Peter Norvig is the director of research at Google and a post on his blog weighs in on the whole Alexa is Awesome vs. Alexa is Flawed debate.
He has published some of his web stats on his blog showing that, although his site gets about twice the pageviews of Matt Cutt’s blog, Matt’s blog has an Alexa pageview ranking that is roughly 25 times better than Norvig’s. That would seem to indicate some unreliable data, at least in terms of measuring web site traffic.
Basically, a site gets credit for an Alexa pageview ONLY when a visitor views that site through a web browser which has the Alexa Toolbar installed. Meaning, no Alexa Toolbar installed, no Alexa pageview credited. And, since the Alexa toolbar only works with Internet Explorer, that means that if most of your visitors are using Firefox (or Opera), Alexa is not giving you credit for pageviews.
In addition, we had a post a while back about how one can manipulate Alexa ranking and how that fact also really devalues its benefit as a true measure of web site traffic.
Unfortunately, many of the companies which a blogger could use to make some revenue from their blog (ie. PayPerPost, Text-Link-Ads and others) use a web site’s Alexa ranking when determining the value of an ad placed on that web site.
So, if your blog does not have a high Alexa ranking, you will not be compensated as well by advertisers, despite the fact that your ranking may not be a true representation of your web site traffic and that an Alexa ranking can be artificially inflated :(.