10 Jan Determining Reachability – A new Google patent
A new Google patent concerning ‘reachability’ has been published here on the US Patent Office website, it was filed in June 2010 and may have become part of the Google algorithm already.
The abstract describes “a system designed to determine a resources reachability score.” The score is arrived at by looking at the resources (webpage’s) outbound links and judging the resource by the quality and user interactions of the secondary resources (outbound links.) The resulting score will then be used as an “input signal” for ranking processes.
The score calculated for the primary resource (which is defined as an image, document or multimedia content) will be calculated by looking at three pieces of data.
- User interactions with the secondary resource.
- Median access time of the secondary resource.
- Click through rate.
The patent goes on to define some of these user interactions in more detail, talking about long clicks –
“A timer tracks how long a user accesses, views, or “dwells” on the resource. For example, a longer time spent dwelling on a resource (i.e., a long click) can indicate that a user found the resource to be relevant for the query.”
and short clicks –
“A brief period viewing a resource can be interpreted as a lack of relevance. In some implementations, the click data is a count of clicks of all types for a particular query and resource combination. In some implementations, the click data includes a count of each click type for a particular query and resource combination.”
This data also applies to video so where the video is 30 seconds long, a short click may be 5 seconds whereas if the video is 5 minutes long a short click may be 30 seconds.
What conclusions can we draw?
For the last 10 years the debate about whether linking out from a web page can affect your rankings has raged and Google’s algorithm has certainly ignored or favoured linking out in different ways during its various incarnations. This patent seems to be following Google’s usual approach in encouraging webmasters to create sites and pages for the visitor rather than the search engine spiders. It will certainly make people think harder about linking to low-quality websites for SEO purposes. If a link remains on a page but is rarely clicked and when it is it only registers a ‘short click’ then it will affect the ranking of the whole page. I believe this can only bring benefits to the internet as a whole encouraging websites to link to good-quality relevant websites.