PageRank Matters – But Not For Ranking Pages. Sort Of
It’s hardly surprising that a metric like Google’s infamous PageRank is so widely mistaken for a key determinant of how well a website page ranks (how high up it appears) in the search engine results pages (“SERPs”).
But it is a naive mistake to assume that a higher PageRank score necessarily leads to better rankings. And, to make matters worse, it is a mistake propagated by many within the SEO industry.
The words “PageRank” and “rankings” are so similar that any novice could be forgiven for thinking they mean one and the same thing. After all, Google achieved its meteoric success and left the competition behind largely on the back of its ground breaking PageRank algorithm which scores (nearly) every page on the web and enabled better results to be delivered to surfers.
What is PageRank?
Google assigns every page on the web a numerical PageRank score and this is a measure of how important Google rates the web page. There are two different forms of PageRank that are quoted. The first one is a quasi real-time calculation continuously updated day in day out by Google. We never see these figures. The second one is a score from 0 to 10 reported in Google’s toolbar which can be downloaded free from Google. This second form of PageRank is called “Toolbar PageRank” or “TBPR”. Many other service providers offer tools which show TBPR and usually display them as a green bar like the one shown in the image at the top of this post (which you can click on to see a larger image).
TBPR is updated irregularly and unpredictably and intervals between updates can be anywhere from days to many months. It’s important to realise that TBPR lags behind real-time PageRank and is therefore always out of date (except when you are viewing a web page that has just had its TBPR updated, of course).
So, Why Doesn’t High PageRank Mean Better Rankings?
So if PageRank is a measure of importance, surely the more important pages must be shown higher up in the SERPs? Nothing could be further from the truth! There are many other factors taken into account. So many in fact that PageRank plays a relatively minor role in determining rankings.
In a recent post over at SEOMoz (see link below), Rand Fishkin stated that on average there is a correlation of around 11% or 12% between the PageRank of web pages and the positions they actually appear in the results for any given search term. In other words, if you enter a search query and then record the PageRank (which means Toolbar PageRank) of every web page in the top 100 results, say, you will not necessarily find that the pages at the top have a higher TBPR than the pages at the bottom.
I carried out a test myself. I didn’t need to validate Rand’s conclusion but I didn’t have access to his data and couldn’t resist seeing it for myself. I tested the top 100 results in google.co.uk for a search on SEO companies. Guess what? The result came in at 11.9%, right on target. To tell the ‘whole truth’, I did 3 tests in all and the other 2 came in at 25.6% (for a search on SEO services) and 4.8% (for a search on population statistics).
Relevance is known to be an extremely important factor in determining rankings. Relevance is an assessment of how closely any web page ‘relates’ to the user’s search query. Relevance takes into account many factors including the content or words on the web page and a whole host of other things like the characteristics of the back links to the web page. This is why an important page which is less relevant to the search query might not rank above a page with a much lower PageRank score but which is more relevant.
What is Google’s PageRank Good For?
If Google’s PageRank cannot be used to understand why web pages rank where they do or used to improve rankings, then what is Google PageRank good for?
Rand Fishkin at SEOmoz explains some of the uses of PageRank in detail and there is no point in simply repeating them here so if you are interested to read more then why not read his post “What is Google’s PageRank Good For?”. It is well worth a visit to see some legitimate and practical uses of PageRank and to witness some popular myths debunked.
The low correlation between PageRank and actual rank is so counter-intuitive that it takes an authority like SEOmoz to bang the drum before many people can accept it. Mind you, in such a competitive world, even an 11% correlation may be too significant to disregard entirely.
[And who would not be influenced by the prospect of a back link from a PR9 domain? Despite having gone to some trouble to explain the irrelevance of PageRank, I was still pleased to see our home page go from a 3 to 4 in Google’s November PageRank update, a few weeks after this post was first published. 🙂 ]
Did you find this post useful? If so, why not help us create more and help others find it by linking to it using the buttons below or subscribing for updates using the RSS Posts feed at the top of the page?