Andrew Watson

Could Google be Considering Ranking Pages on Link Position?

Posted on 12. Jun, 2010 by Andrew Watson in Google, SEO & Search Engines

On May 11, 2010, Google was granted a new patent based on the ranking of documents based on user behaviour and/or feature data.
Initially, whilst interesting to hear Google had applied for a new ranking patent, I did not pay too much heed to patent 7,716,225. It essentially appears to be an enhancement to Google’s existing page rank algorithm.

However after a little further consideration, I think Google could be considering a fundamental change to their ranking algorithm. This in turn could have consequences for SEO tactics, and to an extent the way in which certain website elements are designed, built and developed.

The patent is based on ‘the ranking of documents based on user behaviour and/or feature data’. The ranking of documents based on feature data already occurs to a certain extent within the current architecture of pages. One current component of Googles ranking process, is the ranking of pages based on HTML markup e.g. H1 tags, p tags etc. and it could be argued that these tags highlight page ‘features’ of sorts.

However it is further reading of the patent, that shows which elements Google is specifically concerned with. It is the feature data associated with a plurality of first links, within a plurality of first source documents, that point to a plurality of first target documents – essentially how links appear and how they link to other pages/documents.

This leads me to think that Google is fundamentally considering changing from an algorithm based on links having a equal weighting to one in which each link within a page is not treated as equal.

This could be further backed up with the release/acceptance of HTML 5. New HTML 5 elements could see Google push the ranking of ‘links’ and ‘features’ further. Consider for example the new HTML 5 elements of <nav>, <header> or <footer>. What would happen to pages diligently designed, built and optimised, in which an extended ‘footer’ architecture is used? If it is defined as <footer>, as it should be semantically, Google could simply decide these links are less valuable than other links, and consider improving their search results by applying less weighting to these <footer> elements.

The other aspect of this patent is concerned with ‘user behaviour’. This could also see a change to Google’s ranking algorithm, with those links that are used least, being treated as less important and thus receiving less weighting.

In the example of the <footer> tag, this would could make sense, as a privacy policy or terms and conditions etc may indeed be less important than the site content… however what about content that appears ‘below the fold’. Due to it’s position on the page, it could be ranked as less important, as it is less viewed/used than a link ‘above the fold’. That could have far more concerning consequences for those involved in SEO, and the design and development of websites. More content may have to be pushed above the fold – leaving less room for expansive imagery.