I really don’t trust Alexa.
Here’s why

1. It’s Biased

Alexa’s stats are based on the data collected by the users that installed the Alexa Toolbar. Fine. But who’s intrested in downloading the alexa toolbar? What does the alexa toolbar offers?

  • a search field (present in IE and Firefox, and with real search engines)
  • popup blocking (present in IE and Firefox)
  • traffic report
  • site owner info
  • related sites ( competitors? )

as you can see, the main features here are useful for webmasters, bloggers, site owners, SEOers etc.
This is not stuff for the average surfer!

Alexa’s user base is not representative of the internet population.

2. It’s Biased

Alexa’s toolbar exists only for Internet Explorer (yes, I know there’s the unofficial firefox plugin, but nobody really uses it.) (yes, I know there must be someone that uses it, but you have to agree that’s an insignificant minority).
And this means

  • no macs
  • no linux
  • no win + firefox users
  • no win + opera users

not representative

EDIT: an official firefox toolbar has been developed. The question is if it will be used or not. I guess it will not.

3. It’s Biased

Alexa’s user base it’s composed as follows: ( from http://awis.blogspot.com/2007/02/alexas-traffic-panel-is-increasingly.html)

Country Users (previously)
China 16.44% (18.46%)
United States 14.28% (36.91%)
Brazil 3.82% N/A
Japan 3.64% (3.80%)
United Kingdom 3.11% (4.49%)
Taiwan 2.91% (1.72%)
Hong Kong 2.55% (4.59%)

Once again, not representative.

…And They Know It

This is from their about page (bold addedd by me).

Keep in mind that this is the best thay can say about their service without telling something that’s a complete lie. Just try to imagine what’s the real situation :)

  • Our users are disproportionately likely to visit sites that are featured on alexa.com such as amazon.com and archive.org, and traffic to these sites may be overcounted.
  • The extent to which our sample may overcount or undercount users of the various browsers is unknown. Alexa’s sample includes users of Internet Explorer, Firefox and Mozilla browsers. The AOL/Netscape and Opera browser is not supported, which means that sites operated by these companies may be undercounted.
  • The extent to which our sample may overcount or undercount users of various operating systems is unknown. Alexa sample includes toolbars built for Windows, Macintosh and Linux.
  • The rate of adoption of Alexa software in different parts of the world may vary widely due to advertising locality, language, and other geographic and cultural factors. For example, to some extent the prominence of Chinese sites among our top-ranked sites reflects known high rates of general Internet usage in China, but there may also be a disproportionate number of Chinese Alexa users.
  • In some cases traffic data may also be adversely affected by our “site” definitions. With tens of millions of hosts on the Internet, our automated procedures for determining which hosts are serving the “same” content may be incorrect and/or out-of-date. Similarly, the determinations of domains and home pages may not always be accurate. When these determinations change (as they do periodically), there may be sudden artificial changes in the Alexa traffic rankings for some sites as a consequence.
  • The Alexa Toolbar turns itself off on secure pages (https:). Sites with secure page views will be under-represented in the Alexa traffic data.