Thursday, August 17, 2006

The Long Tail of Google Searches?

Ars Technica has a great article today about the "googlearchy," citing a study that refutes the idea that the more popular something is (the higher in the search results), the more popular it will become (the more sites will link to it).

Using data regarding page rankings from Yahoo and traffic figures from Alexa, the authors tested whether this expectation held up against real-world data. They plotted the traffic vs. inbound links in such a way that the slope of the line should reveal whether page rankings increased traffic. For sites that received anywhere from about 100 to 1,000,000 inbound links, traffic and links retained a linear relationship, but the slope was nothing like predictions. Traffic increased far less than would be expected if search engines were enhancing popularity. It actually increased less than would be predicted if traffic were directly proportional to inbound links. In the end, it appears that each inbound link only increases traffic by a factor of 0.8. The results suggest that the reliance of web users on search engines is actually suppressing the impact of popularity.

This becomes even more clear when the data outside this range is examined: both at the high and low end, the curves go flat. At the low end, this suggests that low-popularity sites are retaining traffic at a rate disproportionate to the number of places that link to them. Meanwhile, at the high end, sites seem to reach saturation; regardless of their inbound links and page ranks, they cannot expand their (already considerable) audience.

Chris Anderson will love this! Google obviously makes things easily searchable, and every website is equally accessible by its users. So it makes sense that the long tail of websites should be flatter than some people expect - and the perfect example of the long tail phenomenon!

No comments: