I have 2 websites of ecommerce in France and for these websites I have a strange behaviour of the alexa results. Here are some statistics about the websites :
Unique Visits January 2012
Number of Search Results google
Links to my site
ALEXA Score
How does it come that website B with 1/3 of the visitors of website A have a much better Alexa Score ( x2 ) then website A?
This is discussed else where (reddit) and I am writing this on the top of my head. Alexa toolbar is installed on certain browsers (IE sucks six), Alexa uses this specific sample for ranking. Reddit users where surprised to find a huge proportion of traffic from certain geographies like India than first world countries.
The unbiased ranking published may be the one taken at a larger scale using statistics from CDNs and Google analytics.