I’ve long considered Google Pagerank, and the, some great, PageRank Sculpting discussions around the net, to be akin to an idea of wealth and cashflow – i.e. should you save what little money you have, cut out the unnecessary expenditure and spread it about to make ends meet, or do you go out and get yourself a better job with more cash? – DaveN touched on the subject of nofollow scultping with Matt Cutts recently and Matt offered up a similar analogy.
Nofollowing your internals can affect your ranking in Google, but it’s a 2nd order effect. My analogy is: suppose you’ve got$100. Would you rather work on getting $300, or would you spend your time planning how to spend your $100 more wisely. Spending the $100 more wisely is a matter of good site architecture (and nofollowing/sculpting PageRank if you want). But most people would benefit more from looking at how to get to the $300 level.
Should you nofollow unimportant internal pages or nofollow external links in an effort to consolidate the Pagerank you have already accrued?
Or should you spend your time getting other quality links pointing to your site to increase the PR you have to start off with (how you get Pagerank).
The long term best impact strategy here is simply to earn more, or you’ll find it a slow rise above the core issues of your current predicament, whatever that may be, and I think the same can be said of the question of maximising page strength by PR sculpting.
In truth you need to do both, maximise what page strength you have by whatever method you use to manipulate PR and on-site relevance, and linkbuild to add conviction to your attempt at making a particular page relevant and give it a shot at those first page rankings.
Page Rank Sculpting Discussions
Joost de Valk has a terrific article on PR Sculpting, as does Dan Thies on using nofollow to sculpt pagerank, and the Mad Hat pitches in on the FUD of Nofollow being a red-flag if you’re trying to maximise the visibility of page in Google. Michael Martinez has an interesting take too.
I’ve used nofollow on internal links to sculpt and concentrate internal PR and from what I’ve seen the results *might* be promising, though very minimal, and not a long term substitute for an intelligent site architecture to begin with and certainly no seo magic bullet, although you have to be careful.
I should point out I never use rel=”nofollow” to prevent the indexing of a page – merely to control which pages any particular page shares it’s link equity with, if you are Googlebot anyway.
It *appears* that the first link you nofollow on a page *might* also nofollow any other link to the same url on that page, although nofollowing the home page link high up in code (when you have another link to the home lower on the page) seems to be treated differently by Google, Yahoo and MSN. Wonder if a ‘Contact’ page is too?
Optimise Your Contact Page, Don’t Nofollow It
As I have said, I’ve been playing about with rel=”nofollow” on this site for 4 months, and in all honesty, in future, I won’t be relying on nofollow to sculpt unimportant pages out of any possible link graph, just optimising those pages better, or leaving them out altogether, like I used to do in 1999.
It can be a useful tool in a site redevelopment, but from here on in, I’ll be keeping nofollow for bad neighbourhoods and, pending further testing, on top level blog pages, using Andy Beard’s Nofollow Dupes although this site is still a linky love / dofollow blog (for regular contributers at any rate).
Nice round-up. A good example of nofollow use could be adding the attribute to all those social network buttons on your blog, but no one has really any proof that is makes a big difference in doing this. Their in lies the whole perplexity of nofollow.
Thanks for the comment, Jaan. It’s a terrifically confusing subject, virtually impossible to test with any accuracy in the real world, and it’s obvious to me benefits are minimal at best, and probably better addressed through an intelligent structure to begin with.
“probably better addressed through an intelligent structure to begin with.” EXACTLY. You know we (well some of us, not me) SEO’s are easily swayed by new concepts in SEO. We want to be the first to use them and show that they work. Especially if the concept is “somewhat” touted by Google (Matt Cutt’s) themselves. In reality testing really is the only thing that can show real results in SEO and in this case this concept or technique is impossible to test its effectiveness.
True. I liked the idea of nofollow, but it just doesn’t quite all ‘fit’, as far as I can see. I don’t think we’re being spoon fed all the info we need to know to take advantage of it’s benefits. I see pages nofollowed still apparently in the link graph, still ranking no1 for (albeit uncompetitive) terms, 4 months after I applied rel=”nofollow”. I’ve decided I want every page on my site to rank for something, long tail or not. That will include my contact page. If I don’t need a page enough to nofollow it, I’m now questioning myself as whether or not there’s an argument for just leaving it out all together.
“I see pages nofollowed still apparently in the link graph, still ranking no1 for (albeit uncompetitive) terms, 4 months after I applied rel=â€nofollowâ€.” That is because just nofollowing a link to a page in your website won’t stop it from getting indexed and ranked. I don’t think Google ever said that would happen. The reason for this is simple, other links without nofollow could be pointing to the page and other links in the past could have pointed to the page.
I am aware of this, but through linking, I’ve purposely aimed to “nofollow-orphan” pages to see if they ranked, based on some others’ analysis and my own, and they still do, if they were indexed before you applied nofollow. I always thought orphan pages was a surefire way to tank in Google, but it doesn’t appear to be in this instance.
Hi, When i build a new niche page i usuaully nofollow the “contact us” and “terms” pages directly from start. What im now starting to wonder is if thats a good ideas? I have heard Google will check if you have those pages if its a small page and if you dont you might get lower ratings. If i have nofolloed them.. does it count as i dont have them maybe? Thanks for a great article. BR, Johan
I’m thinking of perhaps optimising my contact page, and perhaps having terms etc only accessible from the contacts page. :)
Jaan, One of the points of nofollow is that it removes a lot of SEO constraints from the design of your user navigation. This means that you can do SEO with less impact on usability. The argument that you can’t directly test the impact could be applied to almost anything we do in SEO. There are *always* hidden factors that can’t be measured.
“One of the points of nofollow is that it removes a lot of SEO constraints from the design of your user navigation. This means that you can do SEO with less impact on usability.” Thanks for the comment Dan. This is what I liked about nofollow in the beginning, and why I think it can be useful when restructuring an existing site. If I’m building a site from the ground up, I’ll be doing it without nofollow, and eventually always attempt to phase it out over time in redesigns.
“The argument that you can’t directly test the impact could be applied to almost anything we do in SEO. ” Not really. I see results from SEO changes I do all of the time. Better titles, better content, better internal navigation, better syndication of content which lead to links to my website, etc. I used nofollow in a few test situations and I didnt see much improvements at all. Doing the above changes did. Either way we are beating a dead horse here, we know how both of us feel about it.
How do you know that the title changes didn’t impact user behavior that’s being measured in some way? Having seen sites go from a few hundred to a couple thousand pages indexed within a few weeks of a nofollow restructure, and never seeing it go the other way, I believe that it’s working. But I can’t prove it – we can’t prove anything. And all these people claiming big traffic gains make it hard to discuss, because that’s a huge stretch absent any analysis of the cause. But yeah, the horse is dead.