I say that Alexa's numbers are fairly unreliable. But I need to qualify that statement. Let me first explain what Alexa's numbers ARE, then I'll talk about how useful they are.
Unlike Google, which measures a site's value based on how many other sites link to it (and the quality of those sites), Alexa measures your site's value (read: popularity) based on two primary items:
1. Number of visitors to your site
2. How many pages they visit
Makes sense. They have an algorithm that combines those two numbers somehow, and comes up with a page rank. Thus, if your site is ranked #234, there are 233 sites in the world more popular than yours.
Here's the real issue. The only way Alexa knows how many people visit your site is by getting people to install their "toolbar". Strictly speaking, the Alexa toolbar can be considered "spyware", since it keeps track of the websites you visit.
How many people do you suppose use the Alexa toolbar? Probably thousands upon thousands. But the reality is, for every person who uses the Alexa toolbar, I would GUESS that there are at least a thousand people who DON'T use it. So how accurate are those popularity statistics?
I'll give you a for-instance. Last year I was curious to see what kind of a "bump" I could causeQuote Puzzler to take by getting a couple of its regular visitors to install the Alexa toolbar. So I asked two people to do that. And the next week, the Alexa ranking for that site jumped from about 1.5 million to 42 thousand.
Pretty sad, eh?
Okay, so now the qualification to my remarks. The MORE traffic a site has, the more accurate the figures are - that's a simple matter of statistics. Alexa will tell you that the numbers are completely meaningless if your site is listed less popular than 100,000. In other words, if your site is ranked 200,000, and mine is ranked 3,000,000, Alexa says statistically speaking, they can't tell which one is REALLY more popular.
But above 100,000, they say they are accurate.
And I say: By getting two people to install an Alexa toolbar, I got my site listed at 42,000, and they think 100,000 is statistically accurate?
I have my doubts.
So, a round about answer to your question: Alexa stats are interesting, but don't sweat them!
http://www.virtu-software.com/ask-doug/QandA.asp?q=79
Walang komento:
Mag-post ng isang Komento