Why SEO Social Signals Will Never Be a Thing

on under Search Engine Optimization.

Despite rampant claims that social signals influence search results, the fact remains that Google can't crawl Facebook share data, and its ability to crawl Twitter is unreliable at best. In August 2013, dramatic correlations between rankings and +1s had a huge portion of the SEO community convinced that +1s influenced rankings, but Matt Cutts quickly squashed this myth:

Just trying to decide the politest way to debunk the idea that more Google +1s lead to higher Google web rankings. Let's start with correlation != causation: https://xkcd.com/552/

But it would probably be better to point to this 2011 post (also from SEOMoz/Moz) from two years ago in which a similar claim was made about Facebook shares: https://moz.com/blog/does-google-use-facebook-shares-to-influ... . From that blog post from two years ago: "One of the most interesting findings from our 2011 Ranking Factors analysis was the high correlation between Facebook shares and Google US search position."

This all came to a head at the SMX Advanced search conference in 2011 where Rand Fishkin presented his claims. I did a polite debunk of the idea that Google used Facebook shares in our web ranking at the conference, leading to this section in the 2011 blog post: "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)."

If you make compelling content, people will link to it, like it, share it on Facebook, +1 it, etc. But that doesn't mean that Google is using those signals in our ranking.

Rather than chasing +1s of content, your time is much better spent making great content.

Since, then, SEOs have (mostly) switched from claiming that social signals influence rankings, to saying that they will start doing it in the not too distant future.

I may be in the minority on this, but I don't expect Google to start using social signals as a direct ranking factor any time soon. Let's talk about why, and what they're doing instead.

On Unsteady Ground

When Twitter shut off its data fire hose and killed Google's realtime search, Google realized its position in the marketplace. If they weren't already aware of it, they now knew for sure that social signals would never be a reliable metric. At SES, Matt Cutts said:

People were upset when Realtime results went away! But that platform is a private service. If Twitter wants to suspend someone's service they can. Google was able to crawl Twitter until its deal ended, and Google was blocked from crawling those pages for 1.5 months. As such, Google is cautious about using that as a signal - Twitter can shut it off at any time.

Google could never rely on the social signals of other social networks, because they could always block them, and of course take them to court if they tried to find ways around those blocks.

Of course, Google already had its own platform, and it's been heavily pushing Google+ since then. Google+ currently has North of 300 million users (even if some of them are probably just clicking to get rid of the red bell).

Naturally, since Google has access to the information on their own network, SEOs assumed that Google would start using this data as a ranking signal, since they no longer needed to get permission for it.

But then Matt Cutts came out and said they don't use it.

Why?

Google's Own Existence Has Always Been Its Greatest Enemy

When Google first took the world by storm, the search results were phenomenal because they used a brand new ranking signal: links. But it didn't take long before spammers started realizing they could manipulate their position in the search results with dishonest links. The link graph quickly became tainted, and it's been a cat and mouse game ever since.

The irony is that if the search engines never used links as a ranking factor, they would be the perfect ranking factor. Google has become a snake that eats its own tale.

I suspect that Google will never use +1s as a ranking factor for precisely this reason.

Instead, I suspect they have done something rather clever. The hint comes from the massive correlation between rankings and +1s, even though experiments have shown that +1s don't influence rankings.

Rather than using +1s as a direct ranking factor, I suspect that Google is using its machine learning tools to predict what kinds of pages are most likely to earn +1s. In other words, Google is replacing actual +1s with theoretical +1s.

It is certainly possible that the high correlation with +1s is just a coincidence of the fact that it happens to correlate well with other factors. But the fact that it correlates better than any other signal (besides other machine learning "signals" like Page Authority) makes that a tough sell for me, especially when it would be foolish for Google to completely ignore its own data.

If this is true, and I strongly suspect it is, Google would be extremely foolish to turn +1s into a direct ranking factor. This would unleash yet another cat and mouse game between Google and the spammers. The search engine would forever damage it's +1 data. Looking at what's happened with links, we already know there would be no going back.

Google has done something very intelligent. It's created a metric that it has unfettered access to, and it's using it to build invisible ranking signals that we may never be able to manipulate.

Don't hold your breath for social signals.

Image credit: Jason A. Howie

  •  
  •  
  •  
  •  
  •  
  • Hey Carter - nice post!

    Good ideas in here, but there's too much speculation on this matter at the moment. In fact, it's all speculation. As it stands, I believe social MUST play a role of some sort in rankings, even if it's just brand presence.

    The fact is this - social media (regardless of likes, followers or any of those garbled metrics) gets traffic to sites. It gets eyes on brands. It flat out works when people commit to it. It doesn't matter if there are any perceived SEO signals - it's worth being there anyway, isn't it?

    Some are still reluctant to get on board. Some platforms just won't apply to all businesses. But there's no denying it's a part of what we all need to do.

    • Carter Bowles

      I'll agree that social media is an important part of online marketing, and that all SEOs should have a working knowledge of it, at bare minimum.

      As far as social signals directly contributing to rankings, I don't think there's any speculation there at all. Matt Cutts has been pretty clear about that, and while I don't swallow up everything he says, there's no clear motivation to lie. The experiments conducted also seem to have pretty much ruled it out.

      I'll agree that if you're using social media properly, it should be creating all kinds of positive user behavior metrics, and those can almost certainly help improve your rankings if they are positive enough, even if no links result. In any case, nobody should turn down referral traffic opportunities. I still think more emphasis should be placed on email, personally.

      The point about +1s being used to uncover hidden ranking signals is definitely speculative. I'll be the first to admit that. But you have to admit +1's correlation with rankings is suspicious.

      Thanks for the feedback. Always happy to hear insights from others.