Just for kicks I was checking out the Alexa traffic stats for joseph.randomnetworks.com. They let you compare your stats against other sites, so I started looking at how my stats compared to a few others. Most of them easily eclipsed my numbers. Then I compared it to ma.tt:
I’d always had a healthy level of skepticism about how accurate Alexa was. Once I saw this graph any hope of usefulness that Alexa may have had was completely gone. I know for a certainty that not only does ma.tt get more traffic that this site does, it gets a ton more traffic that I do.
The level of disinformation that Alexa is providing has gone from being just useless to actively harmful. So remember, friends don’t let friends rely on Alexa traffic statistics. For anything.
9 replies on “How I Know That Alexa Traffic Stats Are Worthless”
I think it’s a matter of sample size. For sites in the bottom million like ours there numbers are going to probably be pretty off, but for larger sites like WP.com the sample size is large enough that it tracks actual traffic pretty well.
With out Alexa, no one is going to give a rank for sites. It may not accurate, still it have some value in finding rank of your web site.
Hi Joseph,
My experience with Alexa is a bit different.
I use it together with other analytic tools, and get a pretty accurate reading for a site with about 8000 uniques per day.
See comparison:
http://www.alexa.com/data/details/traffic_details/simania.co.il?site0=simania.co.il&site1=steimatzky.co.il&site2=booknetshop.co.il&site3=bookme.co.il&y=r&z=3&h=400&w=700&c=1&u%5B%5D=simania.co.il&u%5B%5D=steimatzky.co.il&u%5B%5D=booknetshop.co.il&u%5B%5D=bookme.co.il&x=2008-06-12T06%3A50%3A09.000Z&check=www.alexa.com&signature=tIl1fDcXLXWl1Mzt5EQxjn9KjOE%3D&range=3m&size=Large
Alexa does not have accurate results for ma.tt for some reason.
For many other sites it gives a good estimate for comparison.
Like any other pole it should not be taken too seriously but trends can be extracted from it.
Regards,
Lavi Avigdor
Simania.co.il
@Matt –
Good point, especially for my site. This brings up the question then, where’s the cut off for useful numbers? Is anything beyond the top 100 (or 500) highly suspect?
@Stefin –
Which is one of the reasons why I posted this. Over reliance on Alexa for “regular” sites is a bit scary.
@Lavi –
I compared it with ma.tt because I knew that site gets a lot more traffic than I do, but wasn’t miles and miles ahead of me (like any of the sites in the top 500).
Hi Joseph, you’re right – Alexa is not really showing accurate numbers. Somewhere I’ve read that Alexa stats are based on Alexa Toolbar reports. I wanted to find out and installed the Toolbar. Instantly, all of my sites went up – as I’m causing disproportionate traffic on them while checking, posting & maintaining.
I’d say it depends very much on the topic of a site and it’s geographic location – as Alexa is not really well-known in some parts of the world resp. readers of non-internet blog topics.
In spite of its impreciseness it may harm webmasters in monetizing their sites, as advertisers use Alexa Stats for checking the popularity of publishers.
I agree with all of you, Alexa is not an accurate way to measure the traffic rank of a site.
Since, Alexa data is collected using Alexa’s toolbar, Alexa has a bias for sites related with I.T. (since most IT people uses their bar. Usually a doctor or a lawyer does not take care about alexa numbers). And also have the other issue about webmaster visiting their own sites.
I think the best way to measure traffic ranks on a site is using the unique visitor numbers (not even hits nor visitors). A site is popular is a lot of people visit it, not if a reduced amount of people visit it a lot of times.
Well, these are my 2 cents. Thanks for allow me to post in your blog.
It does worth it for websites under 10,000. Under that number, then its more accurate. I did install the Alexa toolbar and one of my site rank went from 700,000 to 150,000, So I can say that its a useless tool.
Compete and Quantcast are more accurate, if you are looking for just U.S traffic.
Use compete.com it will give you more accurate results for comparison purposes. It can’t count all the hits or visits but it is a pretty accurate tool.
http://siteanalytics.compete.com/randomnetworks.com+ma.tt?metric=uv
I find Compete to be more and more suspect lately. Like Alexa, the larger the site traffic, the more accurate the rankings. This wasn’t so a year ago at Compete which seemed to be right on the money with several smaller sites whose traffic stats are available to me personally.