This post was originally titled “Is PageRank Sculpting Dead & Can Comments Kill Your PageRank”
Following a confirmation post from Google’s Matt Cutts today, it seems PageRank Sculpting as practiced by many SEOs is effectively dead, and comments, even using links with nofollow CAN have a negative effect on the amount of PageRank that can be passed on to your internal pages.
Link to updates from Matt Cutts plus tips on how to continue PageRank sculpting effectively.
Has Google in one quick swipe removed all benefit of Dynamic Linking (old school term) or PageRank sculpting (when it became “trendy”), and potentially caused massive penalties for sites nofollowing links for user generated content and comments?
I have left a few comments on various blog posts over the last few days, especially on SEOmoz and Twitter, but though it important to solidify some thoughts here, and potentially add a little more perspective.
PageRank Sculpting Formerly Known As Dynamic Linking
The idea of controlling the flow of “Google Juice” around a website to pages that matter, or to other sites that matter has been around for a long time, at least as early as 2003 when Leslie Rohde (Stompernet Faculty) was calling it “Dynamic Linking”.
Those were the days before “nofollow” and you had to use Javascript to accomplish the internal linking control.
In the past I have linked to Michael Campbell’s Revenge of the Mininet which also provides access to Leslie’s original Dynamic Linking membership site. They are both now free, (Michael used to charge $79.95 for his ebook)
I am sure I have sent 1000s of people to Michael’s newsletter signup page over the years, but I am equally confident that 90% of the visitors didn’t sign up. I don’t receive any kind of payment recommending Michael’s work, or Leslie’s dynamic linking.
In many ways I look on at least a passing understanding of these groundbreaking ebooks as required reading for any of my SEO articles
From the comments I see on most SEO blogs, and even many of the articles, I am quietly confident that these seminal works haven’t been truly understood, and of course the content rarely gets referenced.
So I am going to do something unprecedented, and I hope Michael won’t mind – the page does appear in the SERPs.
Once you get there, you can pick up a password to access Leslie’s Dynamic Linking site.
Leslie has always had this great disclaimer:
DISCLAIMER!
Some of the techniques and technologies described in the foregoing are not without their pitfalls and potential unintended consequences. If you are new to web business, inexperienced at search engine optimization, or do not feel comfortable with HTML and Javascript (at least at a rudimentary level), you should not attempt to employ the advanced techniques shown here!
Access The Source of SEO Knowledge
Michael Campbell
I have just given you access to 4 or 5 year old information that in all likelihood is more advanced than you will find discussed on 95% of SEO blogs and forums, and whilst I don’t from principle/ethics join any private SEO content area to avoid conflict with what I blog about, I am quietly confident that it would still be looked on as advanced content for members only, or not even covered in such depth.
But that is just a trickle of knowledge compared to direct access.
Leslie Rohde
I am still eagerly awaiting an update to Leslie’s Dynamic Linking suggested on his SEO Software site.
A major update to this material is currently in process owing to some recently discovered changes in the way Google is processing links. Look for an announcement early next year (2009) — the changes will likely revolutionize on-site linking techniques … again!
Leslie also has an SEO Strategy blog, which he actually updates once in a while. I am going to have to explore the blogging platform he uses, Pebble.
There are 2 other ways to learn more from Leslie:
- Just before the New Year Leslie put together a new site “Optimize Recession” where he introduced the idea of “Zone Based SEO” – I mentioned it on Twitter.
Zone based SEO might seem obvious at first, but it allows you to systematize and possibly even automate specific SEO campaigns, especially now it is possible to extract ranking positions from Google referrer data. - Stompernet – Leslie is one of the faculty at Stompernet, who now offer very progressive SEO and marketing training. Start off just by joining their newsletter and the 7 Deadly SEO Sins course plus free videos, and possibly get their “Stomping The Search Engines 2″ course for $1 (plus a trial to their Net Effect magazine)
The Death Of Dynamic Linking With Javascript?
Of everything that has been discussed about Google making changes to which links they will follow and count going forward, how they handle javascript is probably the one that is worth the most consideration.
The first I read about it was on Search Engine Land in an article by Vanessa Fox (who used to work for Google as a member of their webmaster team) covering Google Javascript Links.
(Note: I know that anchor text is very contrived, but SEO is about helping people find what they are searching for, not snake oil or gaming Google)
Having given a great link, I can justify grabbing a small code example
Some examples of code that Googlebot can now execute include:
-
<div onclick="document.location.href='http://foo.com/'">
-
<tr onclick="myfunction('index.html')"><a href="#" onclick="myfunction()">new page</a> -
<a href="javascript:void(0)" onclick="window.open ('welcome.html')">open new window</a>
Javascript That Is Still Dynamic?
This will probably work
onclick=”myfunction(‘jkhhjstdysd’)
Have myfunction() within a file loaded in the header, or preferably in the footer for faster page loading. You would still want to use the CSS that Michael and Leslie suggest for usability.
Somehow define which destination ‘jkhhjstdysd’ refers to, and that could potentially be broken down into components.
If Google somehow cope with that, and possibly easier would be to just use pure external javascript that pulls in some XML, but that then complicates things if you want to mix real links with dynamic ones.
But this is moot if nofollow actually still works.
Does Nofollow Still Work For Dynamic Linking or PageRank Sculpting?
I am going to lead with the freshest insight I have read, Dan Thies (also Stompernet Faculty) thinks things are being blown out of proportion.
Here are the primary opinion pieces and coverage I have seen, though I am sure there were plenty more
Google Loses “Backwards Compatibility” On Paid Link Blocking & PageRank Sculpting
Didn’t comment, was too busy looking for other coverage.
Google (Maybe) Changes How the PageRank Algorithm Handles Nofollow
My comment on the post:
I can only think that Google have been misinterpreted.
If I have a blog post with 300 comments, and have the links nofollowed (my blog is dofollow but example), then there would effectively be juice lost due to the comment links.
Links are valuable, because they add to the relevance of a comment made, because a reader can follow them to find out more about the person.
However they also form part of disclosure.If this is only for internal links, there are major problems because often a link will be nofollowed because it points to a tracking link that is also blocked by robots.txt
Any sensible knowledgeable webmaster is going to nofollow those links, because they serve no purpose for Google in their current state, and who wants to turn them into hanging pages.
That may also be a workaround, if Google handles links blocked with Robots.txt differently
No Clarification Forthcoming from Google on Nofollow & PageRank Flow
My comment on the post:
I can’t see any evidence that this is affecting external links.
Wikipedia is still a black hole of link equity. If this affected Wikipedia external links, we would see some effect, because due to recursive calculations through internal linking, it could potentially reduce their juice pool by as much as 30%
It would also affect the Ebay group with sites such as epinions.
If it has any effect, it will be internal links only.
The amount of juice lost could be similar to dangling or hanging pages, and due to many poor SEO articles suggesting robots.txt for duplicate content, Google Webmaster guidelines suggesting robots.txt for search results, and just ignoring obvious signals such as TBPR.
Yes, any smart SEO could spot the toolbar showing some green on pages blocked by robots.txt and work things out for themselves.
But the juice goes into the internet ether, and due to macro PageRank calculations, comes back.
If anything, this will help Google surface more long-tail content, and sites with lots of pages will benefit.
Live Blogging Of Matt Cutts @ SMX
Is What’s Good For Google, Good For SEO
Important to read both articles because it gives a clearer insight to the exact wording on lots of different issues.
Alternative Reactions
PageRank Sculpting is Dead? Good Riddance
PageRank Sculpting – Recent Matt Cutts Video
I have to strongly point out that this video was recorded before SMX, and maybe even a week or 2 before. It is on the official Google webmasters channel on YouTube, thus has probably been vetted in some way for accuracy.
Full Transcript
Matt Cutts on PageRank Sculpting:
Rand, In Brighton, and that might be Rand Fishkin, I don’t know asks:-
What are your views on ‘PageRank Sculpting’?
Useful and recommended if implemented right, or unethical?Well I wouldn’t say it is unethical because it is stuff on your website – you are allowed to control how the PageRank flows around withing your site.
Erm, I would say that it is not the first thing that I would work on.
I would work on:-
- Getting more links
- Having higher quality content
Those are always the sort of things that you want to do first.
But then if you have a certain amount of budget of PageRank, erm… you certainly can sculpt your PageRank.
I wouldn’t necessarily do it with the nofollow tag, although you can put a nofollow on a login page, or something that is customized where a robot will never log in for example, but a better more effective form of PageRank sculpting is choosing for example which things to link to from your homepage.
So imagine you have got two different pages.
You have got one product that earns you a lot of money every time someone buys, and you’ve got another product where you make… you know 10 cents.
You probably want to highlight this page. You want to make sure it gets enough PageRank that it can rank well.
So this is more likely to be a page that you want to link to from your home page.
So when people talk about PageRank sculpting, they tend to think nofollow and all that sort of stuff, but in some sense the ways that you choose to create your site, your site architecture, and how you link between your pages is a type of PageRank sculpting.
So it is certainly not unethical to have all the links come into your site, and you decide how to link within your site, and how to make the pages within your site.
Erm, I do think that having more links because you have great content is a better way to rank well because it is a second order effect to be sculpting your PageRank.
It can be useful, but it wouldn’t be the first thing that I would do.
Commentary on Matt’s video I will leave to my good mate Dave
PageRank Sculpting; its all old school baby.
Response From Google After SMX About PageRank Sculpting
None…. yet – regard this as a placeholder
I do have some thoughts though:-
- I think we need a strong statement that external links with nofollow would not cause PageRank to evaporate.
- Nofollow is a simple solution for user generated content and comments, but if it has any effect of PageRank disappearing, we are going to lose the links on tons of blogs totally.
It would be a sad day that an action by Google reduced the interlinking of the web. - I don’t want to encourage use of javascript for PageRank sculpting – it is not really very good for accessibility
- Noscript – Nested embedded object items, containing links or thumbnails to source that may well be descriptive of content? This is needed if RSS Readers and web based email clients are going to continue to strip out video embeds.
- Links that lead to pages blocked with robots.txt and other hanging pages really need to be nofollowed. I think we need to know that in that situation PageRank wouldn’t normally evaporate, but I can understand why that might not be confirmed.
- I would love a much clearer indication of page size that Google will index as there are just vague notions that it can be more than 100 links per page.
If a size is specified, is that gzipped?
Matt Cutts On PageRank Sculpting
Matt Cutts today (June 16th 2009) wrote a post confirming that Google now treats PageRank significantly differently than the original PageRank patent, and that links with nofollow, whilst they don’t pass PageRank to the linked page, also can reduce the amount of PageRank that flows to other links on a page.
Rank Fishkin has already responded with analysis:
Google Says: Yes, You Can Still Sculpt PageRank. No You Can’t Do It With Nofollow
I also missed this commentary from Matt Leonard on why this could potentially make life harder for niche sites.
There are bound to be more posts appearing on Techmeme today.
PageRank Sculpting Isn’t Dead – It Has Evolved
Lets take a look at my WordPress SEO Masterclass.
Those Red links in the Sandcastle structure are not nofollow, they are oneway linkage.
It can be achieved with fairly simple coding, I even posted part of it over 2 years ago though the code needs to be updated for WordPress tagging rather than UTW.
This linking structure still works extremely effectively, but with one major caveat – internal & external links on the tag pages being used to channel juice back to the home page can’t be nofollowed.
If you are using default WordPress “ugly excerpts” they don’t contain any HTML content, no links to worry about other than the links to the posts.
Tag pages should thus be restuctured to highlight your best content, otherwise you end up with 3rd level push. 3rd level push in most cases isn’t a bad thing, if you don’t have a lot of comment links.
Rather than remove links that you previously nofollowed, the key is to add additional internal links to useful pages.
There are ways to handle the comment links, retain the benefit of having the comment content on your blog, and even keep giving your visitors a little link equity (dofollow links), though that solution will require significant programming effort.
Blogspot bloggers are now totally messed up, as even adding nofollow to their tag links isn’t going to retain juice.
Those who based their internal linking on my advice are not significantly affected by this change, and as this actually happened over a year ago, it is one of the reasons they have benefited.
The new PageRank sculpting could be looked on as advance information architecture, which was always the advanced PageRank sculpting
Expect a new WordPress SEO Masterclass soon, but it is unlikely to be free, and I would avoid following the advice of anyone who suggests conning your community using Iframes and Javascript for comments.
Update: Additional coverage worth a read @ Search Engine Land & Search Engine Watch
RELATED POSTS
The post PageRank Sculpting Isn’t Dead But Comments Can Kill Your PageRank appeared first on Andy Beard.