As if I didn't already know, I've just seen absolutely irrefutable proof that website ranking systems like Alexa are extremely inaccurate. By only having access to data from people using their toolbar, their stats can be extremely skewed.
Today, for some reason, ShrinkThisLink (the free link shrinker which I created and run) has spiked to an Alexa "Daily Reach" of 10 per million users. This makes absolutely no sense, since according to Google Analytics and Awstats, two stats tracking systems which I use and which both collect actual data (awstats from log files, Google Analytics from javascript embedded in pages) both show that if anything, traffic on Saturday (which is the day the Alexa information is apparently for) was BELOW previous days. The only explanation that I can come up with is that for whatever reason, a higher number of those visitors were using Internet Explorer (which my stats do seem to confirm) with the Alexa toolbar installed. This proves that it doesn't take a lot to cause reasonable amounts of change in the Alexa statistics, since when you average out the higher percentage of IE users and the lower number of overall users, you should end up with about equal numbers - demonstrating just how skewed these systems are.
Today, for some reason, ShrinkThisLink (the free link shrinker which I created and run) has spiked to an Alexa "Daily Reach" of 10 per million users. This makes absolutely no sense, since according to Google Analytics and Awstats, two stats tracking systems which I use and which both collect actual data (awstats from log files, Google Analytics from javascript embedded in pages) both show that if anything, traffic on Saturday (which is the day the Alexa information is apparently for) was BELOW previous days. The only explanation that I can come up with is that for whatever reason, a higher number of those visitors were using Internet Explorer (which my stats do seem to confirm) with the Alexa toolbar installed. This proves that it doesn't take a lot to cause reasonable amounts of change in the Alexa statistics, since when you average out the higher percentage of IE users and the lower number of overall users, you should end up with about equal numbers - demonstrating just how skewed these systems are.
Comment by Everard,
Thursday 10th of August 2006 09:43:29 PM
[Reply]
I'm pretty sure wikipedia explains why the numbers are all mesed up
Post a comment:
Sorry, comments have been disabled.
Sorry, comments have been disabled.