The sometimes delusional cycle of Web 2.0 companies and VCs looking at Alexa rankings, does often acknowledge that Alexa is a bit skewed, as if its out by perhaps 50%.
Well its a boat load skewed, Alexa is actually only about 5% accurate if one uses data from Gawker.
Because Gawker is transparent about page views, and has a property whose readership is part of the Web 2.0 scene, Valleywag (Trivia fact – I chose the name Valleywag) and one that definitely isn’t, Deadspin, Alexa’s accuracy can be correlated to real data other than Comscore.
Valleywag traffic: 600 thousand page views per month
Deadspin traffic: 4.5 million page views per month
According to Alexa, however, Valleywag ranks twice as highly as Deadspin, with a rank of 5,000, compared to Deadspin’s 10,000 ranking.
Which means that Alexa skews tech. sites such as Web 2.0 favs by a huge factor of 15 even within the top 10,000 sites where the accuracy is higher.
In short, Alexa is almost useless for websites outside of the top 1000, and no sensible investment or reporting should be influenced by it.
Website Statistics and Traffic Graphs comparing www.valleywag.com and www.deadspin.com