Link Graph vs Social Graph

Link Graph vs Social Graph

Time Icon 2.27.2013Tag IconSearch

Google is by no means transparent especially when it comes to topics like the link graph and social graph. This is one of the biggest problems facing professionals in the field of SEO, and it’s been this way for about a decade or more. If anything, the landscape has become increasingly hostile over recent years, especially for those who were using techniques that were outside of Google’s Webmaster Guidelines.

Fortunately, lots of information has been released by Google in the new direction they are headed, and a clearer picture is emerging as to what must be done to successfully rank a page well in Google’s SERPs. That picture looks pretty much the same way it looked over 10 years ago: Create and publish good content, make sure you optimize it properly, then tell your friends about it (back in the day it was forums, newsletters and chat rooms, now it’s social networks like Twitter, Facebook, LinkedIn & Google+, etc…) and within Google’s Webmaster Guidelines and chances are, it will rank. Imagine that!

How well your content ranked circa 2001 depended greatly on various factors primarily dealing with link popularity – that is the number of links pointing back to your website. The way websites were interconnected at a macro level was commonly referred to by seasoned SEOs as the “Link Graph”. Those who are familiar with computer science or have some experience with software / web development know that these are just nodes (different documents and pages) connected by directed edges (links).

Over recent years, Google has changed how it ranks websites. Now it factors social media activity around content and even the authority of the author who wrote the content. This is being referred to more and more as the “Social Graph”. There is no need to start putting too much focus on one over the other (yet), because for the time being both the social graph and link graph are two important parts of how websites are ranked in Google current day.

Boss Hog: The Link Graph

The link graph is still the big dog, at least for the time being. It still has the most influence on where your site ranks, but that is changing more and more every day. With a clear understanding of the link graph, the social graph and author rank system will make much more sense.

Boss Hog

In the past, SEOs have almost completely relied on link building in symphony with on-page SEO to get rankings. While on-page has remained just as significant (if not more), low quality links are quickly fading in their effectiveness to boost rankings, especially any kind of links that can easily be obtained or paid for. Back when search engines relied primarily on links alone to value website pages, they were heavily favoring connectedness of sites in order to establish which sites were most relevant for any given keyword term or industry.

This made it much easier to crawl the Internet because connected sites tell the search engine both who and what is being talked about. Although this connectedness was good for bots, It didn’t help to identify authoritative, trustworthy sources, nor does it prevent spam – so it wasn’t always all that great for people. Further refinement of the algorithm since 2010 has included relevance between linking and linked pages to determine the amount of PR assigned to the interlinked pages as well.

Page Rank in 2013

Each website consists of a domain and pages within that domain. Even if a website only has a single home page, the homepage and domain itself are separate. This is because the content of the homepage can change, but the domain itself must always stay the same. If the domain changed, the website itself would be different. This establishes two layers that can be used to rank pages: a domain-level rank and a page-level rank. What most people referred to as “PR,” or “page rank,” is actually the domain-level rank, and the page-level rank is often left out of the equation in favor of the site’s general rank as a whole.

Page Rank is just one piece of the puzzle, though, because massive search engines also need a way to identify how trustworthy a site is. Trust is based on how connected any given site is to a well-established, highly selective site. For example, Oxford college’s site is highly academic, does not link to spam, is highly referenced among other sites that are reliable and consistently produces content that follows the same trend.

These are all standards that are impossible to re-create without having genuinely useful content that people are interested in, which is exactly what search engines want on the results pages. It’s important to realize, though, that the reliability of inbound links are far less important than outbound links. This is because people running a site like Oxford have no way of controlling what sites are linking into Oxford. What’s more important is that Oxford retains its integrity by carefully reviewing all of its own outbound links.

Trust Factors

As Mark Traphagen pointed out last month, Google has hinted at where this is all going. In order for sites to develop trust and reliability in the eyes of the search engine current day, they can’t associate themselves with spam, scams, unnatural links, link schemes or spyware / malware of any sort. Links current day are said to have less and less influence on SERP results, while others say that certain authoritative, in-content links still count. No matter what people in the SEO space say, only Google knows for sure.

I like to imagine a spectrum with complete trust on one end and complete distrust on the other. The closer the site is to the most reliable, honest and trustworthy sites on the Internet and has content that is commonly shared by trustworthy and authoritative sources, the better it will fair in search engine results. If the site links to questionable content and shows a pattern of spammy, black-hat practices, it moves closer to being blacklisted by the search engines. This is what most SEO campaigns have focused on in the past, because the goal was to get great backlinks from sites with authority.

However, the number of backlinks a website has and the entire concept of page rank is evolving and changing with every algorithmic update. Over the last few years social shares are looking more and more like the new backlinks and the authority and following of people / authors associated with a piece of content are starting to get factored into the authority of web pages. Whether this affects rankings in the SERPs is subjective on who you talk to. I personally don’t believe from my own experience that Google authorship alone has improved my rankings in SERPs, if anything it has increased Click Through Rates (CTR).

We Live in Public: The Social Graph

The Social Graph: We Live In Public

As of now, the benefits of the social graph are apparent, and it is widely accepted at this point that being active and sharing your content on popular social networks can benefit your SEO. However, the exact value of something such as a Facebook like, a “+1” on Google+ or a retweet on Twitter is unknown. Viral content that generates hundreds of thousands of these has clearly affected rankings in the past, but that doesn’t give a clear picture of the value of an arbitrary number of likes, tweets, retweets or plus ones.

Because of this lack of certainty, there’s an easy rule of thumb that anyone can follow: get as much of a business’s natural traffic flowing through social media as possible, using content as the glue that holds it all together. This is the strategy that major corporations like McDonald’s and Coca-Cola use, and these are companies with multimillion dollar budgets. They’re not going to invest six or seven figures into a process that has no inherent value to them. They still get to use Twitter, Facebook and Google+ as forms of advertising and customer relations as well, which is one of the reasons that social media would be valuable even if it didn’t affect ranking.

Is Author Rank the The New Page Rank?

There’s a new metric that’s been popping up in search engines results over the last year or two, commonly referred to as “Author Rank”. It has been speculated that author rank is being used by Google as a deterrent for link farms and black-hat content propagation to tie content to real people. In the SEO blogosphere, some have even claimed that Author Rank is actually being factored into overall PageRank, and that it is not intended as a replacement to Page Rank – but rather a means to make your PageRank a bit more accurate. Which makes a lot of sense, especially within the context of the above patent, but again… only big G knows for sure.

Author rank should not be confused with Google’s authorship – (rel=author – and / or – rel=publisher). Common factors for ranking an author involve gauging the quality, frequency, number of shares and a host of other factors that are still mostly speculation at this point. Signing up for Google’s authorship program will allow your content to become associated with you as an author in the search results.

Looking Forward to 2013 and Beyond

SEO is alive but changing and evolving. Less link building and more link earning. Less emphasis on tactics and more on strategies that put the user first. The symbiotic relationship between the link graph and social graph as well as author rank provides a glimpse into the future of search over the coming years.

2 Responses to “Link Graph vs Social Graph”

  1. Steve,

    Great article, I was curious how author rank was going to tie in to the SEO realm. Thanks for sharing.

  2. Julie Gould says:

    Great article, I found it very informative. Looking forword to reading more of your work.

Leave a Reply