The SEO industry moves quickly. Those of us who work in the industry are always trying to adapt and keep up with best practices. Due to the nature of our niche, we can sometimes get ahead of ourselves in the search for the next big industry shift.
Right now there is a lot of talk about Google using social signals as a ranking factor. However, the truth is social signals are easier to manipulate than links. While social metrics are important and they provide value outside of SEO, there are a number of pitfalls regarding their use as a ranking signal for search.
Links remain at the core of Google's search algorithm and continue to be the driving force behind rankings. Due to the pitfalls associated with using social as a ranking signal, links remain the best factor for determining visibility online.
Pitfalls of Using Social as a Ranking Signal
First off, let's make this clear – Google does not use social metrics for search ranking. Google's Distinguished Engineer Matt Cutts explains this in a recent Google Webmaster Help video.
Furthermore, Amit Singhal, Chief of Google Search, reinforced this in his SMX 2014 keynote.
Again, social metrics (such as +1s, retweets, likes, etc.) certainly are useful and valuable in their own regard, but there are some potential pitfalls to using them as a ranking factor. These pitfalls include:
- Social metrics are easy to manipulate.
- Google doesn't have/could lose access to crawl the data of external social platforms.
- Lacking universal adoption on their own platform (Google+).
1. Social Metrics are Easy to Manipulate
One of the biggest reasons that social metrics can't be used as a viable ranking factor is that they can be easily manipulated.
While links continue to be the principal ranking factor Google uses, manipulation has been an on-going problem with them as well. Cutts is constantly battling the spammers and black hats that pollute the web and give white hat link builders a bad name.
Using links as the primary ranking factor does leave room for manipulation, but the problem with social metrics is they are even easier to manipulate than links.
I mean seriously, how hard is it to create a fake social profile?
In 2012, it was estimated that Facebook had roughly 83 million fake/duplicate accounts and researchers suggested that 1 in 10 Twitter accounts are fake in this recent post.
This leaves a lot of room for manipulation.
Along with creating fake profiles, it's also possible to manipulate social data through bots or automated tools. Using automated software, spammers can generate thousands of likes, +1's, favorites, etc. in a matter of minutes.
As Google continuously battles link spam and adds to their link schemes document, it seems unlikely that they would want to implement a ranking factor that is even easier to game.
2. Google Doesn't Have Full Access to External Social Platforms
Another challenge to using social metrics for ranking is that Google doesn't have full access to crawl the data on external social platforms.
"We're not using [social signals from Twitter or Facebook] right now," Singhal said at SMX West. "We don't have access to the Twitter data, so you can imagine how hard it would be to build a system that relies on those signals."
It's apparent that Google isn't confident using data they don't have full access to crawl as a ranking signal. Even if Google were to gain full access, using social data for rankings still remains problematic – as Google could lose access at a moment's notice.
In fact, this actually has happened before.
Back in 2011, Google was blocked from their special feed to Twitter data after a deal they had with Twitter expired. Google spent a lot of engineering time on that data and once they were blocked, all their work became useless.
The risk of being blocked again has made Google hesitant to invest any more time and effort into using social metrics for rankings. Cutts actually referenced this uncertainty in the aforementioned video:
We have to crawl the web in order to find pages on those two web properties and we've had at least one experience where we were blocked from crawling for about a month and a half. And so the idea of doing a lot of special engineering work to try to extract some data from web pages when we might get blocked from being able to crawl those web pages in the future, is something where the engineers would to be a little bit leery about doing that.
These are some pretty clear indications from prominent Google employees that they lack the access necessary to confidently use social metrics from external platforms for rankings. In fact, there are even some issues with their own social platform – Google+.
3. Google+ Lacks Universal Adoption
Along with the concerns regarding external social data, Google also faces some challenges with their own social platform.
The main issue with using metrics from Google+ is that the platform lacks universal adoption. Although Google+ has been widely accepted within the SEO industry, reception among average internet users has been less enthusiastic. (Although Google+ is the fifth largest social platform, much of this has to do with the fact that a Google+ profile is created automatically with every Gmail account)
Even though Google has unfettered access to this data, the issue here becomes that not enough people use Google+ to legitimately use these metrics to determine rankings. Signals from Google+ simply don't provide a fair representation of the web's overall populace at this time.
That doesn't mean this couldn't change in the future though. Google does use metrics from Google+ for personalized results.
"We think about it from a user's perspective," Singhal said at SMX. "They're looking for high quality content generated by reputed people. We have the authorship program that allows that to happen. They're also looking for content from people they know, and we have the personalized results for that."
Why doesn't Google use more signals from Google+?
"We have found that use of social signals in personalized mode is far more positive than using in non-personal results," Singhal said.
The fact that Google is using these signals for ranking personalized results shows they want to factor social signals into ranking. However, they currently don't have the adoption rate needed to do so.
Links Remain the Best Ranking Factor
While social still has value, within search links are the strongest signal for determining visibility. Social offers another worthwhile marketing channel, but from an entirely SEO standpoint it has little to no direct value. Links STILL matter because they remain the best ranking factor in search.
Links are at the very core of Google's search algorithm - it was built on links. To this day, they remain the primary ranking factor because they provide the best search results. Recently, Cutts reaffirmed this in another Google Webmaster Help video.
In this video, Cutts explicitly states that backlinks "are still a really, really big win in terms of quality for search results."
Singhal also reiterated the importance of links at SMX.
"Links are clearly an important signal about the importance of your content," he said. "They're still very valuable. At the end of the day, we take a holistic look at the value of your site. We're looking to build algorithms that give users what they want."
As long as links are providing the best results in terms of quality, they will remain the chief factor in determining rankings in the SERPs.
Social Signals in Conjunction With Links
As I said before, there is value in social metrics and it is apparent that Google would like to factor them into ranking. While there are currently too many pitfalls to using them as a primary ranking factor, it could be possible that Google is working toward using them as a sort of secondary signal in conjunction with links in the future.
For instance, if a piece of content is garnering a high number of inbound links, but has very few social shares, this could raise suspicion.
On the other hand, if something is getting a large number of links and being widely shared via social, it provides more evidence that it is quality content. Google could potentially use social data to further validate quality and justify a higher position in the SERPs, but again this is clearly not feasible from an engineering standpoint at this time.
Although it isn't viable to employ social signals as a primary ranking factor by themselves, they could possibly be used in conjunction with links in the future.
It seems we in the SEO industry are constantly searching for the next big thing to replace links. However, Google built their algorithm around links and they remain the strongest ranking signal. It could be useful to reference social metrics along with links in the future, but currently there are too many pitfalls to viably use them for rankings.