Thoughts On How Social Networks Impact Search Results

Posted: 22 Sep 2011 05:30 AM PDT

A few weeks ago Nikos asked if I would offer my opinion on social networks and their effect on search results, particularly now that Google+ is on the scene. Who am I to say no?

As it happens I’ve been answering this same question often enough on forums lately. Seems like a popular question.

Before getting started, please understand this post is just my opinion. I’ve done exactly 0 testing on the things I’m about to say. My thoughts are based on my own reading and understanding of the subject and in trying to think how a search engineer might think about social and search.

With that said how about a little search engine history? I promise we’ll get to the social stuff, but some thinking about where search has been will give us a better context once we’re there.

Search Engines and Links

Imagine it’s day 2 of life on the internet. Day 1 was filled with people posting articles and other content and along comes day 2 with people wanting to find the stuff they’re interested in amid all that content.

You decide to build a search engine to help people find what they want.

Putting aside the issue of literally finding and indexing what’s out there, your task is to decide which pages to rank first. I would think you’d naturally begin by looking at the words that are actually on the pages you’ll return in the results. It’s the obvious first step.

You look for matches with with words in the query and words on the page. Perhaps you give emphasis to words in specific places like the page title or at the beginning of the content or in various meta tags. This works really well until people discover the value of ranking first and discover how you’re choosing to rank web pages.

Then comes the spam. And then more spam and still more spam. It’s easy to manipulate the words you place on the pages you control.

Eventually you realize you need to find another way to understand the content on websites and rank it appropriately.

That’s where Google comes in with their PageRank algorithm. Google shifted things and made links the focus. They didn’t remove that “what’s on the page” data, but they shifted the focus.

Their idea comes from the idea of attributions in scientific papers. The more mentions of one paper in others is a good indication of the value of that initial paper. If it works for attributions and scientific papers, why not links and web pages?

The one difference is that spam problem. For the most part scientists are a little more trustworthy than spammers.

If more links meant better ranking then people would figure out how to get more links regardless of the merit of what those links pointed to.

Google adjusts to see some links as more valuable than others. The rest of us figure out what makes some links more valuable and adjust too.

Google adjusts again and looks at the authority of the web page linking out, the distribution of domains sending links to a particular page, where on the page links are located, the words used in the link, how often those same words are used in all links, and on and on.

The rest of us learned and adjusted as well. Some to understand seo fundamentals and some to continue playing the cat and mouse spam game.

In the end search engines continue to add more and more data that they can use as signals to determine whether page x is a better match to a query than page y. The more signals they use, the harder it becomes to significantly manipulate the results.

Social and Search

Social sites like Facebook, Twitter, LinkedIn, and now Google+ offer another set of signals search engines can use. At first it was the links coming from those sites to our content, but now most of those links have the proverbial link condom added.

Social shares can be seen in a similar way to a link. If you consider a link as a vote then a Like or a Tweet or a +1 is also a vote. And just like links, different social shares can be given different values.

And just like links, social shares can be manipulated. Whether or not they’re easier or harder to manipulate I don’t know. My guess is it’s about the same.

Both Google and Bing have publicly stated that they look at social signals and some of these signals affect the order of rankings in search results. All indications seem to point toward them looking more toward social when and where they can get data.

In a recent email to Wired Magazine Google said:

Google will study the clicks on +1 buttons as a signal that influences the ranking and appearance of websites in search results

If you go back to scientific papers and links then search engines probably started with the simple more social sharing means more important content means better rank.

Again this is easily manipulated and so we need to consider possible ways to view some sharing as more valuable than others.

Who’s sharing? What’s the authority of the sharer?
How often is something re-shared?
How quickly is something being re-shared?
What’s text is used to share?
In how many different communities is the content being shared?
Can we apply a social rank akin to PageRank

The above are just a few things off the top of my head. We can probably brainstorm and come up with more. Then we can watch as everyone debates the merits and lack of merits in our ideas and the usual flame wars that will ensue.

SEOmoz collected data earlier in the year to present their findings on correlation between certain search signals and their effect on ranking. As always needs to be said, correlation doesn’t not automatically mean causation.

In the end I don’t think the details as important as the general idea that search engines are looking to how we all share things in social communities and that they’re looking to those communities for signals that will affect search results.

For most of us this will mean a few things.

We need to continue to create the best content we can
We should create content for real people as it’s people who share socially
We should encourage our readers to share our content
We should participate in social communities
We should understand we don’t need to be everywhere

Why Google+

Look again at my last point above. We don’t need to be everywhere. It’s physically impossible and even if it were possible I’m sure it would be exhausting.

What advantages does having a Google+ profile offer that other networks might not?

It’s Google’s data. If you’re a search engine and you’re going to collect data that determines how your search engine ranks web pages, who’s data are you going to use? Your own most likely.
Facebook refuses to share it’s data with Google and Google’s license for Twitter data recently ended. Google only has access to the publicly available data of both.

Face it the social data Google will use is the data it collects directly from its own network. So which network do you think will most greatly affect Google’s search results?

Google is already using Google+ in a way that could lead to more or less traffic. Currently if you’re signed into your Google account you’ll see avatars from your Google+ friends for those results they’ve given a +1. You’re more likely to click through with that recommendation.

You can also +1 content directly in Google’s results and I wouldn’t be surprised if what you and your friends +1 re-ranks your personal search results.

Because Google+ is still relatively new, you’ll have a head start on building an authority profile for when the masses arrive (assuming they arrive). With less people active now you can also build stronger relationships with people who’ll also have authority later.

Summary

The first shift in search rankings was the move from on-page to links as the major factor in ranking. Today another shift is taking place as things are moving to the social scene.

How much of a shift it will be is currently unknown. What’s known is that search engines are looking to social networks for new signals to help them rank content. That alone should be enough to convince you of the importance of social networking and sharing where search results are concerned.

It stands to reason that social signals will follow a similar path as links. it starts with quantity being the most important and then moves into quality being the most important.

The major takeaway for us is to continue to create great content, participate with social networks, and encourage others to share your content. Where Google+ is concerned, now is a good time to build authority before the masses arrive.

I still have some Google+ invites to give out if anyone is interested.