Archive for June 14th, 2010
Written by Eric Enge
On May 11, 2010, Google was granted a new patent that basically states that all links on a page do not have to carry (or pass) the same weight. The concept is that the value a link should pass to a target page will be largely based on the probability that a user would click on it (hat tip to Bill Slawski who wrote a great post deconstructing Google’s reasonable surfer patent).
Some simple examples of how this might work include:
- Links to unrelated content are highly unlikely to be clicked
- Specific types of content, such as “about us” or “privacy” pages, may be more—or less—likely to get clicked, depending on context
- Location of the link on the page (above the fold should get more weight than below the fold)
- Emphasis to make links more prominent, such as changing size or color, should add weight
These are just a few examples, and Bill’s post provides a much more comprehensive list. The question is, how does this affect the way that you look at SEO? Here are some thoughts.
Emphasize your most important links. When you create a new page, or modify an existing one, consider which outbound links on the page you want to pass the most value to. On the home page of the site, these might be links to your bestselling products. Perhaps you have a nice list of these in the right column of a three page layout, and it is currently below the fold on the page. Since these are the products which make the most money for you, consider moving the list to the left or middle columns, and figure out how to get it above the fold.
Leverage image links. For my first controversial idea, imagine the impact on image links. Images are far more likely to get clicked on. Does that mean that image links pass more value than text links, even without the anchor text? Probably not as yet, but it is an intriguing idea..
Use embedded links. Links embedded in the main content of the page are more likely to get clicked than those elsewhere on a page. Remember those key product links I mentioned? In addition to emphasizing them for more visibility when a page is initially displayed, consider integrating one or more of them into the content most likely to be seen by a user when they arrive at your web page.
De-emphasize navigation. It may be that basic global navigation links could pass less value for many reasons, including the fact that people tend to develop “snow blindness” with page elements that are frequently repeated. Navigation links also may be considered to have a lower degree of relevance (for example, a link to your automotive products from your lawn and garden products page).
Forget footers. Footer links will certainly be less likely to pass value. Not only do they have the snow blindness problem, they tend to be way below the fold.
Skip the ads. Anything that may be perceived as an ad, whether or not it is an ad, or is labelled as such, would be likely to be devalued.
Link only to relevant content. While this is not covered by the patent as far as I can see, I’d be willing to conjecture that a page linking to non-relevant documents could be seen as less relevant itself to the original content of the source page. Taken to its extreme, this could mean lower rankings could result from linking to pages that are low in relevance, even if they are not necessarily “bad” pages.”
Reconsider purchased links. Devaluing links to non-relevant content would provide further discouragement for authoritative or trusted sites from selling links. Link buyers will likely see less value from those links on the right rail, or in the footer, and will press harder for in-context, in-content links. Higher quality sites are generally not willing to provide such links.
As a practical matter, processing CSS to understand the details of page layout is a challenging task for Google, or any search engine. Patrick Altoft of Blogstorm argues that Google may use click data from Google Analytics, the Google Toolbar, Google Reader and other sources to evaluate link value.
My own opinion is that Google is surely going to find as many signals as they can to help them understand the importance of a page’s outbound links. If they are using click data as Patrick argues, the users themselves will identify which links are the most important ones. The only underlying requirement is that the page has enough traffic for the click pattern data to be statistically significant. But, of course, pages with material traffic tend to be the most important pages on the web.
The concept of how a reasonable surfer would behave appears to have a lot of merit, and the search engines are investing everything they can in better ranking systems, and ones that reduce the impact of spam. But ultimately, basic usability guidelines still apply. Understanding how a user sees your site and their behavior when they interact with it has always made sense. Now consider the idea that this can influence the ranking of a given page, but also the value of the links it provides to other web page.
Written by Eric Enge