Why does Alexa.com love my client's website?
The simple answer is "Because I control the design". We need to look a bit closer at that question and maybe we will find out something interesting about Alexa.
January 2012 Update at the bottom - Alexa is seriously flawed.
Original article 11 April 2011
Alexa is... Top of the Pops for websites
Alexa is a company that rates websites and
gives them a rank. Number one is the best and the worst must be about 250 million since that's how many websites there are reputed to be
in the world (Spring 2011). I doubt though, that Alexa bothers with all 250 million. It's a bit like Top of the Pops for
websites.
What Alexa thinks of b-u-p
(To prevent this page cluttering Google's results when searching for the client I will refer to my client's website as b-u-p)
At the time of writing Alexa gives b-u-p a
rank better than 900,000. 'Owz about that then? Here is todays ranking figure. Has it improved (less than) 900,000? Click the little box below to go to the Alexa page.
After you have studied the paragraphs containing Alexa's measurements, type in your own web address in the search box at the top to see how Alexa rates your website. Remember that low numbers are better. Don't tell me you are 25 million? Sorry to hear that.
So what is it that puts b-u-p in the top 0.4% of websites in the whole world? Alexa also gives b-u-p a UK ranking better than 40,000 and if there are about 3 million websites in the UK, that puts it in the top 1.4% of UK websites. Not quite so good but it is up there with its competitors.
What is Alexa's ranking?
Alexa'a ranking is a combination of the amount of search traffic, the amount of time spent on websites and the number of pages that visitors view. Alexa says:
A site's ranking is based on a combined measure of Reach and Pageviews. Reach
is determined by the number of unique Alexa users who visit a site on a given day. Pageviews are the total number of Alexa user URL requests
for a site. ...... The site with the highest combination of users (Reach) and Pageviews is ranked #1.
So it is the number of visitors and the number of your web pages they look at. Guess who is Top of the Pops this week? Google of course.
An Alexa user is someone who has the Alexa Toolbar installed on his or her web browser. This provides the user with useful information and also gathers information for Alexa's database.
b-u-p is an infant, just turned 6 months old and Google is cautious about new websites so it does not rank highly for short competitive search terms. The number of unique Alexa visitors must be very low. That means the other factors must be extremely good. Or so you would think.
Now is a good point to take a look at branded-usb-promo.com who sell branded USB sticks to see if you can tell what is so good about the website from Alexa's viewpoint.
Statistical significance
The truth of the matter is revealed on the Alexa website in their FAQs. They have
recently updated that page to make things clear. If you have a ranking worse than 100,000 then the figures are not very accurate.
To be statistically significant you need a sizeable sample of data and we don't have that with b-u-p because it is a new website. So the moral of the story don't take statistics at face value. Not without checking that you have enough data to make decisions. That is particularly relevant to website owners who look at their Google Analytics and I know all about that in my management of Google Adwords accounts.
Is that it then?
Well no. I'm not so sure I want forget Alexa's ranking. Partly because the ranking for b-u-p is good, partly
because it's good fun to watch but also because it looks like Alexa is giving me a pat on the back for doing a good job with the design of b-u-p.
Think of it this way. If you have a new website and you want to test it to see if users like it you could sit and watch them. Usability testing is an art in itself and there are specialists for that. So far only a few Alexa users have visited b-u-p and as a sample size that could be comparable with a good usability testing roster. Those few visitor have behaved in a way that makes Alexa's automated measurements produce a highly respectable ranking.
Since we know the search traffic contribution to the ranking is virtually zero, that means the ranking is statistically significant because it is based on the behaviour of the few that have visited. If the website had hundreds of hits per day the significance would be lost because we wouldn't know which element was contributing most to the ranking. So there we have it. Alexa loves b-u-p because it is a good design. It is a website designed to be found.
The conclusion to these observations is that if your website has very low traffic you could use the Alexa ranking to judge the quality of your website design in terms of usability. Well that's my theory for today.
The design of b-u-p
There are different aspects to design, in particular: function and form. When we talk about design we usually have in
mind the appearance and experiences we have of the object, the look and feel. I am no designer. Any good web designer would conclude that by
looking at this website. The visual impact of b-u-p comes from the creative talent of Belfast Web Designer, Tim Potter who interpreted my design brief and produced the visual framework for the
website at a very reasonable cost.
My contribution to the design was in the functional aspects and it still work in progress. My design brief was detailed and was focussed on my judgement of what the target user wants. Tim's interpretation shows his expertise in this area. At the moment Alexa is showing signs that we are doing well.
What of the future?
Will a moderate increase in visits make the ranking worse as the number of visits take a bigger role in the computation?
And will the ranking improve as b-u-p becomes hugely successful? Does Alexa's computation take the youth of a website into account and should
the ranking stay about the same?
Update January 2012 - Alexa is seriously flawed. Alexa is usually rubbish.
The Alexa ranking continued to improve and began to improve rapidly. Eventually it soared to be lower than 500,000 which was very encouraging.
Then I noticed Alexa reported that Top Queries from Search Traffic included product codes. The type of thing the website owner would type into Google for a convenient way to navigate directly to a product.
Having told the website owner about Alexa he visited Alexa a few times and was persuaded by Alexa that it was a good idea to load the Alexa Toolbar into his browser. He became a consistent and major source of data for Alexa and his low bounce rate, high page views and long time on site inadvertently caused the ranking to soar.
I uninstalled the Alexa toolbar at the website owner's browser on 5th January and the Alexa ranking has plummeted from better than 500,000 to 643,000 in 20 days.
How can Alexa be so flawed? This means a ranking of less than say, 400,000 is so unreliable that it is utterly useless. It still has some use as an indicator of the quality of a young web design. Just don't tell the website owner about it!
Alexa also says the top query is "branded donuts". Utter nonsense. B-u-p doesn't sell them and Google Analytics doesn't report that.
One interesting feature of the Alexa ranking is that it is used as a measure of traffic grabbing quality by some web designers and bloggers who create websites with the sole intention of selling them on to suckers to try to make money out of advertising using Google Adsense. It is easy to manipulate Alexa up to a certain level. We have inadvertently shown that getting a ranking above 500,000 is quite easy.
I shall monitor the rank and update this page for time to time and tweet about updates.
Hillsborough near Belfast, County Down, Northern Ireland. Mob 07972 637 569