A Look at Some of the Negative Feedback on rel=”nofollow”

If you’ve been following the blogosphere you should know by now that the Google, Yahoo! and MSN search engines decided to start honoring the rel=”nofollow” attribute on links to mean that the linked page shouldn’t get any increased ranking from that link. This is intended to reduce the incentive for comment spammers who’ve been flooding weblogs with links to their websites in comment fields. There is another side effect of the existence of this element which is pointed out by Shelley Powers in her post The Other Shoe on Nofollow where she writes

I expected this reason to use nofollow would take a few weeks at least, but not the first day. Scoble is happy about the other reason for nofollow: being able to link to something in your writing and not give ‘google juice’ to the linked.

Now, he says, I can link to something I dislike and use the power of my link in order to punish the linked, but it won’t push them into a higher search result status.

Dave Winer started this, in a way. He would give sly hints about what people have said and done, leaving you knowing that an interesting conversation was going on elsewhere, but you’re only hearing one side of it. When you’d ask him for a link so you could see other viewpoints, he would reply that “…he didn’t want to give the other party Google juice.” Now I imagine that *he’ll link with impunity–other than the fact that Technorati and Blogdex still follow the links. For now, of course. I imagine within a week, Technorati will stop counting links with nofollow implemented. Blogdex will soon follow, I’m sure.

Is this so bad? In a way, yes it is. It’s an abuse of the purpose of the tag, which was agreed on to discourage comment spammers. More than that, though, it’s an abuse of the the core nature of this environment, where our criticism of another party, such as a weblogger, came with the price of a link. Now, even that price is gone.

I don’t see this is an abuse of the tag, I see it as fixing a bug in Google’s PageRank algorithm. It’s always seemed broken to me that Google assumes that any link to a source is meant to convey that the target is authoritative. Many times people link to websites they don’t consider authoritative for the topic they are discussing. This notion of ‘the price of a link’ has been based on a design flaw in Google’s PageRank algorithm. Social norms should direct social behavior not bugs in software.

A post entitled Nofollow Sucks on the Aimless Words blog has the following statement

Consider what the wholesale implementation of this new web standard means within the blogosphere. “nofollow” is English for NO FOLLOW and common sense dictates that when spider finds this tag it will not follow the subsequent link.

The author of the blog post later retracts his statements but it does bring up an interesting point. Robert Scoble highlights the fact that it didn’t take a standards committee to come up with this just backchannel conversations that took a few hours. However as Tim Ewald recently wrote in his post “Make it easy for people to pay you”

The value of the standardization process is that it digs issues – architectural, security, reliability, scalability, etc. – and addresses them. It also makes the language more tighter and less vague

The Aimless Words weblog points out that it is unclear to anyone who isn’t party to whatever conversations that went on between Google, MSN, Yahoo and others what exactly are the semantics of rel=’nofollow’ on a link. Given that it is highly unlikely that all three search engines even use the same ranking algorithms I’m not even sure what it means for them to say the link doesn’t contribute to the ranking of the site. Will the penalty that Yahoo search applies to such sites be the same in Google and MSN search? Some sort of spec or spec text would be nice to see instead of ‘trust us’ which seems to be what is emanating from all the parties involved at the current time.

PS: I was wondering why I never saw the posts about this on the Google blog  in RSS Bandit and it turned out to be because the Google Blog atom feeds are malformed XML. Hopefully they’ll fix this soon.

Comments (12)

  1. SuperJason says:

    In my opinion, the best solution is to use the "nofollow" attribute for anonymous users, and don’t use it for non-anonymous users. The trick is to determine which you are. One of my favorite ideas, but not without it’s flaws, is to use an image where the user has to type in the randomly generated string.

  2. The nofollow attribute looks like another way to artificially ignore people, which is what empires are made of… I would rather be explicitly reduced to the absurd but then again Condi Rice seems immune to such public demonstrations so let the corporate artifice begin…

  3. Anonymous says:

    » Reaction To "nofollow"  InsideGoogle – part of the Blog News Channel

  4. Shouldn’t Google fix a problem in its algorithm by fixing the algorithm?

    When searching Google, I often just want to get to the page that people associate with a certain issue, regardless of whether that source is authoritative or not. Because with this change, you leave it up to other (mostly) unqualified persons to judge the authoritativeness of sources, which will probably yield the same crap in the end.

  5. David Sifry says:

    Technorati already supports the nofollow tag – we still index the posts and links, but we don’t count them in our authority rankings, see my blog for more details…


  6. According to the Google announcement:

    "From now on, when Google sees the attribute (rel="nofollow") on hyperlinks, those links won’t get any credit when we rank websites in our search results."

    This sounds like the indexer will still index the linked page, but the PageRank calculation simply doesn’t take that link into account (the link doesn’t exist).

    But I agree with you that there is a good chance that the three major search engines possibly could have slightly different interpretations. That happens even with detailled standards 😉

  7. Bill de Hora points out an interesting feedback loop: PageRank is based on the theory that links carry influence, now Google is asking the web to be modify to sustain this theory better (in face of adversity).


  8. Lachlan Hunt says:

    The nofollow relationship is indeed very harmful. I also covered many reasons for why it is, and also discuss a much better alternative with more well defined semantics in parts 1 and 2 of Link Relationships Revisited.