How do Alexa rank sites?
For some time now, I have been collating statistics on four websites.
My main area of interest is in the Alexa trends, to see if they bear any relationships to other factors, such as traffic.
It would seem logical that a ranking should track traffic, but this is not apparently the case with Alexa.
There are some interesting things to note about the above graph.
The first is that the top line is pretty much stable, with no more than a 5% deviation throughout the period. This site however has experienced a 50% growth in traffic over the time of the graph. How come this is not reflected in the Alexa stats?
The bottom line is the growth in ranking for this site. It has a very low traffic rate, and very few followers. I post on it a couple of times a month, so it should be languishing at the bottom. Yet the trend is constantly upwards. Traffic to the site over the period is more or less a flat line, so why the growth?
Of the other two sites, one (the green) is a very popular blog though posting is sporadic, while the other (the red) is more of a niche site with regular postings. Again the stats don’t reflect the figures I would have expected.
I hope Alexa know what they are doing, because I don’t!
I think Alexa works like an opinion pollster with a sample it assumes to be representative of the wider population. An opinion poll needs only to include a couple of people from a minority group, who may be completely unrepresentative of the wider population, to give a picture of that group being of far more significance. So if The Other Fellow had someone who was part of Alexa’s sample who checked his blog every day, then his ratings would rise.
That could be, but the disparity between the actual figures and those that would be expected is still very wide.
I’d love to know as well. According to them my 3 month change is up 20%. I’m not sure how they got that information because I block their crawler.
From this page though they say that they get their data from users of their toolbar:
http://www.alexa.com/help/traffic-learn-more
If that is the case, then we have to assume that within the 50% increase in traffic, not one was using the Alexa toolbar. That would be odd?
It would be odd, but possible – how many Alexa users are there? It’s like opinion polls, how many times have you been asked how you were going to vote? I’ve never ever been contacted by a pollster
I think you should put up a blank page and advise the users of the Alexa Toolbar you know to visit each day to seen how far it would climb.
Hah! That isn’t a bad idea. My only problem would be in remembering to visit it every day.
I have an Alexa rating of 264,000 and Google Analytics gives me 1,784 visits in the last month – about 50 a day. I am entirely convinced that the Alexa rating arises from a friend in England who visits the site every day, frequently comments and uses the Alexa Toolbar.
If Alexa ranking can be determined by only one or two regular visitors, then it’s hardly a good yardstick? You are rapidly catching up on Head Rambles!!
In the past three months, I have cut my Alexa rating by one-half – from 487,000 to 243,000 – and it has not attracted a single extra advertisement 🙁
Ian: The recession bites again. In the same period I think I sold two. I used to pick ’em up at around one or two a day!