Alexa traffic rankings are based on two primary components: (1) the reach (the number of unique users) and (2) the number of page views for a given site. Alexa computes these values on a daily basis and calculates the geometric mean of the two quantities over time. The official rank of a site is therefore based on this geometric mean, and hence theoretically reflects both the number of users who visit that site as well as the number of pages on the site viewed by those users.

Here’s the catch: only users who have installed the Alexa toolbar onto their browser have any impact whatsoever on Alexa rankings. In other words, the Alexa ranking only reflects the activity of those users. This is all well and good, excepting the ease with which these Alexa toolbar users can cheat the system.

“Gaming” Alexa is disturbingly simple. One has to merely get a few buddies to install the toolbar and surf a site every day, or, better yet, enlist the aid of Bots (computer programs that simulate human activity) to refresh several browsers every half hour or so. Within a week or so, the said individual’s target is bound to go up several hundred spots in Alexa web traffic rankings.

While Alexa programmers are constantly working to increase their program’s credibility, it is nonetheless in web users’ best interest to not put too much stock in a website’s Alexa ranking. Don’t assume that just because a website has an Alexa ranking of 100,000 it is necessarily more popular than a website with, say, a ranking of 450,000. On the other hand, it’s probably safe to assume that a website with a ranking of 100,000 is more popular than a website with a ranking of 6,273,000.

When it comes to Alexa - or any computer program, for that matter - it is always best to remember that, while the program itself may be nearly perfect, it can be manipulated by humans who are themselves a far cry from perfection. Employ your discretion when studying electronic results, and do a bit of your own research if you can regarding the websites in question. Remember: no one ever benefited from taking a computer program’s word at face value.