Today Alexa has announced that it has made big changes to the algorithm it uses to rank websites. I don’t know if this is good news, bad news, or simply neutral. Most people have known for a long time that Alexa rankings are practically worthless. In spite of that, many advertising services do rely on Alexa numbers to assess web pages.

Why is Alexa so unreliable? The site tries to rank the traffic of websites, which can’t really be done reliably unless a website inserts code that allows someone to track the real page views and visitor count. Alexa tries to approximate site traffic by having internet users install a toolbar that logs all of their web activity. Alexa uses that data to estimate traffic from the web as a whole.

There’s an inherent problem with this. If a site gets a ton of traffic from people who don’t use the Alexa toolbar, it won’t ever get counted. There are also certain types of web surfers who are more likely to install and use the toolbar… that artificially inflates the statistics of sites visited by those surfers.

Alexa apparently is trying to address this weakness in their system with this:

In recent months we’ve heard from our Alexa users that understanding Internet usage beyond Alexa Toolbar users was increasingly of interest. Ask and you shall receive!

We listened to your suggestions, and we believe that our new rankings system is much closer to what you asked for. We now aggregate data from multiple sources to give you a better indication of website popularity among the entire population of Internet users.

I’m not sure how much the information from these other sources will influence the rankings and whether or not those are reliable either. So I’m not too excited. I’m doubtful that Alexa will reveal to us what the “other sources” are, so we can’t really judge whether the new rankings are better or not.

Both of my sites (this one and pinktruth.com) moved up fairly significantly in the new ranking. But it doesn’t mean anything to me.