Whilst this is theoretically possible, it is also unlikely.

To understand why the change has happened

• You have to understand SEO, external and internal linking, and have an advanced knowledge of how PageRank works. This gives me another chance to drop a very blatant affiliate link to Stomping The Search Engines 2 which is probably the best value (just a measly $1) high quality SEO training online. (http://andybeard.eu/Recommends/SEO-Training.html)
• You need to study Twitter Linking structure over a period of time – snapshot SEO is for cowboys – even what I am writing here isn’t going to be highly accurate, because I don’t have access to analytics, server logs etc, and telling a script to analyse every page of Twitter just isn’t viable

I am not writing this just for another opportunity to pimp an affiliate link, but I am sick of poor SEO information out there among bloggers. I have suggested SEOs need to clean up old information, but to be honest, this is about as fundamental as it gets.

I do have a fair understanding of SEO and linking structures, and I do monitor changes, not just the pretty toolbar PageRank Google updates every 3 months or so, or the ranking of my Twitter profile in the SERPs.

Here are a few of my more recent tweets on the topic:-

Changing account name and retaining some of your link juice

@ed Good to see the move accomplished – you might do well to fill up a page of tweets to yourself from Next Instinct (best of ed) = links10:26 AM Jan 10th from Power Twitter in reply to Ed

AndyBeard

A little basic SEO training

In case SEOs are forgetting something, external links+indexed pages = PageRank – content quality irrelevant7:40 PM Mar 25th from web

AndyBeard

The president would not be happy if I could game an indented listing

It should be very easy to make any twitter account appear as an indented listing of a high profile twitter user e.g. Barak Obama8:21 AM Mar 25th from web

AndyBeard

Twitter making changes to Meta Titles was not an April Fool’s joke, but the titles for individual tweets is far from optimal

If you add your twitter profile to every place you get links for your blog, of course it is going to rank high9:51 AM Apr 1st from web

AndyBeard

Whilst I might have followed people, that isn’t necessarily a vote

Dear @ev @mattcutts The people on my Twitter “Following” blogroll are not recommendations – pls nofollow/discount5:58 AM Apr 4th from web

AndyBeard

Little minor details affecting individual tweets, and maybe overall profile rankings due to the recent “replies” changes

@BradWest unfortunately that solution isn’t, as you lose the “in reply to” links10:31 AM May 13th from Power Twitter in reply to BradWest

AndyBeard

So What Changed At Twitter?

Quite simply, the default blogroll links

Previously they were based on 2 criteria

• You had to be following the person
• They were displayed in account creation order

Thus popular tech bloggers who had early beta access to Twitter, plus Twitter founders always made up a high percentage of the default blogroll links across the whole Twitter network.

Even accounts with just a few thousand followers would have a high pagerank if they were created early enough, because they most likely followed a few high profile Twitter users such as Robert Scoble, and Twitter founders, so they gained the benefit of very powerful links.

How Is The Twitter Blogroll Calculated Now?

I wish I knew so I could work out how to take the best advantage of it – seems almost random, other than you have  to be following the person.

What I do know is that is still doesn’t represent my personal choice of the best people to follow, and all the links should be nofollow unless I can determine who can be included.

About PageRank

The PageRank formula isn’t complicated math, but it gets applied recursively for the global internet, which is one of the reasons why Google needs huge banks of servers.

Google calculates PageRank constantly – rankings based on it and 200+ other factors change constantly

What appears on the Google Toolbar (Toolbar PageRank) is

• Just a rough approximation
• Can be manually overridden by Google in the case of a penalty
• Is only updated every 3 months
• Looks to be linear progression, whereas the real numbers are logarithmic in nature

Posts like this one on The Next Web or this one on So Web are correct that there has been a change in visible ToolBar PageRank, but the analysis of why it happened is certainly lacking.

Have you noticed how PageRank updates frequently happen around the same time as Google Engineers are at conferences?

Every single time I write a post mentioning PageRank, I get comments about PageRank not being important for ranking.

These comments very rarely differentiate between toolbar PageRank and the PageRank of whatever kind Google stores on their servers and upgrades on an extremely frequent basis for every page. I know from first hand experience that the toolbar PageRank has very little to do with rankings, and is manually manipulated based on Google’s commercial goals.

PageRank By Any Other Name…

The Ranking Factors article at SEOmoz in many ways skirts around the issue, referring to Toolbar Pagerank, and then ignoring the concept of what is real PageRank by splitting things down into multiple related items.

• Link Popularity within the Site’s Internal Link Structure
• Global Link Popularity of Site
• Topical Relevance of Inbound Links to Site
• Link Popularity of Site in Topical Community
• Global Link Popularity of Linking Site
• Link Popularity of Site in Topical Community
• Topical Relationship of Linking Site
• Internal Link Popularity of Linking Page within Host Site/Domain

The only direct question specific to PageRank was:-

• PageRank (as measured by the GG Toolbar) of Linking Page

Aaron Wall in answering this question actually gave a response hinting at the real importance of PageRank

The toolbar is perpetually outdated, but Google uses PageRank values to help set crawling priorities and to determine if a document should go in the regular or supplmental index.

Some Simple Questions

• Can a page rank without a Title tag?
• Can a page rank without any internal linking?
• Can a page rank even on a new domain?
• Can a page rank without direct external links?

Ultimately with almost all the ranking factors, it is a balancing act, but with PageRank or however you wish to describe “Google Juice”, it becomes a little more fundamental.

No PageRank, No Google Juice = No Index

I realise if you take a purely theoretical stance, that if you created a 1000 page site full of original content, and then point Google to the sitemap for that site, that Google might index the whole site, and if you remove that link, some of the pages might remain indexed for a short or long period of time.
I haven’t done the test, but a random surfer in theory could land on one of the isolated pages, if Google chose to keep the unconnected pages in the index.

PageRank Flow & Real World Indexing

I need a real world example to demonstrate how important juice flow around a website or blog is important, and I decided that Michael Martinez effectively was asking for this by saying:-

I do absolutely nothing to make SEO Theory “SEO friendly”. It is better indexed in Google than most SEO blogs.

Take that for what it’s worth.

I am always up for a challenge, especially when Michael went on to say

My complaints about the poor quality of Google’s search results stem from Google’s willful, deliberate segregation of the Web into two categories: Preferred Pages (Main Web Index) and Supplemental Pages. Preferred Pages are always shown first in search results regardless of how much more relevant the Supplemental Pages may be to queries.

Actually Google seems to have 3 types of pages

1. Main Index
2. Supplemental – apparently being phased out, but it could be all FUD, on that I agree with Michael on
3. Not Indexed

Michael forgot about the pages that are receiving so little juice, Google doesn’t even bother indexing them, even on sites that are “better indexed than most SEO blogs”.

It is possible that Michael is doing some kind of indexing test, or he could also have selectively decided that he doesn’t want his old content in Google’s index.
Thus I am not going to link directly to the following pages which would damage his test results.

That being said, Michael did ask to be quoted on it, and to quote him I am sure he would want the person doing the quoting to provide good, if not conclusive evidence for or against his stance. I am not going to claim conclusive evidence, but at least I have spent a little time on this reply.

Michael links to his date based archive pages from every page in his sidebar, thus they should be receiving a fair amount of juice. However that juice doesn’t flow very deeply and he only has 5 posts on each page of his archives.

If you go just 3 pages deep, Michael starts to have indexing problems.

http://seo-theory.com/wordpress/2007/02/page/3/

Every article listed on that page is not in Google’s index… at all!

http://seo-theory.com/wordpress/2007/02/14/5-ways-to-launch-new-multiproduct/

Not Indexed By Google

http://seo-theory.com/wordpress/2007/02/13/yesterydays-seo-advice-at-todays-prices/

Not Indexed By Google

http://seo-theory.com/wordpress/2007/02/09/how-to-end-search-engine-slavery/

Not Indexed By Google

http://seo-theory.com/wordpress/2007/02/08/how-to-build-long-lasting-trusted-value/

Not Indexed By Google

http://seo-theory.com/wordpress/2007/02/07/best-kept-secrets-in-seo/

Not Indexed By Google

It seems the content on Michael’s SEO Theory blog isn’t as well indexed currently as you might expect, but as I mentioned earlier, that might be due to experimentation

I Have Pages Not Indexed As Well

I decided a while back it would be hard to write a post like this without having some pages of my own to point out, so I did a number of things.

• I didn’t make extensive structure changes to improve things based on my WordPress SEO articles
• I switched off translation plugins
• I don’t include unique article descriptions
• When I upgraded to WP2.3, I didn’t include Custom Query String, so my archives are not as flat as they used to be – I should note there are a few versions of CQS now available for WP2.3+ including Custom Query String Reloaded
• I have been using underscores with my tag_pages rather-than-dashes

I had a tough choice back in October, after being hit with a sizeable fake Toolbar PageRank penalty (currently -3) – continue making changes to my site structure to improve search engine performance, or keep the site relatively unchanged.

It is hard to say whether the penalties are/were material unless you bite the bullet and not make changes required by Google.

The only change I decided to make was to not include CQS when I upgraded to WP2.3+ – I decided that this would allow me to eventually provide some examples of pages falling out of the index, and then I would be able to demonstrate how I improve site structure to fix the problem.

With the changes Google made to the reporting of supplemental results, or if you believe them removing supplemental results altogether, it did take a little while for things to settle down.

I was waiting for a little deeper indexing activity to be visible, and then to wait a week or so for that activity to show in results. I did point out a few months ago that Google Webmaster tools provides these indexing charts, but the scales are still broken.

http://andybeard.eu/2007/01/mybloglog-social-networking-opportunity.html

Not In Google Index

The big difference is that I had to go back 7 pages in my January archives from last year to find a page that was no longer in the index, and my date based indexes are not on my sidebar on every page of this domain.

http://andybeard.eu/2007/01/page/7

Related Links Are Transitory

Related links certainly help passing juice to older related content, but eventually even if you list 10 related pages, and use very specific control of related pages using a plugin such as Simple Tags, the related posts become superseded.
I will probably end up tagging this post seo, wordpress, linking, linking structures, pagerank, ranking factors

I have used most of those tags in the past, thus it is most likely that I will get 10 related posts, but also that some previously related posts will become displaced on the list, and that change will not just happen on this page, but all pages on this domain that are related.

Deep Linking to Older Content

Deep linking to older core content always brings a little fresh life back to them, and gives them a fresh injection of Google juice. Once you get to 500+ pages of content, it becomes harder and harder to give life back to all of them, and thus only what you class as “pillar” content gets a much needed burst of life.
There is a constant ebb and flow, 2 steps forward, one step back.

Temporal Factors

Maybe there are temporal factors taken into account by search engines, and some kind of temporary PageRank assigned to new content.
What I do know is that if content is buried deep in your archives, so deep that it doesn’t receive any juice and isn’t indexed, then a link from that page is totally worthless.
An old link on a TBPR PR10 domain that is buried deep in the archives might still have some value, whereas being 30 pages deep on a blog that receives very little link love, or maybe an archived forum post, isn’t going to be worth much, if anything.

Google may remember old links that have lost juice for a period of time after they have been removed. Donna has spent some time looking into this.

To Be A Contender, You Have To Be In The Game

If your pages aren’t in Google’s index, they can’t rank for anything, even long tail queries.

To be in Google’s index, pages really have to have a certain undefined amount of juice, no matter what other factors you gain merit for.

Thus PageRank is the primary Google Search Ranking Factor, because it is the only factor you 100% have to fulfil to have a chance for your pages to rank in Google’s search results.

To give you a good parting analogy, all plants need water – different plants thrive with different amounts of water, and you can give a plant too much water – I don’t know if you can have too much Google juice, but you might have too much over a short period of time… a downpour which washes away the soil.

I am not talking about original words, but original ideas which can be based either on experience from experimentation or what I would describe as reasoned theory?
Experience based on following conventional wisdom just doesn’t count – you only need to read one SEO blog for accepted best practice, it doesn’t really matter which one of the mainstream SEO blogs you choose. You might choose a few more for industry news with differing opinion.

The kind of blogs I love to read are those with substance, even if I can’t fully understand all of them first time around.

Now with that said I am going to highlight a few choice articles.

WordPress SEO Goodness

I am not the only one working with WordPress SEO dynamic linking structures. Here are 2 great examples

Sebastian has recently (thank goodness) switched over to a new domain and a WordPress installation and it seems creating some customized WordPress linking structures.
I haven’t fully worked out the benefits and drawbacks of what he is doing, but it will certainly ensure all his content gets indexed.

Halfdeck discusses SEO for Fun, and in my experience he only writes useful unique content. He uses his blog as a testbed for linking structures and also provides a tool for linking structure analysis.

Dan Thies recently let the cat out of the bag about Google proxy hacking but I actually want to highlight his SEO ebook. It is full of solid information – it gives you a good grounding for some of my more geeky posts that discuss creating such structures with WordPress. (it is free)

It might be a little easier for some people to understand than Revenge of the Mininet (and the bonus dynamic linking ebook) which for the last 10 months I have been strongly recommending. (Also free)

Jaimie is also a SEO worth reading (though he has been taking some R&R due to illness) and wrote the code to fix the proxy hacking – hopefully he will have a WordPress plugin for those that need it soon.