Matt Cutts and I have a beautiful relationship - I relentlessly pester him and he gives me answers to questions that are pestering webmasters. It's a cross between symbiotic and vampirical (vampiristic?). Thankfully, Matt has once again let me suck away some of his precious time to address some big issues. After this, I think I'll let him have a rest and probably go after poor Tim (his website is so lonely).
I've posted a list of six questions I asked Matt, 3 of which he answered (answers, or lack of answers, are in bold). This is followed by some discussion on these and other topics that warrant a mention, and may be serious news to many folks (they certainly were to me).
- Why was John Chow's website (johnchow.com) penalized in the Google search results?
- A) John engaged in manipulative link acquisition (buying links, affiliate linking, etc)
- B) John's site consistently provided links to paying companies without using nofollow
- C) John's site engaged in other manipulative practices on-page – cloaking, keyword stuffing, etc.
- D) He said mean things about Google and we're soft-shelled here at the 'plex
- E) No answer provide by Matt
- Does Google recommend the use of nofollow internally as a positive method for controlling the flow of internal link love?
- A) Yes – webmasters can feel free to use nofollow internally to help tell Googlebot which pages they want to receive link juice from other pages
_
(Matt's precise words were: The nofollow attribute is just a mechanism that gives webmasters the ability to modify PageRank flow at link-level granularity. Plenty of other mechanisms would also work (e.g. a link through a page that is robot.txt'ed out), but nofollow on individual links is simpler for some folks to use. There's no stigma to using nofollow, even on your own internal links; for Google, nofollow'ed links are dropped out of our link graph; we don't even use such links for discovery. By the way, the nofollow meta tag does that same thing, but at a page level.)
_ - B) Sometimes – we don't generally encourage this behavior, but if you're linking to user-generated content pages on your site who's content you may not trust, nofollow is a way to tell us that
- C) No – nofollow is intended to say "I don't editorially vouch for the source of this link." If you're placing un-trustworthy content on your site, that can hurt you whether you use nofollow to link to those pages or not.
- A) Yes – webmasters can feel free to use nofollow internally to help tell Googlebot which pages they want to receive link juice from other pages
- If I have a website in Spanish that serves Spanish speakers in Mexico, Spain, Latin America & the United States, how does Google recommend that I tell them my content is relevant to searches in all of those countries?
- A) Host a single site in any country you want, but use a general domain extension like .com, .net, .org, etc. Then, acquire links from sites in all of those markets and Google will realize that your content is intended for an international, Spanish-speaking audience
- B) Build a website for each country, but host them anywhere so long as they are registered with the proper top level domain. Feel free to use the same content on each, as duplicate content filters only apply to a single country.
- C) Build multiple websites for each country, host them individually in each country, register them on different top level domains and create unique content for each (even though the content might technically be exactly the same). Then, link build and market each site in each individual country separately. Don't interlink excessively between these sites.
- D) Use a single site for now and target your largest potential market. Then, give us some time to work on this issue – we'll have greater functionality for single sites targeting multiple countries/languages in the near future.
- E) No answer provide by Matt
- Google has noted in the past that a maximum of 100 links per page was wise and would insure that all of the links on a page would be crawled. Does this rule still apply or is there some flexibility?
- A) The rule is still a good one – even very important pages with lots of PageRank should stay away from linking to more than 100 other pages.
- B) There's some flexibility. If you have a high PageRank page with lots of link juice, we may spider well beyond 100 links per page – possibly even 2-300 depending on how valuable we feel that page to be.
- C) The rule really only applies to pages of low importance/PageRank. Googlebot now regularly can crawl 150-200 links per page without breaking a sweat and those numbers can be even higher for pages we consider particularly important.
- D) Although we may crawl more than 100 links per page (maybe even many hundreds), we don't recommend linking to that many because of the dilution of link juice that occurs. Instead, use sub-navigation pages to help ease the link per page burden.
- E) B & D
_
Matt's exact words - The "keep the number of links to under 100" is in the technical guideline section, not the quality guidelines section. That means we're not going to remove a page if you have 101 or 102 links on the page. Think of this more as a rule of thumb. Originally, Google only indexed the first 100 kilobytes or so of web documents, so keeping the number of links under 100 was a good way to ensure that all those links would be seen by Google. These days I believe we index deeper within documents, so that's less of an issue. But it is true that if users see 250 or 300 links on a page, that page is probably not as useful for them, so it's a good idea to break a large list of links down (e.g. by category, topic, alphabetically, or chronologically) into multiple pages so that your links don't overwhelm regular users.
- What is Google's position on the value of generic web directories that market to webmasters as a way to boost link strength, PageRank and relevance? Would you largely agree or disagree with my assertions about the value of generic web directories on the subject?
- A) While Google certainly feels that many directories are valuable, those that are built with SEO purposes in mind are generally viewed as somewhat manipulative and we make an effort to see that their value is limited. We're generally in agreement with your post.
- B) Google doesn't treat SEO directory sites any differently than any other sites – if their PageRank is high and their content is relevant, we pass just as much link value and anchor text weight as any other similar link. So, we differ a bit in opinion with your post.
- C) Mostly A with a little B
_
Matt's exact words - We tend to look more at the quality of a directory than whether it is SEO-related. Of course, plenty of directories that are targeted only for SEO don't add that much value in my experience. I talked about some of the factors that we use to assess the quality of a directory in an earlier post at https://www.mattcutts.com/blog/how-to-report-paid-links/
_
"Q: Hey, as long as we're talking about directories, can you talk about the role of directories, some of whom charge for a reviewer to evaluate them?
A: I'll try to give a few rules of thumb to think about when looking at a directory. When considering submitting to a directory, I'd ask questions like:
- Does the directory reject urls? If every url passes a review, the directory gets closer to just a list of links or a free-for-all link site.
- What is the quality of urls in the directory? Suppose a site rejects 25% of submissions, but the urls that are accepted/listed are still quite low-quality or spammy. That doesn't speak well to the quality of
the directory.
- If there is a fee, what's the purpose of the fee? For a high-quality directory, the fee is primarily for the time/effort for someone to do a genuine evaluation of a url or site.
_
Those are a few factors I'd consider. If you put on your user hat and ask "Does this seem like a high-quality directory to me?" you can usually get a pretty good sense as well, or ask a few friends for their take on a particular directory."
- Are pages with "noarchive" treated any differently than those without the tag?
- A) Yes, Google may not treat links, content or other factors on a page the same if it has the "noarchive" tag.
- B) No, these pages are treated the same as a page that permits us to archive.
- C) No answer provided by Matt
BTW - For the remaining unaswered questions, Matt responded:
The rest deserve longer answers and I'm too swamped to do a full reply, or the question is not how I'd pose it, so multiple answers (or none) could apply or the answer is more nuanced.
So, I guess I need to work on my question-posing abilities.
Here are the big takeaway points from my perspective:
-
Nofollow is now, officially, a "tool" that power users and webmasters should be employing on their sites as a way to control the flow of link juice and point it in the very best directions. Good architectural SEO has always had some internal link structuring work involved, but nofollow and Matt's position on it makes it clear that for those of us who are professionals, we can be use it intelligently without a downside risk.
-
For pages with many, many links, sticking close to 100 links per page is probably still a very good idea, though high PR and link juice pages can certainly get more pages spidered. I note that on a page like the Web 2.0 Awards, well over 200 links are being followed and passing link juice (at least from what I can see).
-
Directories (and all websites) that link out need to be very, very careful of who they link to, as this is a big way that Google's algorithmically identifying and discounting paid and other types of links they don't want to count. For really solid evidence on this, check out SEOmoz's own recommended list - the page is indexed at Google, but can't rank for its full title tag, we're not even ranking for the phrase in quotes. Why? I'm almost sure it's because of someone on the list that we're linking out to that Google doesn't like. I don't think that's a hand penalty - I think it's algorithmic. Full disclosure on this - SEOmoz used to be a directory operator, with a junky piece of crap called SOCEngine. We've officially shut it down at this point, even though it's been barely active for the last 2 years, as it would be totally hypocritical to operate a low quality directory and then proclaim that they're not worth paying for.
Matt also addressed a few other issues that deserve their own blog posts, and those should be coming soon.
You've got Whiteboard Friday's... maybe you could introduce a new service like "Ask Matt Monday's" where you ask Matt a new question each week.
Then when Matt doesn't answer we can all say "Sounds like somebody's got a case of the Monday's"... I'm sure he'd love that
Quick points:
- you might want to highlight the text that I sent back in bold in addition to the multiple choice answers; the text answers give more context than just the answers alone
- I think saying people "should be" using nofollow is a bit strong. More like people can use it for internal links if they're power-user-y enough to want to sculpt PageRank flow within their site at the link level. But I'd say that most regular webmasters don't need to worry about link-level PageRank flow within their site. I think saying "power users and webmasters should be employing on their sites" overstates it a little. It's available if you want to get into that much fine-grained control.
Thanks for the clarification, Matt. That's definitely not as uncomfortable a thought.
However, fine tuning internal PageRank still strikes me as something you'd do strictly for the search engine and not the user. With that in mind, I worry that Google or some other SE might decide at some point in the future that we shouldn't be doing it after all.
This isn't necessarily 100% true. I've used nofollow on internal link to block product "search" pages in the past (before it was a guideline) with great success. It's not only for the search engine, but for the user. I've scrapped through a lot of analytics data and I can safely say that those search pages on our site DO NOT provide a lot of value - the bounce rate is usually high, and the conversion is very low. I'd much rather have someone land on another page of the site.
While that outcome might be what you want for the users, Kwyjibo, you're still doing it purely to get the search engines to treat your site in a particular way.
I think the debate is already lost, however - we do loads of things to our sites now that aren't for users, but are for the engines only (robots.txt, meta robots etc.).
Good points, Matt - I'll go bold your entire responses. And thanks again, stuff like this really means a lot to all of us.
Hi, I just want to mention that most people do not seem to know about the newest Google webmaster guideline which states....
“Use robots.txt to prevent crawling of search results pages or other auto-generated pages that don’t add much value for users coming from search engines.”
So for those of you saying that it seems odd and unguideline like to "tell" a search engine what is important and what is not about your website, I would review that guideline.
Right, but I'm not suggesting that this is specifically against the guidelines. As you point out, it's recommended.
I just feel that it's against the over-arching concept of "make sites for people, not for search engines" and that as such, it's the sort of measure that I wouldn't count on the search engines supporting forever. Any practice can be abused, and if that happens, I see no reason they'd continue to recognize it.
So in my mind, the question becomes do I take advantage of this and make changes, even though I'm concerned that at some point in the future I may find that I have to go back and undo those changes? Or perhaps more realistically, what happens if the SEs stop recognizing nofollows on internal pages? Then I'll have to go back to designing navigation structure the old fashioned way.
It seems to me that the same way some people don't want to use XML sitemaps because they want to know if a given page is having indexing problems, I don't think I want to use nofollow for internal links.
"Build for people, not for search engine" days are way past over. You build for both Google and people. When those goals conflict, always choose usability over SEO, but nofollowing links does not impact user-experience a bit.
What Google tell you doesn't matter. I don't know why so many people who disregard Google's Guidelines always like to quote the Guidelines every chance they get.
Know how Google operates and learn how to take advantage of what you know. If you get hung up on what Google tells you sooner or later you won't be able to sleep because you bought a few links and didn't ask them to be nofollowed. It's your damn website - do whatever you want.
Matt,
Thanks for tht clarification... but I'm gonna need just a bit more....
Wasn't it just a few months ago that Adam Lasnik said in an interview in regards to Internal Page Rank Flow:
So most Webmasters should not be worring about internal pagerank flow?
Also, earlier this year in a Webmaster World post about Outbound Link PageRank leakage Adam Lasnik said:
So is Google now ok with "engaging in link schemes designed to manipulate your site's ranking." if it is only on your own site that you are engaging in these link schemes on?
Adam Lasnik said that as a response to people talking about nofollowing their outbound links after Wikipedia added nofollow to all their outbounds. To prevent newbies from obessing over nofollowing every link to "conserve" PageRank and minimize PageRank bleed, Adam said don't worry about PageRank flow. Get backlinks.
But this is like a NASCAR team telling a driver "don't worry about the car setup, just keep turning left."
Hey Natasha! What both Adam and I were saying is that if your site architecture is at all reasonable (e.g. a WordPress blog, or a typical hierarchical/tree layout), then internal PageRank flow is more of a second-order effect. So I'd worry about other things at a higher priority.
Hey Matt,
Thanks for the reply and clarification. I know I could not have been the only one who read this post and at first pass thought "but didn't someone from Google say something last year about not using NoFollow to manage Page Rank Flow". I just thought bringing it back up and having you clarify would help a lot of people understand where Google stands on NoFollow for pagrank flow management (ie: using nofollow to manage page ranks flow on internal vs. external links) Thanks again!
"The nofollow attribute is just a mechanism that gives webmasters the ability to modify PageRank flow at link-level granularity."
What hypocritical crap.
Google quality guidelines:
“Make pages for users, not for search engines.” / Link Schemes: "Links intended to manipulate PageRank"
New Google quality guidelines:
"Make pages for users, not for search engines, unless it benefits Google then you can design for search engines instead, but if you overuse it we'll change our minds again and ban you. For example, nofollow your items that aren't in season to push the flow of PR to your items that are hot now...that's okay. However, if your blue widget site suddenly is flowing all of it's PageRank to a "order Cialis" page then we'll be mad."
Link Schemes: Links intended to manipulate PageRank, like controlling the flow of PageRange on your site is now okay. Just don't try and manipulate PageRank in some other way. Google reserves the right to decide later that this is a sneaky blackhat scheme and ban you.
John Chow Clause: If someone else comes up with a sneaky idea like linking back to someone who reviews your site, we'll ban you regardless of what we say, even if none of those links were paid for. As long as someone reports them in the Google webmaster tools.
Rand Fishkin Clause: Rand is immune to the above rules as long as they continue to let us push propaganda to the masses... ;)
hahahaha. good one.
Very nicely put. :D
Google also reserves the right not not decide, but pretend that they did, drop a few hints that they decided one way, and then pretend later on that they actually had decided the other way all along, if they think that it might make them look better to do so.
Well said! We should completely rewrite the guidelines to what it "actually" says.
MadHatterKC,
I think that you are right on about rel=nofollow.
I think that if Google thinks about the implications a little bit longer that they will reverse their decision on rel=nofollow.
Robots.txt is for controlling robots, not rel=nofollow.
Using rel=nofollow to "sculpt PageRank" is not part of the rel=nofollow microformat standard and may have other unknown effects with other search engines.
Shouldn't Google and other search engines ignore rel=nofollowed internal links? A site would presumably always vouch for its own pages. Rel=nofollow was intended to prevent spam by saying "I do not vouch for this user-contributed external link".
Sites that rel=nofollow their internal links remind me of those over-SEO'd sites from a few years ago where every time a keyword appears on a Web page it is highlighted in bold or italic -- an obvious sign of search engine manipulation.
It also reminds me of Yahoo's recent attempt to create a new tag called "robots-nocontent". It's not standard, no one knows completely how it works or how the other search engines might decide to interpret it. What if Yahoo or MSN suddenly decide, "we don't trust sites that nofollow their internal links and when we flip this switch in our algorithm we end up weeding out a lot of over-SEO'd junk"?
hmmm... if user_agent==googlebot: add_nofollow() ?
Now now, that would be cloaking.
Only youtube, the new york times, and whoever else gets the Google gold stamp of "do whatever you want" is allowed to cloak ;-).
I'm still a bit surprised at the expansion of the use of nofollow. When it first came out, its purpose was very clear: bloggers should use it for UGC, like comments. It meant "I can't vouch for this link."
Now they want us to use it on ads, even if we can vouch for them, and they're even recommending we use it to control internal PR, which we used to control by crafting a logical navigation structure.
All I can say is "hmmmm".
Rand, nice Q&A - I like how you offered answers in between giving Matt's answers. Keep up pestering him please!
I have one follow up - When talking about the number of links on a page, does that include both internal and external?
I would imagine it includes both but it would be nice to have it verified from Matt.
Agreed.
I'm sure we all appreciate how stacked out Matt Cutts must be. But since he's the most visible Googler, his insights are often both telling and vital.
Additionally, if you do look at rather than simply read the guidance that comes out Google, it's rarely rocket science.
More often than not, it's common sense that follows much the same rules as the social interactions we take part in almost every day.
A topic which I'll be writing about shortly...
"Nofollow is now, officially, a "tool" that power users and webmasters should be employing on their sites as a way to control the flow of link juice and point it in the very best directions."
Can't disagree with you there.
I recently wrote about using nofollow to implement Third-Level Push to explain the concept to one of my clients.
I only have 4 words on this topic.
Matt Cutts Schmatt Cutts
We (webmasters) need some form where we can vote up 10 questions that Matt will take the time to answer, like the way slashdot conducts interviews.
I don't really have a good suggestion on how to achieve this but seomoz has the connection to get it done. Maybe you can just create a topic for this and we can thumb up what questions we really want answered.
Thanks for sending over those questions! Always valuable to get more info out of Matt to squash some rumors and myths.
Great questions, thanks as always for participating Matt, and I agree, turning this into some kind of a regular feature would be great.
The nofollow clarification totally blew me away, for two reasons... starting with the second, that "we don't even use such links for discovery," given that this seemed like a gray area before, that they may or may not be followed, but they just wouldn't pass any link juice.
I can't help but feel a little dirty... just having a hard time grasping how this isn't a form of "manipulating" to impact Google search. So you aren't suppossed to buy a link to drive PR to a page on your website, but "manipulating" the flow of that PR to boost pages on your site is okay?
Perhaps using this as a "tool" has been the direction this has been going for awhile, but I can't help but feel, maybe incorrectly, that this is itself, a bit of positioning-- response or fallout from San Jose?
I feel the same way, but FWIW, it appears that internal nofollow use to nudge a little PR away from here and toward there, like "approved cloaking," is a form of manipulation that's ok with the Big G.
Identity - the only thing I'd say to that is that, as Matt wisely pointed out, the ability to control PR flow has always been available to webmasters (using other types of links that Google doesn't/can't spider), so nofollow is jsut adding to the arsenal of tools, not making a new rule.
Rand,
I don't necessarily have an issue with doing this, just trying to get past what I feel is a redfining of rel=nofollow. And maybe it will just take more clearing of the old views.
This has been defined as a tool primarily focused to help combat blog spammers, then linking to any site, then as a tool to use for paid links. But all along, two things have always seemed to be present... that the sites/pages being linked to were ones that we weren't vouching for and that doing anything for the sole purpose of manipulating search results should be avoided.
Perhaps then it is the issue of ever saying that here is a page on my site that I don't vouch for... interestingly, a community-based section like YOUmoz might be an example of where this might want to be done, except SEOmoz obviously manually controls the publishing of content.
The bigger sticking point for me is exactly the point you brought up... in my mind, this is the distinction between using other linking formats, such as javascript links, to prevent crawling or access by engines... I see this as coming back to intent, that these methods were typically used to prevent crawling, not to influence PageRank, which again is something that Google has positioned as something to help their toobar users and shouldn't be something that webmasters or SEOs should be worrying about or trying to influence.
It may just take some time for this new viewpoint to sink in.
Rand puts it in a good way. There have always been links that the search engines couldn't follow so the use of a "no follow" tag just replicates that effect. Interpretation as to value, either on the webmaster side as a "vouch" or on the SE side as an "endorsement," has come about because of the associated value in ranking well.
Domainers are a good example of links not being a be-all-end-all of valuation on the net though. Another would be the PR changes companies have tried to make to their Wikipedia entries (Wired story). Personally, I think Cutts' argument of links being regulated in a governed matter isn't feasible on the structure of the Web.
Identity, read my comment above. Most people who have trouble wrapping their brains around nofollow should start with understanding the basics.
I can see some newbie SEOs arguing the "nofollow is manipulative but paid links aren't" line, but I don't see professional SEOs batting an eye at either buying links or nofollowing internal links, which, like I said, is technically a cousin of 301 redirects and robots.txt disallow.
I guess I should clarify, it really isn't a matter of understanding, I've just found it interesting at the shifting definition of the use, which in itself, isn't necessarily good or bad... so really just more observations than anything else.
Yeah, the "definition" has been something that's closely followed the money involved.
"the shifting definition of the use"
Identity, the definition of nofollow has never shifted. Matt has been saying since last year (as far as I can remember) that rel=nofollow technically behaves very similarly to robots.txt and META nofollow.
The SEO community has "politicized" nofollow to mean a whole bunch of things while ignoring its underlying mechanism.
SEO should be about making sites rank. Nofollow has been, since its inception, a tool you can use (especially on large-scale sites) to help you accomplish that. That should be the end of story.
Sure, its manipulative, but so are keywords in TITLE attributes - which also influence rankings (alot). So are you going to now start a campaign for website owners to remove keywords from their TITLE attributes?
Which is a better use of your time? Debating whether or not Google is hypocritical about paid links or spending that time buying links? I don't see how this discussion is productive.
Google submitted rel=nofollow to a standards body (microformats.org) and got Yahoo and MSN on board. The meaning of rel=nofollow is "do not vouch" not "do not follow".
https://microformats.org/wiki/rel-nofollow
If Google says "you can gain a competitive advantage by using rel=nofollow in this proprietary way" and Yahoo suddenly decides, "rel=nofollow, when used in this way, indicates too much SEO" then it's a problem.
The standard should not be extended with proprietary usage that might conflict with the meaning that another search engine might give it.
"The meaning of rel=nofollow is "do not vouch""
Says who?
https://microformats.org/wiki/rel-nofollow
"By adding rel="nofollow" to a hyperlink, a page indicates that the destination of that hyperlink SHOULD NOT be afforded any additional weight or ranking by user agents which perform link analysis upon web pages (e.g. search engines)."
Actually, from Matt's blog....
Halfdeck, I appreciate your thoughts and opinions on this, and absolutely, I always think it is worthwhile to discuss and try to understand the underlying drivers that are driving changes in the engines... as long as it is productive, that is.
"Actually, from Matt's blog"
Yes, Identity, I know. He's said that a number of times. But I don't always take Matt's words as gospel, and then there's always context. Just because I said a fork is a nice tool to eat my steak with doesn't mean a fork can only be used to eat steak.
Google also tells you to build for users, not for search engines. If you follow that to a T, why should you implement 301 redirects (to prevent PageRank split), robots.txt disallow pages with duplicate content (to prevent PageRank dilution and reduce supplemental results), or META noindex thin content pages (to prevent wasting juice on pages that might not get indexed due to dupe content filters)?
The purpose of nofollow, no doubt, originally wasn't to control PageRank flow. But the implementation of nofollow allows you do exactly that. Paid links existed before Google. The original purpose of paid links wasn't to manipulate search results; people bought links for traffic and branding. But that doesn't mean you can't use them to rank higher on Google. If there's an opening, you take advantage of it. Are you going to tell me you won't buy links to improve your ranking because that's not what paid links were originally meant for?
I don't nofollow links I sell. I don't even care if the links stop passing PageRank because that's equivalent to an "invisible" nofollow on my paid links. My site won't suffer from that - in fact, Google discounting my paid links means more PageRank for my internal pages...and because the buyers can't tell I still get paid.
You now know you have a tool you can play with - you can take the red pill, or the blue pill. That's up to you. But arguing over ethics isn't going to make your sites or your clients' sites rank any higher.
Matt was one of the two authors of rel=nofollow so in this case his opinion on it probably carries a lot of weight.
"Matt was one of the two authors of rel=nofollow"
And guess why he named it rel=NOFOLLOW instead of rel=NOVOUCH. Nofollow is a cousin of META nofollow; take a wild guess what that tag does.
The nofollow robots meta tag means "do not analyze the links on this Web page". It is not for controlling comment spam. Rel=nofollow is misnamed.
"Rel=nofollow is misnamed."
Ok, now remember saying this?
"Matt was one of the two authors of rel=nofollow so in this case his opinion on it probably carries a lot of weight."
I'm not sure what your point is. It is widely agreed that it is misnamed.
You said that nofollow doesn't mean "do not vouch" and Matt's opinion that it means "do not vouch" doesn't matter. So I replied that he was one of the original authors so his opinion that it means "do not vouch" has weight.
We've both stated our points though. I'd rather not clutter up this thread with a long debate that rehashes the same points.
"I'm not sure what your point is."
According to you, Matt's wrong in one instance but he's right when it comes to something else. To back up your claim that he's right, you used his "authority." Then that should apply in both instances. Why doesn't it? Because in one case you agree and in the other case, you don't. I call that cherry-picking.
"You said that nofollow doesn't mean "do not vouch""
I didn't say nofollow doesn't mean "I don't vouch for this link." I said that just because I can use a fork to eat steak doesn't mean I can't use it for something else.
-- source
The specification wasn't written clearly, and the name is not clear. See the open issues with rel=nofollow.
One problem is that PageRank is only used by Google. Yet Yahoo and MSN also have accepted rel=nofollow. There is no standard being maintained and it will be difficult to tell exactly what effect it will have among different search engines.
A Web site should always vouch for its own internal links. To control robots, use robots.txt, robots meta tag, JavaScript links, or even Google's x-robots-tag HTTP header.
It seems like the endorsement of this use of rel=nofollow will only lead to complications.
Where's the word "vouch" in that paragraph, PocketSEO?
Like I said, the original intention of nofollow was to curb blog comment spamming; I don't disagree with you on that. But the original intention of robots.txt also wasn't to solve duplicate content issues. The original intention of 301 redirect wasn't to consolidate link juice. The original intention of paid links wasn't to improve ranking.
The original intent .. well who cares what something is originally meant for? Things change. We evolve. We adapt. We move forward.
The intention of robots.txt was always to control where robots go. HTTP headers (redirects) are a way to make sure that a site follows the rule of "only one possible URL per page of content". Those are standardized functions of those tools.
The issue is not whether you can use rel=nofollow to "sculpt PageRank". It's that rel=nofollow was intended to have a standard use across all search engines -- that is why Google submitted it to a standards body and got Yahoo and MSN on board.
Now Google is extending the standard with proprietary features that might conflict with the standard as used by other search engines.
What if Google suddenly changed the syntax of how it reads robots.txt rules in a way that directly conflicts with the standard way that other robots used robots.txt? In that hypothetical scenario you could write separate rules for Googlebot and other robots. With rel=nofollow on internal links you might have to cloak to Googlebot.
A Web site should always "vouch" for its internal links -- search engines should ignore rel=nofollow on internal links. There are several other standard ways to control robots that do not conflict.
People often forget to mention Yahoo and MSN when talking about "what works for SEO", but on large sites they can send a significant amount of traffic even if it's a smaller percentage. This new purpose of rel=nofollow might not be considered white hat by other search engines.
EDIT: Just wanted to add an example of how the current nofollow specification is confusing. Take a look at Technorati's internal linking structure. Does Googlebot think that Technorati does not vouch for its own internal links?
How about on a site with user-generated content where users link to internal pages? Does Google think those linked-to internal pages are not important even though they are getting cited often?
Search engines should ignore rel=nofollow on internal links because a Web site should always be vouching for its own content (unless specifically blocking it with tools like robots.txt, JavaScript, robots meta tag, etc.).
"Now Google is extending the standard with proprietary features"
Google is not extending the standard at all. The effect of nofollow is that link tagged with nofollow is ignored as if it doesn't exist on the page. That works for handling blog comment spam, and that also happens to work for reducing PageRank flow to a target internal URL. The way Googlebot behaves when it sees rel=nofollow hasn't changed a bit.
No one is putting a gun to your head forcing to use nofollow on internal links. It's just a tool for more precise PageRank flow control which ordinary SEOs/webmasters don't need to bother with. The fact that the ability is there doesn't hurt you or your website. If you don't like it, don't use it.
You can tell John Chow was penalized when you do a search on google for his name and see his domain appears on page 6, what's funny is that text link buyers still buy links on his site despite being penalized. Whoops!
I wonder if Matt would answer our questions if we emailed him or would he just ignore them? hehe.
I think it's great that he does answer Rand though. Thanks Matt and Rand for lots of good info.
Rand, the recommended page does rank for the full page title in quotes, although it is an indented listing. But it's there. (when i search anyway)
It sounds as though Google is putting more emphasis on the self-monitoring of a websites links. It makes perfect sense that a webmaster or company should have more structure over their site and the neighbourhood that it belongs to.
I mean they are giving everyone the chance to see where their websites fit in the natural searches by just clicking on the "Similar Pages" link.
Just re-read this thread again, arrggghh. 2. I am getting so tired of link policing. Why cant we let Google decide what pages are important and which are not. Requiring someone to use the nofollow attribute for better results in the SERPs is absolutely ridiculous.
So a niche resource directory that would be of value to users but is just starting out and does not have any pagerank should add "nofollow" to all outgoing links even if they are to reputable businesses? What about websites that aggregate news from different sources...all those outgoing links should also be nofollow otherwise they would drain the little pagerank the page has?
I understand how robot.txt and nofollow or noindex can be used to internally manage the flow of link juice but the parameters for outgoing links are still unclear because there are a few factors to keep in mind while doling out links
How can a new directory become viable? What should the link strategy be?
Opti, that is such a good thought about the directory. Isn't it true that a directory has to have links out to build reputation?
There is a 5PR directory site that has all 'follow' linking. This is even on some of the inner pages also. A 'business consulting' page had PR3.
It was a rare thing compared to other inner pages that were 0.
Should be noted that the high PR page did have less links out...
I did have https://www.BusinessOceans.com drop from PR2 to 0 the last three months and wasn't sure why. Maybe less new content was being created by us? The page is even ranking higher on Google than it used to.
How much value does PR have on rank nowadays?
What does seoMoz think is the best answer to this one? Please provide reference to test-cases if possible.
[quote]
If I have a website in Spanish that serves Spanish speakers in Mexico, Spain, Latin America & the United States, how does Google recommend that I tell them my content is relevant to searches in all of those countries? A) Host a single site in any country you want, but use a general domain extension like .com, .net, .org, etc. Then, acquire links from sites in all of those markets and Google will realize that your content is intended for an international, Spanish-speaking audienceB) Build a website for each country, but host them anywhere so long as they are registered with the proper top level domain. Feel free to use the same content on each, as duplicate content filters only apply to a single country. C) Build multiple websites for each country, host them individually in each country, register them on different top level domains and create unique content for each (even though the content might technically be exactly the same). Then, link build and market each site in each individual country separately. Don't interlink excessively between these sites. D) Use a single site for now and target your largest potential market. Then, give us some time to work on this issue – we'll have greater functionality for single sites targeting multiple countries/languages in the near future.
[/quote]
nice one - thanks for that Rand :)
I'm going to wait to see what comes back for Q3 - I would think 'c', but i don't think it matters where you host, as long as you have the right country TLDs , one of my key take aways from the DC and multiple sites session at SES San Jose was that Google (Yahoo and MSN) do not apply the DC filter to identical content IF the URLs have different country TLDs - so scraping sites and duplicating them if required across countries to gain ranks for in-country searches is OK - for some of my clients, that has been a big concern. Greg and Priyank gave that the nod on the panel during the SES Q&A, as did Vanessa when I posed the same Q to her at SMX Seattle.
It might not be ideal, but it can be vindicated, it's not illegitimate practice, and it is sometimes necesary. - this was really a great post.
Thanks for the information
great article.Rand .lots of insight there.
Great post. Thanks
With this current climate over the use of these and different search engines interpretting their use slighly differently it would probably make sense fot he 'no-follow' tag to be added for the 'eyes' of google only. Then they can use it how they want and the other engines dont have the issue.
I have been reading about this for a while with interest and I still cannot decide 100% whether it is sensible to use it or not for 1 very simple reason, to not pass the flow to my 'contact page' This page is for obvious reasons linked to sitewide and therefore an important page in theory.
I think it 'is' manipalating the 'G'pr. Whether it is done for logical reasons or not. It is definatly something that is done for the google serps and not for the visitor. There is no reason at all to do that for a visitor. When it comea to title tags etc then those things definately are done for the visitor so not the same kind of manipulation at all.
If my contact page is one of 8 main nav pages, that is a lot of pr going the other main pages of my site where i want it so it makes sense to use this knowledge but I think realsitically if there are existing ways of doing this that were originally meant for the purpose then it makes sense to use those.
Alternatively, I could remove my contact page altogether and use a pop up piece of flash or something with our contact details on. Something that the engines wont follow the same way anyway. maybe a pic even, im sure links to pics arent treated the same way as html pages of course. Tht could achieve the same result with no issues maybe?
Back of my mind I wonder eventually if not having a contact page might be one of a hundred things that counting towards trust rank.
Okay then, alternatively I could follow the link from the home page but add a nofollow link from the rest of the sitewide navaigation. That would say I have the contact page and it is trusted, but wont be followed from the other 300 pages or whatever.
With potential problems to small percentage of users with flash or java or those that cant be bothered to cut and paste my email address for example then I will probably go that final route.
I love the interview there has been a lot of buzz lately with directories being completely useless from what I got Google still places some authority on directory submission. Though you would think that purchasing a featured spot would be the same as buying links.
I have seen pages that pass link weight to 700 internal links (they are Pagerank 4 pages of a major site) so I believe that the more pagerank (not necessarily toolbar pagerank) you have the more links you can have on a page taht pass weight. www.tnsetop.com/blog
Great article. Thanks.
Incredibly valuable! Thanks!
Thanks Rand. This clarifies most of the questions i had in mind about nofollow :)
The nofollow robots meta tag means "do not analyze the links on this Web page". It is not for controlling comment spam. Rel=nofollow is misnamed. like iraqlife
Thank you for being the connection to a real answer that makes sense. Just like anything in life, there are no black and white answers.Do some thinking as look at the directory or site's potential or lack of quality.
My mind wonders about this as it was from 2007.
Matt, have you talked to the MCutts wizard behind the curtain recently about this topic?
I know this is an old post but check out a comment on this blog post by 'warrior blog' (9 comments in) then compare that with 'chetans' comment here (right at the bottom) I think we now all know what John Chow has been doing.. :S
This was very helpful. I will use the no follow in my new footers. I know this article is a little older but presume it still holds true.
Shouldn't Google and other search engines ignore rel=nofollowed internal links? A site would presumably always vouch for its own pages. Rel=nofollow was intended to prevent spam by saying "I do not vouch for this user-contributed external link".
It seems as though Google is trying make us prove to them that we understand our links and support them. Why else would a no follow link be used. It is simply linking someone to information that may or may not be beneficial to either party.
Thanks, saw the comment feature on the home-page and found this article =) Good information especially the part about the # of links on a page, I previously had thought the 100 was a hard number.
ok thts a very well done post. but i have one question. my site is a web 2.0 and user continously posts data. its an anime art site actually so on homepage there are always changing artwork thumbnails and link to them with anchor texts. i think u get what i mean. Same is teh case with some other parts of homepage like articles for writers on site etc. now whats disturbing me is that this data changes all the time. should i nofollow it ?
How do you see if a website is a no follow link? Can anyone help me on this?
View the source for META Nofollow tag, or the REL=Nofollow on A HREF links. Alternatively install SEOMoz.com Linkscape add-in or SEOQuake.com add-in into Firefox. These will visually highlight/crossout links on a page that have Nofollow on them. They are very useful tools for other reasons.
Great and insightful
I'm new to all this so this page has been a great help
Jamie
I used to think nofollow would help my site. Then after 50 pages I got google slapped. I wasn't sure what I did wrong. It could possibly have been that my posts weren't considered original enough. I still haven't got a clue though, but it managed to creep back up on the google search on its own and I even stopped posting. Not sure if it's helpful, but javascript links doesn't seem to need nofollow tags. Things keep changing though so can't really tell what to expect.
Thank you for the sharing of this interview with Matt Cutts, which bring us a lot of useful details.
I should have a question: It`s a post open in 2007, how much more all these details are valid in present, with the "simpathy" of Google for the video or other media content?
I put this question because without using SEO techniques, some days ago I brought a just started web site in the Google index in less than three days (in fact it was in 36 hours, but I can`t prove this), here is the example.
The power of media and Social Networks have changed some rules, at least so it seems to be...
All my respect for all of you
Nofollow is a great tool to help resolve the conflict between usability (run of site links to privacy / contact / (empty) shopping cart / etc.) and SEO, but these issues aren't really significant on smaller sites. On larger sites, maximizing the distribution of PageRank to important pages (product pages for example) can help get more pages indexed. It works as advertised, but it's not necessary for every site and every situation.
Very interesting information. Thanks Uncle Rand!
bobrains rocks!!!
Think for a minute. You are all affiliates or product owners, or whatever. Have you ever noticed that Wikipedia is the most loved site in Google SERPs. Google has this twisted idea that they know better than anyone else what a user is looking for. I might want to buy a solar panel. I am in the market, but Google thinks what I really want is a bunch of "Green" bull@%#$ to make me feel better about my time on Earth. Screw Google! I want to buy a damn solar panel and the only results I get to actually buy a solar panel are from paid search results that have been so badly PPC-guru-ized that I can't find a decent place to buy a solar panel there either.
Google wants their top search results to fit into their Star Trek Federation utopian approach to dishing out SERPs. They don't want you to show up as a valid result. How could free enterprise possibly bring people the result they want? Nobody really wants to buy anything. UNLESS the merchant uses google checkout. Hmmmm, maybe that plays a role in SERPs. Google:"The more they spend with us, the higher the pagerank. If they use analytics so we can spy on their traffic, we'll give them a nudge in pagerank."
I can guarantee this. Pagerank has far less to do with SERPs than you think. If you spend all your time running around chasing the PR grail, then you won't cloud up Google's utopian SERPs. I have a couple of sites with PR 0 and I am getting #1 SERPs in HIGHLY competitive markets. (No I won't tell you which sites.)
I love these guys that get on these forums and say, "Nofollow is worthless. Anyone messing with nofollow is wasting their time." Then with a little digging, you find their sites and they are loaded with nofollow.
Haven't you realized the "gurus" are misdirecting you so you don't encroach on their income? You pay money to be lied to so you don't infringe on the guru's niche. HA!
I know 50 of you are going to run out and search for solar panels and tell me your results don't reflect my scenario, but please remember it was an example. I don't want to waste my day looking for examples. I have better things to do.
IM doesn't really stand for Internet Marketing, it stands for Internet Misinformation. "If you can't succeed in the real world, you teach." "If you can't succeed in Internet Marketing, you become a guru."
If you ever actually pay money for these guru IM scams, make a real study of it. If they are showing you sites where they have great SERPs, study what they are saying, but also study THEIR site and see if there is something else lurking between the lines of code and the BS in their ebooks that is creating their actual success. Also, don't count too much on the sites they show you. The ones they are REALLY making money with will never be shown in an ebook. If they do show you their real success, they are idiots and who would want to listen to them anyway?
Pay as much attention to what Matt Cutt's doesn't say as what he does say. He doesn't answer some of the questions because there is no way to really answer those questions without uncovering an exploitation Google is aware of, but they don't want you to know about it. Where he doesn't answer, spend a day or two digging around within those questions and you will find a way to exploit Google before they can address it. Exploit! Exploit! Exploit until Google fills the gap, then go write and ebook and continue to make money on the exploitation even after it is gone.
MWa-hahahahaha!
I fear you are underestimating the sneakiness.
"Have you ever noticed that Wikipedia is the most loved site in Google SERPs."
Cant say I have. *giggle*
Lol, I am not sure how to treat your comment, seriously or as comic relief. lol, well written though.
Rand, great questions and I agree with the others that the nofollow info is the the biggest news, and certainly it hasn't been put clearer than that before, great question Rand and thanks for answering it Matt (even though it means I have to update feedthebot again)
One other thing as a compliment to the usefulness of this post...
This post has literally the most of the "old school" Google forum people respondimg that I have ever seen on SEOmoz in one place. Hello all!
D) He said mean things about Google and we're soft-shelled here at the 'plex
Classic :)
This is valuable stuff, no matter that Matt gave incomplete answers. Hope he gives us the rest later on! :)
"Feel free to use the same content on each, as duplicate content filters only apply to a single country."
Sounds like an avenue for content scraping and then slipping under the gaze of the all-seeing lidless eye that is Google.
"... even very important pages with lots of PageRank should stay away from linking to more than 100 other pages."
So how does this affect directories?
Thank you Matt!
The only ones I want to know the answer to, he didnt answer. Oh well.
"Directories (and all websites) that link out need to be very, very careful of who they link to, as this is a big way that Google's algorithmically identifying and discounting paid and other types of links they don't want to count."
*sigh* I must have listed one of the sites I've been trying to rank in a "bad" directory. I've seen some of the progress I made fade in the last 2 weeks after listing the site in a couple places.
I have seen pages that pass link weight to 700 internal links (they are Pagerank 4 pages of a major site) so I believe that the more pagerank (not necessarily toolbar pagerank) you have the more links you can have on a page taht pass weight.
Great post but I'm a bit gutted you didn't get an answer to question 3! It's my number 1 question at the moment - any advice gratefully received...
Yeah...and answer to #3 would have been nice...
Here, here... another vote for #3 to be asked again, should you ever get such an opportunity :)
Interesting reading, Rand! Just (too) curious: why the nofollow on the link to John Chow?
Probably in case Google thinks johnchow.com is a bad neighbourhood.
For anybody interested in how Rand thinks John Chow can get his rankings back read this interview.
Thanks for the link, mad4.
Wow, I am glad I found this site. There is so much info out there and so many people selling solutions to SEO -- I have met John Chow...and followed some of his advice, but was surprised recently to see his site is not ranked by Google. It is so true...becareful who you follow.
I would nofollow a link to John Chow. His site certainly isn't in a good neighborhood and He is definitely penalyzed. But it probably would't hurt Seomoz to link to him.
That's really helpful that you are able to share that info. I am in the process of evaluating directories and might not have thought about all that. Also, I like the input on the nofollow stuff. Thanks Rand and Matt!
For a while now I've found nofollow useful for pages that are of the Terms, Copyright, or Blank Form variety. Someone my go digging for those on site, but they don't have much value in drawing search engine visitors.
Doing some sort of seasonal nofollow trying to tailor product emphasis and such seems sketchy to me. I'd rather try to find ways to deep link to emphasized pages or imptove overall site structure.
"Doing some sort of seasonal nofollow trying to tailor product emphasis and such seems sketchy to me."
It does indeed seem sketchy, but internal link manipulation for PageRank appears to be okay, so let's abuse the hell out of it.
"For a while now I've found nofollow useful for pages that are of the Terms, Copyright, or Blank Form variety. Someone my go digging for those on site, but they don't have much value in drawing search engine visitors."
It may not have much value in drawing visitors, but having things like your privacy statement, terms of service, etc could possibly be useful when the SE's are determining a trust value for your domain. Just my 2 cents!
I doubt putting a nofollow on those pages would hurt any sort of trust value...
1. The copy is pretty much duplicate lawyerease from several other sources.
2. What would be the yard stick for measuring the trust via terms and privacy? Just the existence of the page, certain wording, guarantees? All that seems too easy to manipulate if you write it for the search engines.
As for abuse, you can't increase your page rank with the nofollow technique; rather, just direct your availble authority to your own internal pages.
To clarify, when I say sketchy I don't mean in a black hat way, I mean in the way where if I did it I'd probably break it or forget to throw the switch at the proper time or the SEs wouldn't cooperate with my timing...So 'sketchy' in the Indiana Jones jungle bridge sense. :D
<<Would you largely agree or disagree with my assertions about the value of generic web directories on the subject?>>
Rand, I think you were (are) on the money with your "value of generic web directories" post. Years ago (2000) I was interviewed on the radio and asked about the future of SEO. I mentioned that we would move in the direction of industry specific portals - which we see lots of today. I beleive those type of "directories" are the only ones of significance in the second tier realm, as they usually add value and bring valuable resources to a specific (non SEO) industry.
- OldSchool
Useful follow up and well asked questions - its hard to keep second guessing G. But even for questions Matt didnt answer, it would have been useful to see your own views on?
Exactly, Rand. It's just another tool. One that happens to be very easy to implement, although apparently difficult to understand if I must judge by some of the comments.
I don't see anyone objecting to the idea that site structure changes can make a difference in SEO, except for MM who is thinking about something other than PageRank when he says:
Which implies that the only reason to use nofollow would be to completely prevent a page from getting indexed.
"Which implies that the only reason to use nofollow would be to completely prevent a page from getting indexed." Which is not true. As Matt said above: "More like people can use it for internal links if they're power-user-y enough to want to sculpt PageRank flow within their site at the link level." Please define "sculpt" Matt. So we can control PR or hoard it as stated above on a internal page level? If so why would we want to? We as webmasters internally link out-bound for good reason, especially internally. I guess I am having trouble grasping why we as webmaster have to now differentiate high quality pages from low quality pages. Isn't this Google's job? And as Michael stated who could possibly guess how to do this correctly and most efficiently in our industry? Does someone know how and when to "regulate" page rank internally and exactly when? Please step forward. Since Google reps are recommending these actions I would appreciate a clearly example or description of how and when EXACTLY to nofollow internally to "sculpt" PR.
Love it :D basically it was D wasnt it? :P
Excellent bit of info Rand thanks! I try to get clients to nofollow links that they feel have no relevancy towards their sit, its always good to have some of these other questions clarified too
This link tells us all we need to know about link farm operators and the scams they are running, these bidding directory link farms are a ponzi scheme and a con game.
These bidding directory operators are scamming each other, they are all interlinked link farms and they are all doomed, if you like sailing on the Titanic, buy some links in bidding directories.
" if you like sailing on the Titanic, buy some links in bidding directories. "
Or if you feel like turning your competition into the Titanic, go buy them a few thousand.
Rand,
Placing SEOCO on the recommended page will more then make up for any bad neigborhood you might be linking to, you will rank no.1 for recommended SEO within 3-4 working days.
we'll have greater functionality for single sites targeting multiple countries/languages in the near future.
Great info. Especially on the global sites.
Thanks!
John Chow is an awesome place if you ask me and some of what he does get everybody to link to him or talk about him, giving him more traffic.
Maybe that is what he want.