Green pixels - who needs 'em?

Source Title:
What price PageRank?
Story Text:

A ClickZ article from Mike Grehan about the confusion witnessed when telling certain marketers about the 'worthlesness' of Google PageRank.

What's more, one question I was asked last week had me slightly dumbfounded for a moment: "If you're saying I can't value links that I buy based on PageRank, how do I value how much I'd be prepared to pay for them?"

This is the stark reality: People still actually buy and sell links valued against Google's PR score. Which is ridiculous. It's just buying into the myth.

Once I realized the guy was serious, my answer was simple. "A link from a popular and well visited Web site can be invaluable if it sends you qualified traffic that converts. The true value of a link has absolutely nothing whatsoever to do with PageRank."


I was there when he said it

...and the look of panic on the faces of these (quite senior in some cases) people responsible for the web strategy of some major brands was a joy to behold. It tells me that I am so far ahead of the competition, I could take a 3 month holiday, and it just wouldn't matter. Wish I could sell THAT upwards :)

We just ditched pure PR as an in-house metric. Take that how you will....

Still and but ....

I of course know that Mike knows his stuff, and puts up a very approachable presentation that masks great technical insight, but it seems to me that this is slightly disingenuous. It may no longer be just about PR, but that don't mean that a link is, these days, just about delivering targetted traffic.

Themes, blah, LocalRank, blah, authority, blah, signals .. and on and on.

Just look at the links on the left :)

Google Pagerank Check
Huge selection of Powerful Links Relevant Links on Authority Sites


Yes, it is obvious that PR doesn't matter.

I just searched for "mortgages" and some great PR3 sites which have a lot of quality content (but haven't done much link building) come up in the top ten. PageRank is just a "myth"! We can all rank with great content alone! Gee how nice that Google is no longer relying on link popularity algorithms, that makes things easy for us squeeky clean whitehats.

So long as the Google

So long as the Google toolbar displays that little green line, there's always going to be a market for it. All they have to do is remove it.

Not using any PR indicators anymore - no toolbar or extensions - but I work links by volume anyway, so individual PR of sites was never a particular issue.

The Pagerank Challenge

Although I agree with Mike that people place too much emphasis on pagerank by itself, I think it’s overly simplistic to say that pagerank means nothing at all. Of course it means something. It is a rough estimate of the amount of link popularity a site has. It’s true that Pagerank doesn’t display whether a site is relevant to yours or not, but then again, it’s not supposed too. It is supposed to show the amount of link popularity, and it does that well. It is just one tool. Like a stethoscope to a doctor, or a wrench to a mechanic. I doubt there is any professional link builder out there that does not at least glance at the pagerank of a site prior to doing an in-depth analysis of the other factors.

I commend Mike for trying to spread the word to the novice SEO’s out there that they need to look at other factors besides pagerank. I don’t however agree that making such a bold statement is in the best interest of some. I think it would be better to explain that the pagerank bar is a good indicator of link popularity but nothing else, and then go on to explain the other important factors that should be looked at.

In order to prove my point that Pagerank has value (whether it is a small value or not is for another challenge) I challenge anyone here to answer the following question honestly.


Lets say that you had a friend that owned two websites, and this friend offered to give you a free link from just one of the websites. You had to choose which site you wanted the link from without doing any analysis. You had to base your decision solely on the pagerank of the site. One of the websites had a pagerank of 1 while the other had a pagerank of 7. Which site would you pick?

Build a strawman, then miss the boat

Certainly no one with any understanding of this business believes the "true value of a link has absolutely nothing whatsoever to do with PageRank". Pagerank is a representation of value factors that are deducible, its not pixels.

There isn't a novice SEO in the world who thinks pagerank is everything and its just foolishness to pretend such an animal exists. Likewise there isn't an experienced SEO who thinks pagerank is nothing. Pretending such strawmen exist is just a waste of time.

Well written, Ssite.

Well written, Ssite.

>> there isn't an experienced SEO who thinks pagerank is nothing

No, there's one right here. Let me say this again for the hard of thinking


I think far too many people confuse causal links with casual links (re-read that sentence). When you lok at a high ranking site, and you see that it has a high PR score, do you think "Oh, that must be why they rank well then?". If so, you are missing the point. High ranking sites tend to have relatively high PR because of some overlap in how the two are calculated, but high PR is neither a necessary nor a sufficient condition for good rankings

>> Which site would you pick?

Neither. If I haven't done any analysis, I don't want the link. For instance, what anchor text am I going to get? What is the theme of the linking page? Where do the other links on that page point? What anchor text are those links using? How many other links are on the linking page?

Unless I can answer some of those questions, I'm not interested. And before you pipe up Jarrod, that IS an honest answer. You can buy a PR7 link cheaply enough these days on the rare occasions that you just need some green fairy dust for whatever reason, so I'd NEVER take a blind link

PR optimisers.

On the other side, I've come across lots of newbie SEOs who (seem to) optimise for PR alone. Maybe it's a vanity thing, or they really believe that a #1 ranking will soon follow after they promote their homepage to PR6 (?). Nothing that 10 minutes of active reading on a SEO forum can't cure, but they do exist - if only briefly ;-)

Keep beating the strawman


And there still isn't an experienced SEO who thinks pagerank is nothing. This posturing does no one any good.

To simplify it, if you think a PR1 page gets crawled as often as a PR9 page, you are sniffing too much glue.

It's not going to bake bread, but PR is something.

TallTroll, Thanks for the


Thanks for the reply. Sulum ut suus. Or roughly translated, "to each his own".

I have no doubt that you are an experienced SEO, and it is true that the anchor text and the theme are just as important, if not more important then the pagerank. Does that mean that pagerank has no value?

With that being said. I can tell you story after story of a single non-related high pagerank link taking a site's rankings from nowhere land to top ten. As a matter of fact I can probably tell you about 5000 stories, each from one of my repeat customers.

If anyone doubts that a single non-related link can have an effect on a site I give you another challenge. Go buy Traffic Equalizer, and build a 3000 page site. then go find 1 quality PR7 of any topic, and point it to your new site. Give it 3-9 months and watch your Google adsense money roll in.

I dont condone traffic equalizer, as matter of fact, we wont sell links to traffic equalizer sites, or any site that is machine generated. But the fact remains that it is true that a single non-related link will add value to a site. I know one site that is making $3000/month from a single PR7 link.

"To Each His Own" is a great saying. It really bothers me when people try and preach "their way of doing things" as though it is the only way. I know a lot when it comes to this industry but I would never be so bold as to say "This is the only way to do it, Fool".

Jarrod Hunt

Lets just clarify this...


Is absolutely true - no arguments.

This does not mean the same as 'PageRank is worthless'.

If you are using 'toolbar reported PR' as a metric, then you are building a castle on sand.

Why would anyone use a figure that is:
*Out of Date
*Easily Faked
*Only a small part of the equation

.... as standard of measurement.

Sure, PR matters - but as a metric its like using a pine cone to predict how many sales your new range of umbrellas will make.

I'll accept that PR gets you spidered

... but visits from Gbot don't equate to rankings, or traffic. Of course, an unspidered page cannot rank at all, by definition, so I'll concede a very small part of that argument. When you see a well ranking site that has good PR, don't leap to the conclusion that the two are more than peripherally linked.

The way that PR is calculated, and that rankings are calculated have some crossover. This means that high ranked sites often will have high PRs, without there necessarily being any meaningful connection between the 2.

>> It really bothers me when people try and preach "their way of doing things" as though it is the only way

This is a public forum. People can read here, and elsewhere and make up their own minds. I'm reasonably sure that my posts here have substantial basis in fact, and are therefore valid regardless of any opinion. I also know that there are many different business models that you could use to be successful, and if something works for you, great, but just because it seems works, that doesn't always make it so...

Something Michael Martinez

Something Michael Martinez alluded to a while ago was that a Google ranking relies on both popularity and relevancy. It's a nice abstract way to explain search engine ranking criteria to non-SEOs:

Popularity is solely down to off-site factors; i.e. backlinks, regardless of anchor text. Popularity is purportedly shown on the Google Toolbar. How accurate that is is a different matter.

Relevancy is a mix of off and on-site factors; including the topic of the linking site, anchor text and on-site content and SEO.

Both factors are vital to rank well, but considering them as separate entities can explain low-PR site that rank well, and high-PR sites that don't.

Is this Google world-view so far off the mark? It certainly doesn't seem to contradict either TallTroll or Jarrod's posts ...

I like that explanation a

I like that explanation a lot jetboy

Having listened to many more knowledgeable people than i talk about this, recently here and elsewhere, that seems to be a good summary..


?Hey folks - I think that

?Hey folks - I think that a very important point has been missed here, or glossed over.

There are TWO Pageranks.

1. The Pagerank Number(the little green bar) that the public sees via the google tool bar. This Pagerank runs on its own server it is viewable by the public, although rarely updated and google has said publicly, several times that this Public Pagerank Number is considered to be for “entertainment purposes only”.

2. Then there is the Pagerank that the public does NOT see, it is used in the actual calculations that determine a pages rank for a keyword set. This google Pagerank is “important” if you are concerned about your pages ranking. But there is no way to tell exactly what this Pagerank Number is, unless someone at google tells you (but you can come close with an educated guess or two).

BTW - I know how Mike feels, I started talking about this over a year ago, back then some folks wanted to feed me to the lions for blaspheming against google... lol


Toolbar = unreliable entertainment metric.
Real in-house Google PageRank = useful

I'll take high PR for a few reasons, all else held constant.

One, higher PR gets you get more timely crawls. Much has been made lately that, in the world of syndication and copy theft, that the first document to get recognized by Google trumps all others (even the original author) in potential dupe content issues.

Two, as GoogleGuy pointed out, PageRank is often used as a "tie-breaker" in 302, reputation, or canonical issues.

So, like lots0, I pretty much ignore toolbar PR but the benefits of a strong "in-house" Google PR does still present benefit.

TallTroll, Just to clear


Just to clear things up, when I said " It really bothers me when people try and preach "their way of doing things". I was referring to Mike not you, I should have made that clear.

Thanks for all of the debate, it was the first one in about a month that enticed me to chime in. I would get bored out of my head if I could never find someone to go back and forth with on an issue.

Jetboy, thats a good summation which coincides with my view. My only reason for joining this post was to try and sway some of the hard rights towards the center. I certainly do not preach that pagerank is anything more then a shell of its former self, but at the same time it is short-sited to think that it has no value at all, either as a rough indicator of link popularity or as 1 of several ranking factors.

To me pagerank is like watching the weather report on the news. It is by no means perfect but it sure is a good place to start when trying to plan a weekend outing.

A metric is not all or nothing

Some people insist that everything is black and white. If you can't understand the rather obvious value of toolbar PR as a metric, then you flunked math in school and have a hard time understanding the depth of this business.

Just to use the obvious example again, some type of content requires regular, even daily, crawling. To get this to occur, it is critical or at least helpful to aim at least one high PR link at these pages or site sections. Multiple, even scores, of lower PR links will not have the same effect.

Now of course the I-don't-understand-PR crowd will not understand that one of the many metrics of a page's value is how it can help the value of other pages or other website sections. The value of actual pagerank in just this one example is plain to anyone who takes a moment to look, and toolbar PR, while flawed, is still an excellent tool to determine value, even if it is just one tool in a big toolbox.


if you can't understand the rather obvious value of toolbar PR as a metric, then you flunked math in school

I'd back my mathematics against yours any day.

Websters Definition:
Metric: a standard of measurement

Toolbar PR cannot be used as a standard of measurement, it is at best an indicator of which sites need further investigation.

You don't build metrics on 'entertainment' value only figures.

The fact that google now

The fact that google now puts the toolbar Pagerank on its own server speaks volumes, if your listening.

With the publicly viewable toolbar Pagerank on its very own server it can be COMPLETELY disassociated from the actual ranking process. Making the use of the toolbar pagerank as a “metric” or measurement of a sites “worth” useless at best.

Look, we all know why the toolbar pagerank is no longer useful/accurate, the same reason that google no longer shows the current up to date backlinks, google finally figured out that letting us view the actual Pagerank and the actual backlinks gave away to much of their algo and made it way too easy to figure out how to manipulate the results.

So google no longer lets us see the current backlinks or the "REAL" Pagerank that is used to rank your pages.

For those of you still wanting to use the toolbar pagerank as a "metric", I can tell you that we have at least 16 pages that are ranked by the toolbar as PR7 - but we know (have tested) that the pages are not valued that much as they don't have the backlinks to qualify them as true PR7's unless google has lowered the value of Pagerank(oppsit of what they are doing) and these pages do not have the ranking or the linking power that they should have if they were "real" PR7's.

If your useing the toolbar Pagerank as any kind of "metric" that involves money, you can bet the googlites are laughing their asses off at you.

Again, of course if you insist on black and white

Then you can't get what is available from the toolbar, or from pagerank itself. There are few 2+2=4 things in this business, although clearly some people crave such simplicity.

Pagerank remains critically important, particularly in longtail industries. Toolbar pixels of course matter less, but they still are an obviously useful metric to help in decisionmaking. Denying such is too absurd to take seriously. Lots of alleged marketers are dead lazy, and need something painted on their forehead to "get it". Toolbar PR is a minor tool that augments other data, none of which is black and white on its own, but all of which can be used as a measurement of value to some degree. Is it a definitive metric, of course not, but people looking for one need to wake up and smell the coffee. The business is much more complicated and has no easy, clear, 100% reliable answers.

SSite,I feel so close to


I feel so close to you right now :)

Anyways, If anyone is curious I have about 5000 SEO reports we have ran over the past few months that show that there is still a relation between quality and High Pagerank. The relationship may be indirect but the correlation between high quality sites and High PR is irrefutable. This should not be interpreted to mean that all high pagerank sites are quality, but that the trend tends to be, Higher quality Higher Pagerank. It is real easy from our reports to tell which ones are fake.

Yes, there are some sites that are high PR that arent worth the links that they bought to get them there, but that doesnt mean you should discount all High PR sites as crap.

As Ssite says, there are few 2+2=4 things in this business. Or any business for that matter.

An indicator is a good indicator if you understand its weaknesses and its usefullness in the broad scheme of things.

Man, it is uncanny how this debate is starting to sound just like the debate the US polititions have over the usefullness of what is probably the most famous indicator in the world. the CPI or Consumer Price Index. For those of you that dont know what the CPI is or what its weaknessess and strenghs are, I urge you to read the following As I read through it I cant help but laugh at the similarities.

Jarrod Hunt

It's rare for me to disagree with Mike but...

I was watching a program last night about people that cheat at blackjack. I actually don't really like blackjack and I'm not a gambling person but there was stuff about card counting in there that struck me as interesting. What it boils down to is some fancy brainy chappy realized that the probability of winning the game shifted depending on what cards were removed from the deck and that each number of card (i.e. A,1,2...Q,K) had a different value by which the probability shifted. Obviously sitting and playing a card game whilst doing fraction multiplication isn't particularly easy! So he broke the cards into groups and gave them a simple +1 +0 or -1 scoring system. This makes it pretty easy to do the maths. What struck me is that it threw a heck of a lot of information away, but by doing this it made the calculation simple and it turned the odds in the player's favour.

PageRank strikes me as about the same thing. It abstracts away a heck of a lot of data (e.g. the relevance of the link, whether it is a "this page is good" link or a "this page is bad" link), but it still serves as a viable source of information because on the whole it provides useful information. There's a heck of a lot of room for improvement but the data that's abstracted away makes it an easy calculation.

The connection between toolbar PageRank and actual PageRank is much debated. Frankly, I normally avoid such debates because the answer is blatantly obvious. Whilst the value the Toolbar shows is clearly not the actual PageRank, it is also clearly derived from it. In general, if we look at the PageRank of pages with the toolbar then it is what we'd expect. It's therefore obvious to conclude there is a relationship. It's also obvious to conclude that as the Toolbar doesn't show the actual PageRank then some information isn't being provided to us.

Therefore, we come to the question of using Toolbar PageRank as a metric to buy links. Clearly the answer is simple: if I was to buy a single link or a few links then using PageRank as a metric would be of little worth to me. However, if I was to buy a LARGE number of links then buying based on PR would cause my collection of links to tend towards being those that are higher trafficed and/or of better quality. At some point the cost of my time reviewing possible links is more expensive than the loss of quality caused by just buying based on PR alone.

Additionally PR, although much of the worthwhile information is abstracted away, carries one more quality that is extremely useful in comparison to some metrics. It provides information before buying the links :) Granted there are other metrics that do that but it's certainly worth bearing in mind. Mike's quotes show this is precisely where we diverge:

"Once I realized the guy was serious, my answer was simple. "A link from a popular and well visited Web site can be invaluable if it sends you qualified traffic that converts. The true value of a link has absolutely nothing whatsoever to do with PageRank.""

How can he know for sure whether it "converts" until the link has already been bought? Granted you can make an assessment for just a few links...but many? Mike talks about the value of a link (that value is established). I fully agree, but at the point you originally buy the link it is not established. Given we cannot know the true value of a link at the point of purchase the question really should be does the Toolbar PageRank value tend to relate to the true value of a link - the answer is it must. Sure, if you've bought a link from a site before then the data of previous conversions is a heck of a lot better but if you haven't then I see nothing wrong with using Toolbar PR as a guide.

Scale is everything.

Summary: Toolbar PR is, of course, a valid metric for the purchase of links in some circumstances. That said, anybody who relies on just one metric is as foolish as anybody who ignores one :)

Looks like a split

Looks like a split decision.

Ok, I agree that there is some current small "value" to the green bar. But only because as Chris pointed out, the green bar derives its current numbers from the pagerank numbers.

But, the longer the green bar is "entertainment" only and remains removed from the actual ranking process the less accurate the green bar will become.

Who knows, it could be that soon, clicking on those smile and frown faces on the google tool bar will have more effect on the green bar pixels than pagerank... would that be such a bad thing?

Bravo Chris, Very well

Bravo Chris, Very well articulated.

This is nothing new

Mike's been saying "Don't look at PR" ever since I've known him. I published this interview with him in Sept 2003. He's said the same thing at every conference he speaks at... what's interesting is that it continues to spark controversy from people who can't imagine NOT using the green bar as a "guarantee" of a good link.

I think the people Mike is

I think the people Mike is trying to preach too are a small group. I think he is used to preaching to fanatics who think Pagerank is everything; people that approach him at conferences with their big dreams of pagerank driven paradises.

I think he writes articles such as this last one in order to try and convince these people to think about something besides pagerank. I think his message comes across clearly to those who he is targeting, and comes across as "way overboard" for those of us who understand that Pagerank isnt the holy grail.

Maybe in his next "Anti-pagerank" article he should begin with a forward dedicating it to all of those people who are obsessed with pagerank as being the only thing that matters for Search Optimization. Maybe then all of us Pagerank moderates wouldnt get overly aroused and defensive.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.