SEO Book has an interesting article cautioning about reading too much into Alexa rankings. While the statistics provided by Alexa are widely touted by many marketers as a “must-use” tool, there are several good reasons for treating it all with a grain of salt.

The practice marketing section of a popular site for psychotherapists also has a revealing look at Alexa statistics, with an even more in-depth commentary on the inherent limitations of Alexa data. Referring to humanly-edited directories which make a big deal of their Alexa rankings (Joeant.com comes to mind), the site comments:

For example, other things being equal, the more people involved in the production of a website, the more the Alexa data will be skewed by the involvement of those ‘insiders’, as distinct from external, ‘real world’ visitors. Some humanly-edited directories are great examples of this effect, especially those with very poor designs that make it difficult for editors to accomplish anything without a high number of page views. A site can easily get a boost to its Alexa rankings simply by opening up the production process to a team of volunteers, whose activities immediately add to the page view data being collected by Alexa. (I think it would be fascinating to see the ‘real’ traffic information for some sites with large armies of human editors, once the activities of the large production team are subtracted out.)

The bottom line is that yes, Alexa is a great tool — but don’t believe everything it tells you, particularly when what it tells you isn’t backed up by your own server logs!