In Defense of Alexa

Ken's Blog

Alexa must be the most maligned, useful service online. Webmasters seem to enjoy dismissing it as "useless." They usually give their own sites as examples, as if a sample size of one means anything in the world of mathematics. Or they explain how it can be gamed, as if that totally invalidates this valuable tool.

In fact, when you do a proper study involving tens of thousands of sites, such as we do with Site Build It! sites, there is an excellent correlation between the number of visitors that a site receives and the Alexa traffic ranking.

And Geoffrey Mack, who writes the Alexa blog, showed a different kind of correlation between ranking and traffic in a post two days ago.  His post focused on the tight correlation of traffic and Alexa ranking for two sites with excellent rankings (Top 1000) over a period of time.

While useful, it was a touch disingenuous. Any two sites can be chosen to prove a point. And top-ranking Alexa sites have, due to the volumes of traffic, less scatter (i.e., a better correlation.)

So let's get to the bottom of this. Alexa is a tremendously valuable tool, if you understand how it works, including the nature of its sampling and therefore its bias. With that under your belt, you will find true value in Alexa.

Alexa is not, of course, perfect. Perhaps Webmasters seek perfection? If so, then many things in life, business and yes, even the tech world, are going to disappoint.

Alexa results do, of course, display scatter, meaning that some sites with better Alexa traffic rankings get less traffic than other sites with poorer rankings. That's normal in the world of statistics.

And the worse (higher) the Alexa traffic ranking, the more scatter you get (i.e., the wider the variance between Alexa ranking and actual traffic).

So what? That's normal in the world of statistics, too. Factor it in when interpreting.

What about those who worry about "gaming" Alexa? My answer is, "to what point?" Manipulating Alexa to generate a better traffic ranking does not change your actual traffic. So why do it?...

To fool potential advertisers? Any half-smart advertiser will, in a one-month trial, be able to detect impressions and clicks. So suppose a (falsely) great Alexa site has no traffic. The results will quickly show up... No impressions and no click-throughs. No payment and no ongoing relationship.

Conclusion
Alexa is a powerful tool, if you know how to use it. Those who dismiss Alexa as "useless" are simply missing out on a great tool.  Please see this article that explains more about Alexa, how and why we use it, and how you can get the most out of it.

All the best,