The many dangers of NoFollow

NoFollow linking has never been so prominent, and never has it been so dangerous for both ethical and practical reasons.

I don’t like the NoFollow attribute.  When it was introduced in 2005, it made so much sense.  But since then it has been abused by both webmasters and the search engines, and that abuse looks poised to make a quantum leap sometime soon.

Therefore, there are two mes that don’t like NoFollow:

  • The ethical me, who much prefers to be honest when I promote a website.
  • The practical me, who much prefers not to be slapped down, tied up and fed to a herd of half-starved ninja gators when Google wakes up in 2015 or 2016, or gets displaced by an upstart.

I will cover three things in this blog post.  Yes, I’m organized!

  1. The history of NoFollow, which many newer marketers today are unaware of, and many who were around in 2005 might have forgotten.
  2. The ethical case to avoid using NoFollow (As a matter of fact, it is important.)
  3. The practical case to avoid using an attribute that could blow up in your face in a few years.

The short, tumultuous history of NoFollow

The NoFollow “tag”, as it has often been called, is not a tag.  It is an “attribute” (for those interested in correct use of language), which can be added to any <a href=””> tag.  It tells the search engines not to follow the link, because the owner of the website on which it appears cannot vouch for its trustworthiness.  Just to be clear, NoFollow does not necessarily mean that a link is bad.  It only means that the link has not been vetted by the website’s owner or administrator.

NoFollow's sordid history

The NoFollow attribute was introduced in early 2005 to stop blog comment spam, or at least to make it easier for the search engines to distinguish between links from legitimate comments and links from spam-happy bots.

Here is the direct quote from the Official Google Blog:

Q: How does a link change?
A: Any link that a user can create on your site automatically gets a new “nofollow” attribute. So if a blog spammer previously added a comment like

Visit my <a href=”http://www.example.com/”>discount pharmaceuticals</a> site.

That comment would be transformed to

Visit my <a href=”http://www.example.com/” rel=”nofollow”>discount pharmaceuticals</a> site.

Q: What types of links should get this attribute?
A: We encourage you to use the rel=”nofollow” attribute anywhere that users can add links by themselves, including within comments, trackbacks, and referrer lists. Comment areas receive the most attention, but securing every location where someone can add a link is the way to keep spammers at bay.

Matt Cutts, Google  chief “web spam” spokesperson, said:

“Wherever it means that another person placed a link on your site, that would be appropriate.”

Matt Cutts confirmed this in 2009 on his own blog:

“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.’ In Google, nofollow links don’t pass PageRank and don’t pass anchortext.”

In other words, if you are not moderating your blog comments or other user-generated content, this will allow you to continue being careless or lazy or otherwise occupied without gumming up Google’s rankings.  And it’s not just Google.  MSN and Yahoo were involved in announcing simultaneously their support of the attribute.  In 2005, Google had about 37 percent market share, Yahoo had 30 percent, and MSN had 16 percent.  AOL and Ask Jeeves were still players, with ten and six percent respectively.

PageRank Sculpting

It was not long before some webmasters with overactive imaginations found a way to use NoFollow to their advantage through a method that came to be called “PageRank Sculpting”.

As you are probably aware, PageRank is the relative value of a page, and is the most visible of over 100 ranking signals.  Very roughly, the PageRank of a page is calculated based on the value of all the pages linking to it.  Each of those pages has its own PageRank, which it divides up evenly between all the pages it links to.  If you need to read up on the subject, I suggest this post by Danny Sullivan.

The key thing to understand about PageRank is that If a page contains 20 links, it divides its power 20 ways.  However, if it contains only 15 links, it divides its power 15 ways, sending more PageRank power to each of the 15 pages.

PageRank sculpting is the process of NoFollowing certain internal links, so that other internal links are more powerful.  The theory is that if every page of your website points to the contact, about, terms, and other administrative pages, that means a lot of PageRank power that could be going to money pages is being poorly directed.  By adding the NoFollow attribute to those admin links, webmasters believed that they were funneling more PageRank to their money pages.

To the best of my knowledge, nobody got penalized for doing this, but in in 2009 Google changed the way it read NoFollow links to make PageRank sculpting useless.  Webmasters got the idea, as PageRank Sculpting quickly went out of style.  But an important question about all this PageRank sculpting has to be asked, “What were they thinking?!?  NoFollowing links to their own pages from their own pages?  Telling the search engines that they can’t vouch for their own contact and about pages?  Saying, “Hey Google, I am such a shifty character that I don’t even trust myself”?  </rant>

Just wait for the other shoe to drop.

NoFollowing paid links

It was not only webmasters who played fast and loose with the rules.  Google took its turn, too.  In fact, Google now advises:

“In order to prevent paid links from influencing search results and negatively impacting users, we urge webmasters use nofollow on such links.”

Quite apart from the inconvenient truth that almost every link has been paid for in one form or another (yes, “earned” links can be very costly to “earn”), the fact is that there is no link more firmly vetted than a paid link.  A webmaster has to think much harder, “Is this money really worth possibly harming my site’s trust with visitors and the search engines?” than when they link for free.

NoFollowing unnatural/suspicious/random links

But Google seems to have moved past encouraging NoFollow just on paid links.  They seem to be quietly encouraging people to add NoFollow to a very widely defined array of low-quality links, unnatural links, suspicious links (those that might actually be natural, but Google really can’t tell the difference, so why not discredit them just in case) and seemingly random links.

Oh, and press releases.

These days, it seems that almost any link could be flagged as “unnatural” by Google, with so-called “manual” penalties being the result.  Many of Google’s recent manual penalties seem designed to upstage Monty Python.  Recovery from some of the more ridiculous penalties seems almost as random, and I have heard many people saying that by simply adding NoFollow to links, they have been able to recover.

In fact, many people writing about manual penalty recovery can be seen offering advice like this:

“After disavowing or no-following links, webmasters must submit a reconsideration request to Google. If the problem is not completely cleared, Google will send a denial message.”

Or advice like this:

“If it’s high quality, but just linked in the wrong way, ask the webmaster to add a nofollow attribute assigned to it.”

If you are wondering, “What’s next?”, so am I.  At this point, I have seen at least one example of almost every type of link drawing a penalty, and Google seems to be accepting  the NoFollow attribute as a way of crossing the blurry line of what is and is not acceptable on every third Tuesday, if the wind is blowing from the northeast with a faint whiff of Lavender in the air. In fact, Google has said that the Disavow tool is like a huge NoFollowifier.  Here is what Google’s John Mueller has to say on the matter:

“You don’t need to include any nofollow links…because essentially what happens with links that you submit as a disavow, when we recrawl them we treat them similarly to other nofollowed links.  Including a nofollow link there wouldn’t be necessary.”

Which brings us to today.  I watch, mouth hanging wide open (but not drooling on myself, just to reassure you), the mass NoFollowing of links that some desperate webmasters are doing.  There are plugins for WordPress, such as WP External Links and External Links.

I’ll go into why I think this is crazy below, but some highly respectable people have been driven by Google’s seemingly random penalties to actually use these tools.  Lisa of Inspire to Thrive  explains why she installed the WP External Links Plugin:

“I don’t agree with their nofollow policy or shall we say HINT of it but I don’t want to be penalized by this giant and I’d love to see how long the process takes so we can all learn something from this one.

Why NoFollow is unethical

You should not tell a lie.  NoFollow is ethical on user generated content, not because that is why it was created in the first place, but because it tells the truth.  Unless the website administrator moderates all user-generated content, such as on good quality blogs, the truth is that he or she cannot vouch for the links.  NoFollow truthfully communicates that to anyone who wishes to read that attribute, including search engines.

If NoFollow communicates that you cannot vouch for a link that you have in fact approved, that is a blatant lie.

NoFollow - what about your users?

Google is not the Internet. The main reason most people are adding the NoFollow attribute where it does not belong is in response to Google’s displeasure with certain links or the website administrator’s fear of Google’s displeasure with certain links.  Numerous statements by Google have led people to believe that Google wants people to add NoFollow to the links that Google has chosen to find irritating.

The problem is that Google is not the Internet.  There are other search engines and possibly other applications that will use your NoFollow attribute as a signal, too.  NoFollow tells others that the link is not trustworthy, too. It’s not just Google being lied to.

Read Google’s Webmaster Guidelines.  Google’s official guidelines, as vague as they are, are a lot more ethical than its enforcement is (Oh, that’s a whole other ethics topic that the company whose motto is “Don’t be evil” probably would rather I don’t get into).  Let’s see what the Quality Guidelines say:

“Don’t deceive your users.”

So if you are telling the search engines, I won’t vouch for this link, are you telling your users, too?  Just asking.

“Avoid tricks intended to improve search engine rankings.”

Like adding a hidden attribute, for instance.  Those who are old enough to remember what a search engine penalty meant before 2011, will recall that it meant you had done something sneaky and deceptive.  You were a dirty rotten crook, serving up different information to the search engines than to real people:

  • Doorway pages.
  • Hidden text.
  • Hidden links.

Search engines penalized websites for serving up different content to users and to robots, and rightly so.  So what about NoFollow links, where the link is viewable by users but not by search engines?

“A good rule of thumb is whether you’d feel comfortable explaining what you’ve done to a website that competes with you, or to a Google employee.”

“So, you see, I inserted this hidden NoFollow attribute because I don’t want to get in trouble with Google, but I’m OK sending my readers there.  Yes, I know that means I’m either a scammer sending users to a crap link, or a total wuss allowing Google to bully me into blocking robots from following a perfectly good link.” Hmm.  Sure, that’s what I would tell my competitor or a Google employee.

“Another useful test is to ask, ‘Does this help my users? Would I do this if search engines didn’t exist?'”

Seriously, would you put NoFollow in a link if search engines didn’t exist?

NoFollow means not taking responsibility for actions.  There are two main constraints that keep us from linking to bad neighbourhoods; because users might follow the links and because search engines might follow the links.  Putting NoFollow on bad links does not solve any real problem (it might help Google solve its problems), and makes it 50% more tempting to post a bad link.  In other words, far from cleaning up the Web, it is likely increasing the number of poor quality links, especially those posted on poor quality sites.

Why NoFollow is dangerous

Now, it might be that ethics are a less pressing worry on your mind than where you’ll find money to pay the rent, so maybe you are more interested in getting back lost rankings than in being 100 percent authentic and ethical.  Well, here are five reasons why NoFollow could bite you in that soft fleshy padding you sit on.

1. It might not work.  I have seen no official statistics on how many websites recover from different types of penalties, but it certainly sounds like a majority of those that bother trying don’t succeed on the first or second try.

2. You might lose rankings at Bing, Yahoo and other search engines.  Of course, they don’t have the same market share, so you might be willing to sacrifice all of them in order to access the 66 percent of searches that Google delivers.  But what if you NoFollow all your links and instead of winning Google’s approval, you simply lose your Bing and Yahoo rankings? Oops.

But that is just the short-term, and short-term is short-sighted, even if your main concern is next month’s rent.  The long term is what really counts.

3. Google might get you later on. At the current rate, half the Internet will be disassembled, Disavowed or NoFollowed before long, all because Google doesn’t want to count certain links in its algorithm.  What then?  The disassembled part (links people have removed) will no longer be there, but Google will have a huge database of domains that have been disavowed once, twice, thrice or 673 times.  Google will have a huge database of websites that have tons of NoFollow links pointing to them.  It won’t be hard to add into its algorithm a trust factor to account for how often a particular domain has been disavowed or NoFollowed.

Google will also have data on which websites NoFollow their links.  Ah, let’s follow the logic trail.  Google tells websites to NoFollow crappy links.  Website A has 300 NoFollow links on its site.  Website B has 3 NoFollow links on its site.  In Google’s mind, NoFollow means crappy links.  Hmmm.  Which site will Google consider more trustworthy?  Which site will Google see as less trustworthy?  It’s kind of a NoBrainer.  When you look at it from that perspective, is it worth sending such a negative message about your own website? Wikipedia will always be able to get away with it, but could your website?

Don’t believe this could ever happen?  Go back a few years when the best practice was to have keyword-rich anchor text in most of your inbound links, only to make sure you varied your text.  Now, websites are getting penalized for doing just that.  Go back a few more years when the best practice was exact match keyword anchor text.  That will land you in even more trouble today.

Google is now punishing websites for links that were built in accordance with their guidelines as far back as 2004 and 2005.  What you do today can come back to bite you tomorrow and even a decade from now.

4. Other search engines might get you later on. It’s just too easy.  Not every NoFollow link is crap and not every DoFollow link is amazing.  But if a search engine plays the averages, they can reduce the trust of websites littered with NoFollow links and increase the trust of websites that are clean.

5. Google’s rule is ephemeral.  I know it seems like Google rules the world.  But there are other search engines like Blekko and Duck Duck Go (as I wrote about here), and who knows where Bing or Yahoo might be headed?  Google controls 66 percent of search traffic now, but what if that share was to fall?

Can’t happen?  Think again.

Remember when Alta Vista ruled search?

Remember when MySpace was social networking?

Remember when Netscape was everybody’s browser of choice?

Remember when Digg was synonymous with social bookmarking?

Remember when Google ruled search?  It still does, but sooner or later, that question will come up.  And all the NoFollow attributes placed just for Google’s sake will serve as … what?

Your turn.  What do you think?

I would love to hear from you.  I certainly don’t have the last word on this.  I have not liked NoFollow from the start.  I called out Wikepedia on this in 2007, even going so far as to say Wikipedia should be spanked (I really do like the site; I just don’t like their NoFollow policy).

NoFollow made sense for what it was designed to do, but I have always thought that it sends a very bad signal to anyone watching, including search engines.  Obviously, not everybody feels the same way.  Some people might even today be using it for PageRank sculpting.

I would love to hear your thoughts in the comments below, or on blog posts of your own.  Support me.  Refute me.  Let’s get this out in the open and discuss it logically.