Collin Maesson at realsceptic.com has written a couple of articles about the value, or lack thereof, of using Alexa rankings to monitor blog popularity. You can read them here and here. Since Collin is in the business of software development, he knows about such things as statistical monitoring and web page ranking services and about how best to monitor readership. Knowing something about readership is important if you care about your readers. If you know your blog is losing readership, you can do something about it (if you care).
Recently HotWhopper's Alexa rankings began to drop off from a low of around 382,000 late last year (lower is better) to the current 1,003,243 (as at Feb 4 2014) - which is where it was back in the first week of July last year. Thing is, back in June last year, HotWhopper was only six months old and, according to Google Analytics got 29,358 page views for the month. In January this year it got 59,800 or double June's page views. So it seems rather strange that HotWhopper's Alexa ranking went down so much while its page views doubled.
Collin noticed the improvement in Alexa ratings of his own blog and was curious as to why HotWhopper appeared to be dropping in popularity. He asked if I'd allow him to analyse our stats, which I was more than happy for him to do. He wrote the results up on his blog here. After Collin approached me I did a Google search and came across this article, which had almost identical results.
Alexa doesn't monitor readership directly. It doesn't use cookies and it doesn't have access to the servers on which websites are hosted. It can only guess the popularity of a website. As I understand it, Alexa 'guesses' the visits to any website via the number of hits counted by people who have the Alexa toolbar installed. So if a site doesn't get visitors from people with the Alexa toolbar installed, it won't rank very highly. In other words, Alexa doesn't rate the popularity of websites at all. It only measure the tendency of websites to get visits from browsers with the Alexa toolbar installed. Collin explains this in his first blog article on the subject here. (Other site statistics services use different direct and indirect ways to count visitors. It could be cookies, javascript, server logs or a page element or any combination of these.)
I'm guessing that the sort of people who use the Alexa toolbar are not the same people who visit sites like HotWhopper. The people who use the Alexa toolbar don't care (or maybe don't know) about the spyware that gets installed with the toolbar. Anyway, Collin speculated that Alexa may have got HotWhopper's ranking wrong compared to his blog, writing in part:
...It’s probably because I’ve written about Alexa a couple of times and I have a high ranking for my Alexa content in Google searches. My website is often displayed on the first page in Google, I can see that with the statistics I have in Google Webmaster Tools. What you need to realize about this is that people who search for information about Alexa rankings tend to be Alexa users, which means that there’s a good chance that they also have the Alexa toolbar installed.
So this article is a little experiment. It's about Alexa, so it might pop up in web searches for Alexa or Alexa ratings or Alexa toolbar or Alexa popularity. If it does then maybe, as Collin surmises, HotWhopper will get a few visits from people who have the Alexa toolbar installed. If that's the case, then HotWhopper's ranking on Alexa should rise. We'll see.
PS I'm burying this by pre-dating it by a few days, so most normal HotWhopper readers probably won't see it. It's just an experiment after all.
Sou
3 February 2014
I already wondered why I had not noticed this post, while WUWT got me interested in Alexa.
ReplyDeleteAnother interesting fact about WUWT, the favourite local topic, is that the moment WUWT introduced the Alexa spybar, the ranking of WUWT went up through the roof. Like this post argues that is most likely a measurement error.
Later Anthony Watts sold his Alexa stats as proof of his popularity. As the yearly Wordpress statistics showed still later, this was wrong. Surprise. There is actually a small decrease in the number of page views of WUWT.
Thanks, Victor. I buried this article because it's just an experiment. So far not much change, although the weird dive in Alexa rankings seems to be tapering off. I've added the Alexa toolbar to one of my little used browsers, so it's no longer a controlled experiment :)
DeleteI have the feeling that Alexa does quite a lot of processing, trying to improve the noisy and biased signal into something more useful. As a consequence it may take a few months until you see an effect.
DeletePeople with the Alexa toolbar also need to be able to find this article. Thus you also need links to this article, to make Google think this is an important page. Thus hiding it too much may be counter productive.
The dive only took a few weeks - from November to now. In fact the dive sped up a lot lately - the ranking has halved in not much more than a week - which is really weird. There's no rhyme or reason to it that I can see. I know from past experience it can improve just as quickly.
DeleteBTW I only hid this article by backdating it a few days. It's not a main article so I didn't want it to distract from the blog. It's got lots of mentions of Alexa etc. It won't rank all that highly on Google because this isn't a webmaster blog or a computer blog.
If I remember what is known about their algorithms correctly the Alexa ranking is calculated with data over a 3 month period. It will take a while before you'll notice something, if anything.
DeleteCollin, I read that too. However the "3 month period" conflicts with the sudden drop in ranking since November despite HW visits increasing, and especially over January when HotWhopper's page views hit an all time high while the Alexa ranking plummeted.
DeleteGood information about Alexa Rank improvment tricks.
ReplyDelete