Is it time to retire rel="nofollow" ?

[seoskunk]

I just checked my wordpress account and it had 1000's of comment spam all using rel="nofollow" in the link text. This hasn't deterred spammers one bit at all. In fact in many ways its made it worse, to the user they can't actually tell which links from a site are untrusted and for a webmaster its almost impossible to get a genuine dofollow link from anywhere.

I think its time this code was retired....

[aakk9999]

It is true that does not deter automatically generated comment spam, but I think spam would be worse without it. It would also remove a tool given to site owners to declare links they are being paid for.

[n0tSEO]

seoskunk, there is nothing that wil stop spammers from *trying* alas. :(

The best tool that has worked for me so far with Wordpress blogs is the All In One WP Security & Firewall plugin, that almost completely stops spammers from posting.

It would also remove a tool given to site owners to declare links they are being paid for.

In all honesty, I'd love to see a replacement for that use; like rel=sponsored for example, or rel=advert. A bit of semantics wouldn't hurt (and could still work as a filter for link value if implemented well).

[aristotle]

This whole issue has developed into a farce. For example, consider these two hypothetical links:

1. A link from a front page article on the New York Times website (nofollow)

2. A link from Grandfather Jones' "My Favorite TV Reruns Blog" (dofollow)

According to Google, its algorithm disregards link #1, but might give some credit for link #2. This non-sensical approach is one of the reasons why Google's search results are so messed up.

[EditorialGuy]

I just checked my wordpress account and it had 1000's of comment spam all using rel="nofollow" in the link text. This hasn't deterred spammers one bit at all.

Rel="nofollow" may not have deterred the spammers, but it has kept their comment spam from transferring "link juice" (which is exactly what it's supposed to do: no more, no less).

[dethfire]

Most big sites nofollow everything. Hard to get good backlinks now even with good content.

[GreyBeard123]

Rel="nofollow" may not have deterred the spammers, but it has kept their comment spam from transferring "link juice"

Please be so kind to post proof of your comment

[piatkow]

Maybe the spammers are just quaint old fashioned things who are posting links for real people to follow rather than bots.

@EditorialGuy - I thought it was well established that nofollow doesn't prevent "link juice" from flowing out, it just lets it pour away rather than benefiting the recipient of the link.

[EditorialGuy]

I thought it was well established that nofollow doesn't prevent "link juice" from flowing out, it just lets it pour away rather than benefiting the recipient of the link.

That's why I used the word "transferring" instead of the phrase "flowing out."

[Robert Charlton]

Rel="nofollow" may not have deterred the spammers, but it has kept their comment spam from transferring "link juice"

Please be so kind to post proof of your comment

This in fact was the point of rel="nofollow", so it's not clear which aspect of EG's observation is being questioned... ie, whether the question is intended to suggest that the link juice is transferred, or perhaps that comment spam without nofollow is used so widely anyway that it hasn't made any difference.

IMO, nofollow has done a lot to help prevent comment spam distortion. Unfortunately, it hasn't stopped the spammers.

Regarding the link juice aspects of it... way back in 2007, Matt Cutts had stated publicly in an interview with Rand Fishkin that nofollowed links are dropped out of Google's link graph, and that Google didn't even use such links for discovery. This was true not only for the nofollow link attribute, but also for the nofollow meta tag at the page level. So far as I know, though some aspects of Google's treatment of nofollow have changed, this has not.

One of the things that has changed subsequently was the treatment of nofollow for PageRank sculpting, which makes piatkow's comment a little ambiguous....

I thought it was well established that nofollow doesn't prevent "link juice" from flowing out, it just lets it pour away rather than benefiting the recipient of the link.

The figure of speech we came up with is that nofollowed PageRank disappears into a "black hole", and use of it internally to sculpt PageRank flow within a site can, IMO, hurt the site that's using it.

Using nofollow on outbound links doesn't really effectively change anything with regard to internal PageRank flow. Google has hinted strongly that it does reward pages that link out to good resources... in the context of freely given editiorial links... and I think that such dofollow links are a good idea.

Our original discussion of PageRank sculpting here...

Google Changes Treatment of PR 'Saved' by rel=nofollow Sculpting
June 2009
http://www.webmasterworld.com/google/3925952.htm [webmasterworld.com]

There's also the consideration of the noindex meta tag, and its use with and without nofollw. In general, I suggest that if you do noindex a page, you should not also add nofollow.

[seoskunk]

Thanks Robert I didn't know that about pagerank sculpting. I was using "nofollow" on a site that I'm struggling with to prevent thin pages being listed in serps. I have now spent all evening stripping out the "nofollow" tags.

It seems "nofollow" has no real use, you can't use it internally on your site anymore, its failed to reduce spammers (in fact they actively seek nofollow links) and it has meant pagerank is no longer distributed throughout the web as it once was.

I would also like to add from a usability point of view there is nothing to show a link is nofollow in a browser and does not have the same level of trust as a do follow link. I think that in itself is misleading.

In Summary I say NOFOLLOW = NO MORE!

[EditorialGuy]

I would also like to add from a usability point of view there is nothing to show a link is nofollow in a browser and does not have the same level of trust as a do follow link. I think that in itself is misleading.

The link works the same way, from a user's perspective, with or without "nofollow." Whether it transfers PageRank (or the Yahoo/Bing equivalent) is unlikely to be of interest to anyone but the site owner or somebody who's involved in SEO.

[phranque]

I would also like to add from a usability point of view there is nothing to show a link is nofollow in a browser and does not have the same level of trust as a do follow link. I think that in itself is misleading.

trust is irrelevant to "nofollow".
as far as google is concerned, nofollow is a signal for links that are not editorially given - i.e. paid links or UGC.
there are browser plugins that will highlight links with the rel="nofollow" attribute if you are interested.

[Robert Charlton]

seoskunk - phranque and EditorialGuy explain the purpose of "nofollow" for links to external pages. I'm thinking that you may not be understanding, though, how best to keep your own thin pages out of the index....

I was using "nofollow" on a site that I'm struggling with to prevent thin pages being listed in serps. I have now spent all evening stripping out the "nofollow" tags.

"Listed" really isn't a defined term, but I'm assuming that what you're really wanting to do is to keep thin pages from appearing in the index. For a bunch of reasons, rel="nofollow" is not the way to do that.

For your purposes, I think the best way to keep a page out of the index is probably to use the robots noindex meta tag, but without the nofollow attribute. Note that the robots meta tag is "follow" by default, and "follow" allows the page to continue to circulate PageRank within your site.

If do you use noindex on a page, note that you should also not be using robots.txt to keep the same page from getting crawled. I'm going to suggest that you read this discussion, which explains the details. Maybe read it twice.... ;-)

Pages are indexed even after blocking in robots.txt
http://www.webmasterworld.com/google/4490125.htm [webmasterworld.com]

In Summary I say NOFOLLOW = NO MORE!

Not exactly. phranque and EG explain precisely why dropping it is a bad approach.

I almost never use rel="nofollow" on internal links. There might be an occasional rare reason to do so, but I can't think of one right now.

I also (almost) never use rel="nofollow" on links to external pages that I'm effectively recommending via editorial links.

If a link in a blog comment was user-placed and is a self-link, then under strict interpretations of the guidelines as phranque lays them out, it needs to be nofollowed. Either you gave the link editorially or you didn't. If the comments are moderated, then conceivably those comment links can be interpreted as editorial links, and this is perhaps why you're bring "trust" into it.

That may be stretching it, and I'd tend to go with phranque's strict interpretation, particularly say in a signature link, as oppose, say, to a link to an article that you're permitting in a comment. A lot may depend on how well the site is otherwise moderated, and how restrained that poster otherwise is in building his/her links.

I suggest that you continue to use rel="nofollow" for user-generated blog comments, etc. The 1000s of spammy blog comment links that have accumulated are something that you need to remove manually in the course of site maintenance.

Allowing the spam to sit around hurts both when the links are follow and when they're nofollow.

[aristotle]

I agree with those who say that the best approach is to put a noindex, follow metatag on each page that you want to keep out of Google's search results. Also, as others have said, the "follow" part isn't necessary, since it's the default.

But I would like to add that I don't think it's a good idea to noindex a large number of pages, certainly not a high proportion of all pages, because Google's algorithm might regard this as unnatural and downgrade the site. Normally, you should only noindex supplemental pages like "Contact" and "About this site". So I suggest that you avoid noindexing a lot of content pages, even if they are "thin".

[rish3]

trust is irrelevant to "nofollow"

Matt Cutts has made several statements indicating that "nofollow" IS about trust.

Two examples:

Nofollow is method (introduced in 2005 and supported by multiple search engines) to annotate a link to tell search engines “I can’t or don’t want to vouch for this link.”

I think this is pretty smart. Digg isn’t adding nofollow to everything, just the links that they’re less sure about. Once a story looks real to them, I can imagine that they would lift the nofollow.

He's arguably one of the two, er, inventors, of rel="nofollow".

[MikeNoLastName]

follow, nofolllow, all just clues to G which can easily backfire or change meaning down the road. After doing an inventory a couple years back after getting hit by Penguin/Panda, and running sw to find broken links, I was amazed/appalled to see how many old "trusted" external links now pointed to hijacked domains in bad neighborhoods. So from that point on I made it a point to not directly link ANY external domain that I did not control and which did not belong to a government agency (amazingly even a couple of these for things like national parks and attractions had been moved and the old ones scooped up by spammers) or media source or charity org which has been around longer than 3-5 years. In these cases I use the URL itself as the anchor text. With all the others I simply list the URL as plain (unlink-ified no HREF) text (usually formatted with color=blue and underlined) and the user can take the second or two to cut and paste it if they are truly interested, and G can't penalize us either now for an unknown bad neighborhood, or for looking like a paid link, or in the future if the good link turns to mush.

[EditorialGuy]

G can't penalize us either now for an unknown bad neighborhood, or for looking like a paid link, or in the future if the good link turns to mush.

On the other hand, they can't reward you for linking to useful and relevant resources, either.

Back in 2009, Matt Cutts wrote:

In the same way that Google trusts sites less when they link to spammy sites or bad neighborhoods, parts of our system encourage links to good sites.

Source:

[mattcutts.com...]

[seoskunk]

I simply list the URL as plain (unlink-ified no HREF) text

I wouldn't be so sure google doesn't crawl those urls as it now can crawl variables in javascript that contain urls. I think Googlebot crawls everything it possibly can, its kinda its job.

Agreed though that this type of link may not get you penalised, but then again it might. Depends how google feels in next update.

[MikeNoLastName]

edguy
> they can't reward you for linking to useful and relevant resources, either.

True, but better safe than sorry, I feel, at least in these days of penalties. Which is why we DO link to the almost-certainly 'safe' sites mentioned above directly... just not with keyword anchor text which could also be a possible bad signal if G doesn't seem to THINK it matches the destination, who knows. For instance, say this month a well known goody-goody charity is sponsoring a holiday contest and you can sign up online, so 'enter (holiday name) (charity name) contest for fruit basket' might be a perfect anchor link on the archived November edition of a newsletter. The following month when the contest is over they may 301 redir it to their home page or change it up ready for the next different holiday and prize drawing, and now the description no longer applies. Probably not the best example, but you get the idea about how quickly things change on the net and can affect your text link.

Should one guess that 1 accidental bad link could wipe out the outgoing ranking-goodwill of say 3 useful, good links? Why risk it?

Does anyone have any input or evidence that a url anchor text is any better or worse than say an overused keyword anchor for the LINKER? Is G going to assume that since I used 'widget' as a keyword to link to another site, that MY site is about that too? Hope not. Maybe I'm linking because I DON'T have or want to cover that content or product myself, but the widget is my recommendation as a good cheap tool to calibrate my product (amongst many other unrelated applications). This might be one good use of a nofollow? They aren't paying me to sell or recommend their product, but it is to my advantage to link to it in order to sell my own, but does G KNOW that?

Admittedly, the one disadvantage to listing the URL rather than providing a HREF is that you can't automatically open a new tab/window, so the visitor may tend to replace your URL with the cut-pasted one and leave your site.

seoskunk:
G can crawl anything they want, goodie for the linkee. Does it subtract from my rank juice? Don't know, as long as _I_ don't get PEANLIZED in the process. It's not that I don't trust the URL (at the time I first mention it) or don't want the listed URL to get traffic, (I wouldn't mention it at all, if that were the case) I just don't trust G to make the right judgement as to what is good or bad, the majority of the time. As paranoid as they seem to be lately about search results, I would assume they often tend to err on the side of 'when in doubt, cut rankings first ask questions later.'

'Do no evil' requires that one ALWAYS KNOWS what is evil.

[seoskunk]

Have you tried info:http://www.yoursite.com on Google, it now returns

Find web pages that contain the term "www.yoursite.com/"

Preperation for penalising text urls maybe?

"Do no evil" requires an inflated ego and judgement on others IMO

[tangor]

Unlikely any html (or G) thing will be deprecated on the web. Can say it will be, just never is! The web will (want to) grow its regs just like guberment (sic) always does... getting more tricky and less reliable along the way, increasingly granular and intrusive, increasingly deranged.

It is Human nature to "fix this, fix that" "one perceived outrage to the next" "depending on who has the whip hand or not" and never fixing anything.

Run a clean site, valid links, and get on with life. That will always continue to work. "Open range" is better than "fenced" or "managed by committee".

[CaptainSalad2]

The only problem with rel="nofollow" lies in Googles PR spokesmen’s inability to communicate in an non ambiguous manner.

There is to much confusion on exactly when it should be implemented thus resulting in an increasing number of sites adopting a global nofollow policy!

Site A reference site B for the benefit of site A’s users not for site B’s rankings, since it makes no difference to site A’s users how the link is marked up a “better safe than sorry” mind-set is being adopted unless some sort of afflation / relationship exists between the site owners IMO.

Why should any site owner care to leave off the nofollow attribute unless they have some sort of "arrangement"? Of course there are the odd exceptions but these are becoming fewer and further between IMO.

[bumpski]

A webmaster should either stand by a link or REMOVE it!

Remember Google's clearly stated corporate philosophy:
From another thread:

[google.com...]
Ten things that we know to be true

We first wrote these “10 things” when Google was just a few years old. From time to time we revisit this list to see if it still holds true. We hope it does—and you can hold us to that.

4. Democracy on the web works.

Google search works because it relies on the millions of individuals posting links on websites ....

When were there ever millions of individuals posting links that resulted in actual citations from the individual. I have many pages listed as references, sources, citations, and probably, in many cases, the author creating the actual citation has no idea that the webmaster (corporation) of the site has invalidated that citation with a "rel=nofollow". If a "webmaster" allows links from individuals on his site he should stand by all of them or selectively delete them.
"rel=nofollow" partially invalidates Google's own stated corporate principles.

[EditorialGuy]

Why should any site owner care to leave off the nofollow attribute unless they have some sort of "arrangement"?

Maybe because some links are actually citations, as opposed to comment spam, ads, or purchased links?

The Web default is not to use rel="nofollow." The attribute exists as a convenience for blogging and forum platforms that want to discourage comment spam, and for site owners who sell sponsored links or ads directly and don't want to bother with ad-serving software. Unless you have a good reason for using rel="nofollow", ignore it. Nobody (certainly not Google) is pressuring you to use it when it isn't needed.

[rish3]

The Web default is not to use rel="nofollow."

It is perhaps still the "web default", but there are several major sites where editorial links DEFAULT to rel="nofollow".

[EditorialGuy]

It is perhaps still the "web default", but there are several major sites where editorial links DEFAULT to rel="nofollow".

Sure. Owners of Web sites, even big ones, do all kinds of stupid things. I've seen corporate-owned sites that had warnings not to link to them without permission. (How dumb is that?)

[rish3]

Owners of Web sites, even big ones, do all kinds of stupid things.

What's notable in this case is that they are doing stupid things based on fear. And the source of the fear is curiously quiet on the topic.

[EditorialGuy]

What's notable in this case is that they are doing stupid things based on fear. And the source of the fear is curiously quiet on the topic.

Google and the other search engines that support rel="nofollow" have said when and how it should be used. It isn't their job to respond to fear-mongering and conspiracy theories.

[CaptainSalad2]

>>Google and the other search engines that support rel="nofollow" have said when and how it should be used. It isn't their job to respond to fear-mongering and conspiracy theories. <<<

A company that basis it number one ranking signal off the correct usage of said signal doesn't need to respond/explain how it should be used to prevent misinformation? Interesting..

[edited by: CaptainSalad2 at 5:10 pm (utc) on Sep 20, 2014]