but its not sticky. As soon as someone creates quality content that gets shared by quality people, especially if it answers the same question your shitty shared page does, consider yourself alt+f4ed.
It’s been a journey, hopefully you learned something, I know I sure did.
Tools/Sources/Links from this page.