There's been a lot of talk in the SEO community about the growing importance of social metrics, and some have been starting to ask whether social signals will one day replace link metrics as the primary method of measuring relevancy. Well, to continue my opinionated streak, I'm going to say no, the social graph will never replace the link graph. Here's why.
The Myths of Social Media
Let's start by confronting some misconceptions about social media.
1. A social presence is harder to fake
This is simply untrue. It's already possible to buy fake followers on Twitter, Facebook, and other social networks: often hundreds at a time for about $5. The profiles are fake, of course, and critics argue that Google's algorithms could easily identify this because the profiles are obviously artificial.
They are certainly correct that such profiles would leave all kinds of artificial signals, but there is no reason to expect that these signals would be any stronger than they are for artificial websites. A series of new websites that link to each other and are otherwise isolated from the rest of the web looks just as artificial as a series of new social profiles that all follow each other and are otherwise isolated from the rest of the network.
2. People prefer getting information from friends than from algorithms
This isn't so much a reason why Google would use social signals as a critique of SEO and search engines in general. But if there's any truth to this statement, it doesn't carry over to social networks. It turns out people use social networks for pretty much anything but finding information, as we recently discussed.
This does have implications for the search algorithms, however. If people don't use social networks to find information, why should Google? Most people use social networks to entertain themselves, and a smaller number use them for self-expression and communication.
Marketers seem to believe that the primary reason people use social networks is to share information they find on the web, when in reality most of the web doesn't exist to entertain, and so it will never become social.
And that brings us to...
3. Content that isn't share-worthy isn't worthy at all
There appears to be a now deep seated notion that people will naturally share any piece of content that they find useful or enjoyable. This isn't true. People don't share content because they like it or they find it useful. They share it because they think it will help them maintain or build relationships with their peer group. This is a very different thing.
Consider the fact that 94 percent of teachers claim students equate "research" with "Google." That's a lot of people turning to the search engines for information, the one thing they almost never use social networks for. Educational material, as a whole, will never be share-worthy. It exists to inform, not to entertain. That certainly doesn't make it worthless.
The same is true for the vast majority of commercial content on the web. It takes a very "special" kind of person to get so excited about a deal on Amazon that they just can't wait to share it with all their friends on Facebook. Most people dread seeing commercial content in their social feed.
While it certainly benefits brands to be as social as possible, the reality is that commercial content as a whole will never become share-worthy, despite the fact that Google would look very stupid if no commercial content ever showed up in the search results. (Yes, the search engine does have some incentive to move all commercial results onto its own properties, but that's a different story entirely.)
4. The social world is bigger than the linking world
Yes and no. There are certainly more people who use social networks than who run blogs or websites. But when it comes down to it, social networks are peer groups. People don't share something they find interesting or even enjoyable unless they think it will impress their peer group.
A blogger will link to an obscure and technical research paper because it shows that they have done their research and because it is good practice to cite your claims. Your average citizen will not go to the trouble. Links will and always will span a wider presence on the net than shares.
When Social Does Matter to Google
It's not as though any of the above means the social graph is useless to Google. It means that social signals are only important for subjects that are supposed to be social. For the most part, this means anything that is supposed to entertain or to pass the time should generate social activity on the web. That's because these are the reasons most people use social networks in the first place.
What does this mean for SEOs? Primarily, it means that you should only expect your rankings to be affected by social metrics if somebody else in your niche is really excelling with them. That sends the message that the subject should be social, and that you aren't making the cut.
But there's the danger of running too far off with this. How would consumers react if even educational search results were flooded with fluffy posts that favored entertainment over facts? (Ancient Aliens, for example.) How frustrated would they be if a search for "buy gold jewelry" led them to a list of hilarious videos about people buying gold jewelry?
Links carry with them a more scholarly or commercial connotation than social shares do. I believe Google understands its users are there for more than entertainment. Search engines serve a different purpose than social networks, which is why they still exist. If search engines started to resemble social networks too much, they would become redundant, lose their unique selling proposition, and disappear. That is why links will never be replaced by social metrics.
Where is Google Really Headed?
If anything, Google's future is looking less human curated. The search engine's original breakthrough, the use of link metrics, was successful because it was curated by humans, but that turned out to be the major disadvantage of links as well. Because links are curated by humans, they are manipulated by humans. Replacing links with "likes" or "followers" would do nothing to solve this problem.
Instead, we have seen Google tackle this problem by placing less emphasis on human curation and more emphasis on patterns in what humans tend to like. These patterns aren't centered around deliberate and obvious actions like links and shares. They are centered around subconscious signals like time on site, repeat visits, co-citation, query sequences, and other statistical data.
If your pages are doing well in social media without any manipulation on your part, it's probably a good sign for the future of your brand and search rankings. But should you invest everything in social? No.
Image credit: Simon Cockell