|
One of the most fascinating aspects of the web is its dynamism. We all know
that it develops at an astonishing speed - yesterday's craze is today's old
news, and bigger and better things seem to be springing up every few days. Some
of them crumble quickly into dust, while others seem destined to tower above
the rest.
Naturally, search engines also follow this pattern. Some of the early search
engine giants remain with us today, but many of them are gone - and every so
often, a new champion seems to emerge. Recent years have seen the growth and
development of a search engine that puts all others to shame. It might have
once stood at the same level as its rivals, but there is no doubt that for now
at least, Google rules the web.
Many of the companies we work with see more traffic from Google than all the
other search engines put together, and there are more than a few Search Engine
Optimisation services who focus almost exclusively on this one engine.
What is Google's secret?
So why is Google so successful? The answer is simply that when a user goes
searching on Google, they're likely to find what they're looking for, and more
quickly than on any other search engine. Exactly how Google manages to do this
is trickier to answer, as they tend to guard their secrets well. They don't
want us to know too much about how they determine their search results, simply
because they don't want anyone to be able to manipulate their own ranking.
Of course, human nature dictates that many of us aren't satisfied with this.
We desperately want to be able to affect the ranking of our sites, and some
of us will go to great lengths to do so. We work hard to find the perfect keywords,
tweak our meta tags and optimise the content of our site to what we hope is
Google perfection.
But recently, a new word has entered our vocabulary, and is surrounded by so
much hype that very few people actually have a realistic understanding of what
it is - or what it isn't. PageRank is where the attention is focused today,
and many companies are determined to find a means of improving their magic number.
"I want to be an eight," they say, as if PageRank was a dress size
that they could grow into with the help of some heavy-duty calorie shots. Unfortunately,
it's not quite as easy as that.
So what exactly is PageRank? There's a surprisingly simple answer: it is Google's
way of estimating how important a web page is. On a basic level, Google decides
that if one page links to another, the second page must be considered important.
If one page on one site has 15,000 pages linking to it, it must be for a good
reason, right?
Page Rank is about pages, not websites
Let's begin by straightening out a few basic points. First of all, PageRank
is assigned on a page-by-page basis. A whole website does not have this score,
and different pages within a site can have very different PageRank values assigned.
Another important point is that the rating (out of ten) assigned is essentially
little more than an approximation of a given page's PageRank. The actual values
cover a far greater range than zero to ten.
Before going any further, we should take a look at the most important point
of all, often overlooked when we get caught up in the PageRank frenzy. PageRank
is only one factor that Google takes into account when displaying the results
of a search. There are still other factors of equal significance in performing
well on Google - so don't make the mistake of thinking that you would live happily
ever after if your PageRank was a little bit higher. Other factors include a
page's title, and the use of keywords within the page's text - not in the keyword
meta tag.
PageRank is still one of Google's more ingenious strategies, and is certainly
one of the many reasons that it stands head and shoulders above the rest. Partly,
this is due to a combination of two factors. Firstly that the very nature of
PageRank is difficult (but not impossible) to manipulate, and secondly that
the exact details of how the value is assigned is a closely guarded secret.
However, there is one very useful source of data - an academic paper detailing
the formula used to calculate PageRank from Google's early beginnings as a university
project. This formula will have certainly been altered and expanded over the
years, but it is generally accepted that it still represents the essence of
their PageRank system
The Page Rank Formula
The exact details are lengthy, and far beyond what I am capable of dissecting.
But the basic formula is as follows:
PR(A) = (1-d) + d (PR(T1)/C(T1) + ….. + PR (Tn)/C(Tn))
PR(A) is the PageRank of a particular page (A) - not a website as a whole.
1-d is the dampening factor, as explained below.
PR(T1) is the PageRank of the page that links to our (A) page, and C(T1) is
the number of links contained on that same page.
The formula is repeated throughout every single page that contains a link to
this (A) page.
Two important points to take into account. First of all, if you're thinking
that the formula would in practice be an infinite loop, then you're correct.
This is the very nature of the web itself, and is also why Google has introduced
the so called dampening factor.
The second point concerns the way that PageRank is awarded by one page to another.
The generally accepted means of understanding this is to consider that a given
page has, according to its own PageRank, a certain amount of voting power. If
the page in question links to five other pages, then each of the pages being
linked to receive their PageRank "award" of one fifth of the original
page's voting power. It's also worth noting that the number of links on a page
includes a website's internal links.
Link farms don't work
This makes it quite obvious that the so-called link farms, where each page
of a website contains many hundreds of links in an attempt to artificially boost
so called "link popularity", are doomed to fail from the start. In
addition to this, Google has its own system for not only minimising the effect
that these sites have, but eliminating it altogether. As the formula shows,
PageRank works as a multiplier of a site's overall value, so Google has made
sure that link farms have their own value of zero - which means that a link
from them counts for nothing, quite literally.
There is a scare story doing the rounds which claims that being listed on link
popularity sites, or for that matter any site with a large number of links,
can get your site penalised or even banned from Google. This is simply not the
case. If it were, you'd effectively be able to wipe-out your competition's Google
presence with one afternoon's work. It doesn't work that way.
Having links to your web pages on sites with a low page rank and a large number
of links means that the benefits are quite effectively minimised to zero. But
this will not detract from your current PageRank at all.
Obviously, what people really want to know is whether PageRank can be manipulated.
In the past it was often considered impossible to do so, but nowadays this is
not always the case. There are two simple factors involved:
Firstly: who links to you, and how they choose to do so. Secondly: your own
website's navigation and internal links.
Clearly, the sheer number of pages linking to you will not influence your PageRank.
Of far greater importance is the PageRank of each of these pages, and how many
links appear on them. Common sense certainly needs to be applied here. In theory,
one simple way to improve your PageRank might be to have Microsoft link to you
from the front page of their website. In practice, this might be a little difficult
to achieve.
It is already quite clear that linking out to another website, even if it opens
in a new browser window, actually involves potentially giving away a lot more
than a little space on your website. My advice would be to look at your link
exchanges as you would your food. You always want to make sure you're not leaving
yourself hungry, and if you do choose to share, be selective. Exchanging a piece
of your sirloin steak for a small piece of stale bread, shared between hundreds
of people, is far from an even trade. If you're doing so to help another site,
as an act of charity, then this is fine and well, as long as you know what you're
giving away. Choose wisely.
Well-known websites and their PageRank
Now that we have a basic understanding of how PageRank works, let's take a
look at some of the more well-known websites on the web today, and see how their
main pages perform.
Finding out a page's PageRank is couldn't be simpler. Follow the link to Services
and Tools from the Google home page, and find the Google Toolbar. After installing
the software, a bar appears at the top of the browser showing a value for each
page you're visiting. Hold the mouse over the bar, and you'll be told the page's
PageRank - a score out of ten. As already mentioned, this figure is little more
than a representation of a page's actual PageRank.
Not surprisingly, very few pages score ten out of ten, and those that do includes
the likes of Microsoft, Yahoo, Google itself, AltaVista, Adobe, AOL, Mozilla.org
and others. In other words we're looking at the biggest of the biggest websites
- and not something that most of us could ever hope to achieve!
Of course, there is a simple reason that search engines and directories have
such a high PageRank. Not only do they link to a huge, ever-growing list of
sites and pages, but more importantly, a truly staggering number of these sites
and pages link back to them. When you consider the importance of reciprocal
linking, you start to understand why they do so well. With Adobe, you only need
to consider the sheer number of web pages out there that link to a PDF file
(with links to Adobe for their free reader software), and you will see why they
have achieved such a high number.
A nine out of ten score still puts you within a very small minority of the
web. Should you be able to achieve this high a PageRank, you'll be rubbing shoulders
with the likes of MSN, BBC News, Winzip and Internet.com. We're talking about
the web's upper classes here - not really attainable for the majority of normal
website owners.
Eight out of ten starts bringing you to the "reachable" web. You'll
find sites such as CNN, TuCows, Simtel, the Association of Shareware Professionals,
the Shareware Industry Conference site and Lockergnome.
A PageRank of seven is starting to appear reasonably attainable, as long as
we're willing to work hard on the content and reputation of our site. The sevens
include companies such as D-Link, MSNBC, CNET's Download.com and our very own
SharewarePromotions.com.
Don't lose your perspective!
At this point, a little perspective might be in order. A critical point to remember
is that PageRank only plays a part in performing well in Google. PageRank's
primary aim involves ranking the results of a search - but in order to show
up in the search to start with, your site needs to be properly optimised and
have good, solid content. So contrary to popular belief, the era of Search Engine
Optimisation is far from over. It's only had a new, interesting factor thrown
into it.
Finally, a note of caution. This article has been an attempt to very briefly
summarise an enormously complicated subject. Aside from constraints of space,
much of the workings of PageRank remain shrouded in mystery. The ideas presented
are based on available data, known facts, speculation and my own experience
- but none of it should be considered as insurmountable fact!
PageRank is undoubtedly an important factor in how much traffic you will receive
from Google. It is, however, merely one component in your arsenal of tools to
win the battle for one particular search engine. Even with the constantly evolving
web, and the ever-tightening systems employed by the search engines to quantify
the usefulness of a website, content is still by far the most important factor,
and will invariably form the base on which everything else is built. Be seen,
be sold.
About the Author:
Dave Collins is the CEO of SharewarePromotions Ltd., a well established UK-based
company working with software and shareware marketing activities, utilising
all aspects of the internet. http://www.sharewarepromotions.com and http://www.davetalks.com |
