Forum Moderators: phranque
Can anyone please explain a little more about Alexa.com and exactly what it is I am looking at as I am unclear to what it is saying to me :S
Many thanks,
In short, the Alexa stats/ranks are not extremely meaningful since their numbers depend on users who have the Alexa toolbar installed in their browser.
In other words, their stats don't really account for the millions of people who don't have the Alexa toolbar installed.
Alexa is the only publically available traffic rankings on the web today. A reported 7 million people have the toolbar installed.
The biggest group of users are the 'amazon' crowd. The toolbar was pushed heavily to them and also through MSN at some point. Thus, the traffic rankings are a bit - ummm - female, younger, and less tech savvy than the average web user.
Under about the first 5k listings, the traffic rankings are pretty accurate because there is a wide enough sample. I don't trust anything below about 10k.
Who uses the data? Just about anyone who ranks sites.
The top 5k mesh pretty good with most all of the private rankings services (hitwise, counters...etal)
The data is also heavily used in the financial services sectors. VC, Banks, and other finance people use the alexa private data feeds heavily. Thus, if you are seeking vc funding, a good alexa ranking is almost mandatory.
How many of those 7 million have found the toolbar to be less than useful and have uninstalled it?
How many of those 7 million have found the toolbar to be less than useful and have uninstalled it?
Not to mention the folks who have Spybot or any other trojan-remover software installed that disables the Alexa toolbar by default.
I like Alexa because it gives me a very crude indicator as to site popularity. If someone wants me to advertise on their site and says it gets huge traffic, a quick look at their Alexa ranking will tell me if their claim is remotely possible.
As BT notes, accuracy in the 5-digit and up rankings in pretty dicey. On the other hand, for comparing, say, nba.com to nfl.com, it's probably fairly accurate.
one thing that will make Alexa data more accurate is heavier usage
That is completely untrue. You can't fix a biased sample by increasing it. You can only fix a biased sample by balancing it so it begins to approximate a random sample.
However, since nobody including Alexa knows the composition of toolbar user viz a viz general Internet users, there is no way to know how to balance the sample.
It is absurd to think that a device originally made to gather web pages for the Internet archive could be used for ranking purposes.
This thread was started December 19th...
...and it's not hard to see an impact in the Alexa rankings here...
[alexa.com ]
That taught me quite a bit about the accuracy of the top 500.
