After my post last night and the many follow ups from around the SEO sphere, I (along with most observers from the world of search) strongly anticipated clarifying statements from Google's representatives at the SMX Advanced conference. Unfortunately, there's very little to report. The best I have comes from secondary sources, albeit relatively trustworthy ones:
- A large number of people queried the head of Google's web spam team, Matt Cutts, about the issue. Responses ranged from "no comment" to "the PR team would like me to stay away from that."
- Another member of Google's team noted that PageRank sculpting on unimportant pages like "register," "login," "privacy policy," etc. is OK and shouldn't hurt your site.
- A great many webmasters and SEOs are still shocked by the announcement because as recently as the last few days, they've still been seeing positive effects when employing the tactic.
The official line from yesterday that PageRank that was thought to be "conserved" by nofollowing unimportant internal links now "evaporates" rather than flowing to the remaining live links is among the most talked-about and, in my view, confusing subjects the SEO world has faced in a long time. It's very unlike Google to publicly message something in such an offhand fashion without accepting and answering questions on the subject more deeply.
This issue is compounded by many SEOs here who entirely disbelieve the messaging and think it was either a slip-up or purposeful misdirection. The evidence most frequently cited is the video Matt Cutts released only a few days prior to SMX Advanced indicating that while PR sculpting may not be the best activity, it's certainly OK to use it and you won't "evaporate" PageRank by placing a nofollowed link on your page. This video was well covered in this Huomah blog post:
Right away he talks about it being your right to do as you please with your website, including controlling “how the PageRank flows around within your site”. For those that are still unconvinced that PR truly does flow around a site, it should clear that up for ya.
He then mentions that it “is not the first thing that I would work on”. That he would work on “getting more links” and developing “higher quality content”. That is interesting on a few levels;
- While internal/external linking is important for your on-page SEO, one has to consider the ROI for any given SEO activity; no budget is infinite. All things NOT being equal, working on more links might be a better use of staff time.
- Part of any modern link building/SEO program is content (creation, syndication, placement) and one should be putting these aspects fairly high in the activity pecking order.
We read about all the finer details of the SEO process all the time in this industry – but rarely do we hear about resource management. One has to think about SEO in terms of cost efficacy of an activity as there are budgetary considerations to be had.
This is excellent advice, and consistent with Google's previous messaging around link sculpting - it's not the highest ROI activity for SEO, but you should use it as you see fit and where you find benefit. The sudden change during yesterday's session has certainly caused some to question the credibility of the statement. As I noted in yesterday's post on this topic, a modification of this scale (affecting nearly 3% of all links on the web) should have had a massive impact that webmasters worldwide could feel in their site's indexation and rankings, as well as see in the toolbar PageRank update from last week. As neither of those have been reported, the new statement's credibility is in even more doubt.
To paraphrase an SEO I respect a great deal (but whose permission for attribution I didn't get):
Google is frustrated with PR sculpting and they're seeing too much of it. Thus, they're using this messaging to try to stop people from using it as much. If you're a good SEO and you know what you're doing and you run the tests, I bet it still works fine. It's just that Google wants non-advanced webmasters to stop screwing around with their link graph and stay away from excessively using nofollow in ways that hurt indexing and relevance.
I find this to be an exceptionally good insight and the argument I personally subscribe to. Just as Google's messaging about dynamic URL rewriting was overly cautious and their advice about paid links is a bit overzealous, so too is this message about link sculpting. These messages don't mean we can't rewrite dynamic URLs to be static, don't mean that we can't engage in link building that has some commercial crossover (so long as it's not direct link buying) and, most recently, don't mean that we can't sculpt with nofollow. They're simply warnings to be cautious and be aware that Google is watching these issues and worries about them - fair enough.
The last point I'll make is that, like Danny Sullivan, I worry when Google changes messaging around SEO activities and best practices. To, in his words, "lose backwards compatibility" is extremely frustrating for site owners, SEOs and everyday webmasters who only occasionally dip their toes into the SEO field. It's very simple for those SEOs who want to sculpt with nofollow to switch over to something like iFrames blocked by robots.txt (or cookie-based links or Flash or external Javascript calls, etc.) to "hide" links from Google that they show to visitors and receive the same benefits. It's also frustrating that, if true, this now means one can sabotage a competitor's SEO by adding many nofollowed links in comments or other UGC areas (by "evaporating" percentages of the PageRank that will flow).
Let's hope that clearer messages on this issue emerge soon and that they resonate with the hundreds of tests many SEOs are surely performing on this subject as I type. Inconsistency builds distrust and all of us want Google's messages to be trustworthy, even if they slip up from time to time (after all, who among us hasn't?).
Disclaimer: This is not a valid test. I'm not claiming that this proves anything. It's just something interesting I noticed recently.
I have a very small site. 3 pages total:
I started playing around with nofollowing about 65% of those outbound portfolio links thinking that it would feed more juice to the clients who are actively working on their SEO.
One week later, I dropped from my usual #3 spot for my primary search term down to #7. I haven't been at #7 in over 2 years.
I took the nofollows off all my links, and one week later, back to #3.
Could very well just be normal ranking fluctuation, but I'm going to try putting the nofollows back on to see if it has the same effect.
it can take 2-3 weeks to see the positive result.
When I used the nofollow link attribute my site went from page 3 to page 4/5.
About 2-3 weeks later it was on page 2. I had the same result with lots of different websites.
NEVER try using any new techniques on money making sites or client sites. Any new techniques should be used on beta sites, and you should wait a few weeks to see the final result...not 1 week.
I tried this on internal use as I already had the majority of external links on nofollow
Thumbs up to enourage you to share those results with us in new YOUmoz blog post, not here in comments.
Respectfully, I think the data is a little too weak to warrant a YouMoz post. It's total speculation. My "test" is total bunk. Not controlled enough. It's really difficult to be able to test this to the point where I could be confident enough to share a conclusion.
I'll try though. If I can turn it on and off like a tap at least 3 times, and I notice the exact same effect each time. Then maybe it'll be worth a post.
Going to add the nofollows now...
Hey there Whitespark
Surely as these were outbound links adding nofollow should not have a negative effect to the site they are sitting on.
Under the old understanding of how nofollow works we could have expected it to increase site ranking, by increase PR flow around the internal links on the site.
Under the new (possible) understanding of how nofollow works we would expect it to have no impact at all.
Maybe I don't understand this subject as well as I thought (hopefully not) but thought I'd stick my neck out and mention this.
I would suggest something else may have caused this ranking change...
Well adders, what you say makes perfect sense, and, I have been testing my site with the nofollows on for the past few weeks, and it seems to have no effect. So, yes, agreed, no effect at all.
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.
You make some good points Andy, that will also stop PR from going on blogs with lots of commenters.
But the problem comes now is - will this be applied across the web or will it be a filter that trigger happy nofollowers can trip?
I just wonder if this had anything to do with the sudden drop of Wikipedia from the short tail keywords. For example, keyword "real estate" with 38700+ links from external site went down from #3-6 to 3rd or 4th page (tons of examples like this). Again, I have no analysis but I had always wondered what happened in the last few months to get Wikipedia drop so much.
I don't know which are your sources to know about this drop, but i have checked alexa.com
https://www.alexa.com/siteinfo/wikipedia.org
Set time span to "Max".
Look at the tab "Reach" - I can't see significant drop.
Look at the tab "Search %" (The percentage of visits to wikipedia.org that came from a search engine)Yesterday   65.3%7 day avg   65.5%1 month avg   65.6%3 month avg   66.6%3 month change   0%Again, nothing significant.
I think folks need to not expect any type of formal comment until probably toward the end of next week, at the earliest. That's typically how it goes with Google when things come up at conferences. People actually have to travel back home, there's required dig-out from being gone, plus you're just tired.
I'd especially urge everyone to relax. I'd rather have Google take the time to provide some fuller explanation to questions that knee-jerk quickly put something out that might result in confusion.
I'd also stress that there's no need for anyone to do anything about this. Some think Google's not really changed anything. If not, you're good. Some thing it has. If so, then "desculpting" isn't going to gain you anything, any more than "de-metatagging" old meta keywords tags you have on your pages are going to.
Folks should relax and wait for more info to come. And great party last night at The Garage last night, Rand. Though next time, I want a game of pool with you :)
Thanks Danny - would love to play pool with you as well sometime :-)
I do agree that taking action before there's clarification seems unwise. We've been telling our clients to hold off until there's A) better messaging and B) results from active testing.
Good Advice.
I'm going for a ride. It's 90 plus outside. If my wife want's to know where I'm at tell her Danny said I should relax.
Thanks
"Responses ranged from "no comment" to "the PR team would like me to stay away from that."
No, the "PR team would like me to stay away from that" was regarding a couple of specific interviews, not talking about this topic.
I'm talking with a few crawl/indexing/PageRank folks before I work on the blog post; given the interest in this subject, I want to check on all the details before I blog in depth about this. My hope is that I'll publish the blog post in the next week or so.
P.S. Every time I comment over here, I want the comment window to be draggable in Chrome. Any guesses if that's coming down the road? :)
Can we make a deal? I'll show you how to drag a window in Chrome if you show me how to rank (1st-2nd position) for my top 10 keywords.
Just let me know if you're ready for the deal :)
Let's hope we get some definitive clarity here. @MattCutts come thru for the win buddy! Can we expect video on that post as well? It could reduce your writing time.
i also would like the dragging fixed
use this option when you make the dragable and add a class to be the drag handle
 from scriptaculous
(option)handle:
string or DOM reference, not set by default. value may be a string referencing a CSS class value. The first child/grandchild/etc. element found within the element that has this CSS class value will be used as the handle.
 pretty sure should fix your ie / chrome drag issue
Is this nofollow madness a hoax? No.Â
 If you are using nofollows on links that target pages that YOU CREATED, then you are waiving a giant flag announcing that you are an SEO'er. We individually have to deal with the addd exposure that we have brought on ourselves.
Nofollows still work remarkably well. Tested and retested. Take my word for it -- or panic and pull down all of your nofollows to help improve MY rankings. Your call.
If you want to continue to use nofollows then you're just going to have to clean up your shady stuff. Consider nofollow page sculpting just like applying Google Analytics to your shady stuff. If you don't want Google looking under your hood, then don't toss them the keys to your 1965 convertable affiliate spam machine.
I completely 100% agree with you here SearchGeeksOrg! It is manipulative and shady indeed. You can't expect Google to be okay with some SEO telling them and potential visitors what content is most important to them.
 I don't use PR Sculpting. Never have, probably never will. I also RARELY use nofollow. My method is generating great content that people want to link to, which has a much more positive effect on important pages than PR Sculpting. If you have to use PR Sculpting to indicate the importance of your pages, you probably don't have important pages. That's just my opinion.Â
My important pages rank high. Solidly. My pages that I don't care about like TOS, login pages, etc, don't rank unless the query is unbelievably specific and relevant.
So, supposing I had a blog post with high PR... The more (nofollowed) coments it receives, the worse it is for the blog?
Perhaps I'm missing something, but this seems awfully counterintuitive.
Nofollows are meant to be used on blog comments and forum links. You aren't placing the link, you don't know the target, and you don't want to "vote" for the external website. No problem there.
For this conversation, consider your internal page sculpting. Only an SEO'er would nofollow a link to a page that he/she created. There may not be a penalty for internal nofollows today, but you are certainly increasing your exposure for review if you choose to use them.Â
What if internal links, which are simple redirection to another sites and which are blocked by robots.txt, are nofollowed? Exposure increased? Review? Penalty? PR evaporation?
Hopefully nothing of them, but I'd appreciate to hear another voices.
I believe I just made the subscription to this theory as well.
 But just to argue with you a bit, since we seem to be questioning you these days...
Don't you think that it is a probably likelihood that this is becoming a soft-rule? That, in fact, Google is capable of observing the tendencies of these sites the employ nofollow PageRank sculpting, and could therefore make a decision based on the statistical data to somewhat void the nofollow sculpting on the sites of those who seemingly abuse it?
Does Nick have any ideas on how Google could seemingly "pass judgement" on a site that employs too much PageRank sculpting? Would it be something that is easily detectable via crawling?
Definitely possible, but I'd think the messaging would be different, e.g. "we are looking more closely at sites that overuse PR sculpting and may take action to ignore some of that or even restrict the juice that flows when it appears to be manipulative or poorly applied."
A message like that would make me (and, I think, all of us) much happier than just saying that PR now "evaporates" through nofollow links. The implications of that are just too far-reaching and potentially sabotage-inducing from my perspective.
I also belive more in that line than in a radical change from Google around the nofollow tag; I think would make more sense trying to fight the overuse of the tag.
For that, would be reasonable the possibilty of Google starting counting the number of nofollow links per page, as I have read in some comments.
Wouldn't it be better to have a new set of guidelines regarding the nofollow. Harvesting fear on people working in this direction generates distrust, frustration and more important always finding new ways of getting the most of PageRank and not always in the best ways....
Why isn't everybody talking about Google finally being able to read onsite javascript...I know the impact on the nofollow matter is much more important but so is this for programmers and SEOs.
What is Matt Cutts suggesting, we should now go to our sites, remove all the sculpting work and update every single page were we had it. Wooow then a new Google Dance will happen for sure.
Im sorry for the people who doesn't have any login, register or privacy policy pages....mmmm
I agree that the idea of Google allowing the PR to "evaporate" is slightly far fetched. I'm just thinking that even Google is trying some different things to deal with the implications of misused PR sculpting.
Hopefully we will see some form of a response in the near future. Until then I might just have to go ahead and test the affects of PR sculpting on some of my own sites.
Maybe others will do so as well and report their findings.
Maybe the takeaway of this has nothing to do with NoFollow at all. Maybe it's a sign that the number of links on a page no longer matters very much. If each link is taken as its own entity without being affected by the other links on the page then his statement that NoFollowing some links doesn't help the others, makes a lot more sense.
Just a little food for thought...
- Evan
I agree that this is a cloud screen. I don’t think for one moment if PR sculpturing is done correctly for non important pages, Google would somehow take that as bad.
On the other side of things I can understand why Google made the statement; many sites use quite underhand use of Nofollow to pass PageRank to their top level pages.
I certainly won’t be changing what I do on this. Non important pages have use for PageRank.Â
Hi Rand & everybody, you'll be glad to know we've had a first unofficial confirmation today (coming from a member of the Quality Search Team at Google Italy) of what Matt Cutts said at SMX Advanced about how Google handles nofollow: https://www.everfluxx.com/google-confirms-nofollow-wont-help-flow-more-pr-to-your-other-links/
Just revisted this post to see if I could get further clarafication on this.
Thanks for the comment Everflux, that makes things pretty clear.
I guess I'll wait just a little longer to get this from a few more sources but if they back this up I will be using nofollow a lot more cautiously in future.
I was just thinking, what's the craziest thing Google could do (maybe) in the next update. Well, looks like they did one better... that move="nofollow" - Joe Perez
Great insight. I never tire of your thorough analysis Rand...I'm still scratching my head on this one though.
Hey Rand... gang... how’s it going? (tried to add pic in profile, wouldn't take)
Anyway, thanks on that Rand, it was kind of interesting to be revisiting that topic before the latest drama blew into SEO Town. We were kinda light on drama the last while - at least we have this eh?
I'd have to go with the 'wait and see' approach at the moment. Some more concrete statements might make more sense of it all, at this point I am still trying to sort it out. I am just trying to make sense of it algorithmically; something just isn't adding up. Would I go as far as to don tin foil accessories about inverted FUD? Prolly not...
I have always been concerned about how the nofollow was impacting the link graph, even at 3%... but 'evaporating PageRank'? I am curious as to what constitutes the rules (conditions) for such an algorithmic approach. It reminds me of when they 'algorithmically' stopped Google Bowling, which it didn't... One has to wonder about this round. I really think some clarification would help.
Speaking of that 3% - what percentage of those are using the NF tag aggressively? I’d imagine a lot of those are blogs with NF plugins and that those really being concerted and aggressive with the sculpting would be a fairly small number. This all begs the question of why they’d go to the trouble of implementing such an adjustment. There is as much chance of damaging the link graph as actually improving usability of the engine. There is just something troubling with the whole ‘evaporation’ concept IMHO.
Once more, I have never really leaned very hard on over use of the NF tag anyway. Understanding internal linking and prominence has been around and evolving for a long time and this is but another evolution in the chain. It should be interesting to see how it plays out.
Ciao... have a great weekÂ
Great Post rand summing up the SMX sentiment. I share your views. Thanks for being a great host and throwing a terrific party!
I like your last point the best, and wonder how Google answers the question of percentage drain due to nofollow on comments. It's become common practice for blogs and pages that support commenting to nofollow any links posted there. According to this new model an industrious and competitive SEO could go to a higher ranking site that allows comments, have useful comments with relevant links added to the page, and it would still penalize the higher ranking site. What sort of sense does that make?
Maybe this is a move in support of one-way-conversation brands.
Interesting how Google want themselves to be perceived with their messaging strategy...
I'd have to agree with the comments made earlier...
* Experienced SEOs generally know through benchmarking and testing what affect this may have on their websites...
And now...
What is going to happeng with wikipedia (it's not anymore a black hole seo)
Facebook, twitter?they will drop?
Should I take out the social bookmarks on my site? now they are outbound links.
Google is pushing us to make dangling pages
PR Sculpting is one of the coolest strategy in SEO. Com on Google, do not remove this from me !!!
This unfolding drama has been like a soap opera to me. Since starting SEO, this is the biggest controversy.
Very interesting stuff afoot. I love the commentary.
Why does anyone even listen to Google's FUD Czar?
It still works, ive no idea what he (matt cutts) is talking about - it still works I've tested it, and the PR didn't 'evaporate'. He said its been in place for over a year. MC seems to send people on wild goose chases.
This would just be an absolutly massive undertaking and the right wing approach of messaging that matt has taken here of "shock and awe" seems a slight bit too transparent.
 I am sure google will now be doing some analysis over the coming weeks, months, etc to measure the kind of impact matts statements have made and too whom... ( which site simply knocked all the no follows off site wide for example )
 I really feel google is missing the boat here not to mention trying to mislead an entire industry that has more than a few dark hats sitting in the closet.  As they (Goggle) states over and over most webmasters, etc have better more productive things to do than funnel pr, yet they drop an a-bomb @ SMX on the very topic.....
 I am interested in watching some of my friends in germany for the next few months with the moves and results they see.
 But for Mr. Cutts all I can say is you have brought me back to at least dusting off half of that black hat of mine....