Alexa Statistics – accurate for Hits, inaccurate for Visitors
An article a few days featured a discussion on the credibility of Alexa statistics. One person suggested to compare real user statistics with the Alexa statistics. This is what I have done: I analysed hits and visits for the last six month and matched the graph with the graph given by Alexa. And the result is fairly accurate.
The result of my endeavours you can see on the left: the bold blue line is the Alexa graph, the thin lines are the usage statistics. The Alexa graph is fairly accurate for Hits and partially inaccurate for Visitors, but tends to underestimate Visitors – especially during the last couple of months. Also, when the hits fall below a certain threshold, Alexa seems to report no data at all.
Of course, debate should continue. Lots of bloggers have commented on this issue and many have had negative things to say on Alexa. However, Alexa allows to compare the popularity of two sites quite well.
See for arguments on Alexa: Perfektionist, Robert Basic, Marcelo, Sabahan, Dirk Olbertz, SEOPitfall, Peter Norviq from Google, the Technology Blog and Garry Conn.
