In the heart of search conference season, there has been great discussion about how the algorithm is diversifying through the integration of social data, Query Deserves Freshness (QDF) and Query Deserves Diversity (QDD), and personalized/instant search.
While these changes have undoubtedly changed the landscape, they can often serve as a point of distraction or confusion for webmasters.
Truth be told, the core of search has been tinkered with, but has largely remained unchanged over the past four years. Search is still largely driven by domain authority, keyword rich content, and links.
The link is still the heart of search. Links are the primary driver of domain authority and the key to ranking quality content.
What has been refined significantly is the way links are calculated. It's fairly obvious and well treaded that not all links are equal, but what are the exact factors that influence authority?
We've moved far beyond the simple passing of PageRank currency. A significant number of factors affect the value of a link.
This a working list, so please feel free to add below in the comments.
Page Location of the Link
Several search engine patents have been filed that indicate page location affects weighting. Typically, the highest weighted links on a page are directly within the content of the page. The links most likely to be de-weighted are in the footer or the navigation structure.
Anchor text is largely considered one of the most important factors in determining inbound link importance. Google often seems to filter links based on over-optimization. For example, a sudden rash of 50 incoming links with identical anchor text from low quality sites is often a pretty resounding indication of paid links.
One of the more underrated components in determining link value, the text surrounding the link itself is important in determining relevancy and weighting. Beyond understanding the importance of a link, Google uses this relevance understanding to evaluate images and video.
Theme of Page/Site
Search engines respect community-based interactions and recommendations. It should come as no surprise that a link from a keyword/domain relevant site means more than one from a page with a list of loosely related links.
Many webmasters mistakenly look to bloggers as a great way to pick up link juice. From the outside, it makes perfects sense. Blogs are frequently updated and run by individuals, increasing the likelihood of receiving a response to a link request.
This is now a bit of a fool's errand. Most blogging platforms default by adding the nofollow attribute to every link.
According to Matt Cutts of Google, "In Google, nofollow links don't pass PageRank and don't pass anchortext."
Are we really supposed to believe that Google ignores the link value of the entire blogosphere? Cutts' assertion is definitely carefully worded. Perhaps these links fall into the nebulous and unexplored "social signals" section of Google's algo.
Authority of Linking Domain
Undoubtedly the most important factor, the authority of the page on which the link sits drastically impacts the weight of the link. Depending on the authority of the linking site, it can almost completely mitigate the dampening factors.
This is a bit redundant because there's definitely a correlation between age and domain authority, but older links are generally considered more valuable. Conversely, a bevy of fresh links for QDF results is vital for Google News rankings.
An incoming link from a page excluded in the robots.txt is virtually meaningless.
No Index/No Follow
Another common tactic from reciprocal link shysters, incoming links from pages that have the "noindex, nofollow" tag are also worthless.
One-way vs. Reciprocal
Reciprocal links are typically de-weighted between smaller sites, but that's a bit of an oversimplification.
When credible, high ranking websites send deep content links to each other, there's little-to-no de-weighting involved. De-weighting typically occurs among smaller, lower authority sites that suspiciously have reciprocal links pointing at each other within a two-week period.
Total Number of Links
The more outbound links on the page, the less valuable the link. External links essentially split the PageRank. Like beams of light, the more splitting that occurs, the less powerful the link. Link builders are often better served by focusing on finding content that's not inundated with other outbound links.
I can't definitively prove that markup affects link value, but it's quite possible search engines de-weight links that don't appear to be links to users. By this logic, links that look like links in the traditional sense (color change, underline, bold) will have more weight.
For those interested learning about more theoretical possibilities in assessing link value, I would strongly recommend Bill Slawski's analysis of Google's 2004 "Ranking documents based on user behavior and/or feature data."