Google’s embarrassing mistake.

I think it’s time we all agreed that the “nofollow” tag has been a complete failure.

For those of you new to the concept, nofollow is a tag that blogs can add to hyperlinks in blog comments. The tag tells Google not to use that link in calculating the PageRank for the linked site.

In other words, if I post a comment on your blog, and my comment includes a link to my site, people can click on that link to see my site as usual. Ordinarily Google would see that link and view it — as it views most hyperlinks — as an implicit endorsement of my site. This would ever so slightly boost my site’s ranking in Google search results. But if your blog software adds the nofollow tag, Google won’t give my site any added weight at all.

The half-baked idea was that if everyone adopted nofollow, it would quickly make comment spam pointless. The thinking was that comment spam is aimed at creating lots of links to a certain site, thereby boosting that site’s rankings in Google searches. Example: If I fill your comment pages with links to haiku, then maybe Google will start to believe that my haiku site really does have something to do with haiku.

Since its enthusiastic adoption a year and a half ago, by Google, Six Apart, WordPress, and of course the eminent Dave Winer, I think we can all agree that nofollow has done … nothing. Comment spam? Thicker than ever. It’s had absolutely no effect on the volume of spam. That’s probably because comment spammers don’t give a crap, because the marginal cost of spamming is so low. Also, nofollow-tagged links are still links, which means that humans can still click on them–and if humans can click, there’s a chance somebody might visit the linked sites after all. Heck, if we really wanted to eliminate comment spam, why don’t we just get rid of hyperlinks altogether?

Worse, nofollow has another, more pernicious effect, which is that it reduces the value of legitimate comments. Here’s how:

Why should I bother entering a comment on your blog, after all? Well, I might comment because you’re my friend. But I might also want some tiny little reward for participating in a discussion, contributing to the content on your site, and generally enhancing the value of the conversational Web. That reward? PageRank, baby. But if your blog uses the nofollow tag, you’ve just eliminated that tiny little bit of reciprocity. Thanks, but no thanks. I’d rather just comment on my own blog. And maybe, if you’re lucky, I’ll link back to you.

In fact, the solution to comment spam is simple. I’ve used it both on this blog, and on my haiku site. Here’s the step-by-step solution:

Step 1. Automatically moderate any comments that include hyperlinks.
Step 2. There is no step 2.

That’s it. Just throw any comments with hyperlinks into a moderation queue. If you feel like it, you can visit that queue every now and then to let legitimate comments through. Or if you’re busy or cranky, ignore the queue. Either way, no spam will get onto your site (well, almost no spam), spammers’ PageRanks won’t get boosted, and legitimate commenters will still be able to post their comments. And if you moderate up any comments that do have links, the commenters will get the PageRank lifts they deserve.

PS: Hmm, it seems that the great minds behind Technorati also seem to have been involved in nofollow.

Google’s embarrassing mistake.
  • http://www.silentblue.net/ quanta

    Whoa whoa whoa – who says nofollow has failed? Specifically, which studies?

    Comment spammers spam because links give them Google juice. It’s unlikely that any one person will happen upon your blog, spy a weird unrelated entry with nonsense words about poker or pills, and be compelled to click that link.

    If anything, nofollow fails because not ENOUGH people are using it. There are too many abandoned self-managed blogs with obsolete MT/WP/pMachine software that can be exploited.

  • http://dylan.tweney.com Dylan

    I don’t buy it.

    In fact, I’m getting comment spam where the spammers have *built in* rel=”nofollow” statements into their links. Where’s the Google juice in that?

    I *turned off* nofollow on my site because I think it’s a futile, self-damaging strategy — like cutting off your nose to spite your face.

  • http://www.silentblue.net/ quanta

    When I implemented nofollow, I saw a dramatic drop in comment spam, but a major assault on my trackbacks. I don’t believe nofollow is the ultimate answer – spam continues to evolve – but it doesn’t seem to cause any harm. It follows the programmer’s koan – never trust user input.

    I cannot explain your spam with nofollows in it – perhaps it was just nuisance spam or a test trial. All I can say is that my personal experience with nofollow has been favourable. Have you considered using JavaScript countermeasures?

  • Pingback: How to disable in WordPress - JW

  • http://www.shmula.com Peter Abilla

    I just posted about a fascinating study on Splogs and Spings by the ebiquity group at The University of Maryland. Great stuff; they look at weblogs data and conclude some fascinating things about authentic blogs and splogs and also propose a solution.

    http://www.shmula.com/?p=99

    Peter Abilla

  • http://gabrito.com Todd Huss

    I wrote a bit back in February coming to the same conclusion:

    http://gabrito.com/post/nofollow-hurts-bloggers-more-than-spammers

  • Pingback: Michael Tsai - Blog - Comment Spam and

  • http://dylan.tweney.com Dylan

    JW above suggests NoNoFollow as a tool for disabling nofollow in wordpress. I did some research into this before posting this item (wanted to make sure I could turn off nofollow before I started complaining about it!) and I’m now using DoFollow. It seems to work just fine and it couldn’t have been easier to set up.

    Also, Akismet does indeed filter spam to a fare-thee-well. Unfortunately setup is not so easy on that, since you need a WordPress API key, which you can only get by signing up for a free blog at wordpress.com. That’s strange, since many people who want to use Akismet might already have a blog, not hosted by wordpress.com. In my case, I have another blog at wordpress.com and was able to use that blog’s API key here. But the process is still confusing.

  • Pingback: Nofollow Has Noeffect at Binary Bonsai

  • http://www.spatiallyadjusted.com James Fee

    Actually Dylan, you can get a WordPress.com account without getting a blog.

    http://wordpress.com/blog/2006/05/18/sans-blog/

  • Pingback: The Nofollow Rule = No Good? : Business Blog Consulting

  • http://randomdude.com/blog/ Dustin Quasar Sacks

    I agree with you 100%. Spam is still useless space wasting garbage whether or not it has a no-follow tag in the links. I don’t see how a ‘solution’ to spam comments can still leave all those comments in place, like the no-follow ‘solution’ does.

    I’m glad to share a little page rank with people who leave good comments on my blog. If people leave spam or shitty comments then I delete them. It’s not fully automatic, but it works much better then site filled with no-followed spam does.

  • http://msittig.wubi.org/ Micah (via Zawodny)

    NOFOLLOW isn’t for people who moderate their weblogs, it’s for places where links are created but not monitored. All those abandoned weblogs and wikis that become link plantations for spammers? NOFOLLOW does a great deal of good there.

    And who’s to say that link spam wouldn’t be worse than it is now without NOFOLLOW?

  • Pingback: links for 2006-05-31 | Edward O’Connor

  • Pingback: Google’s Embarrassing Mistake at jarkolicious

  • http://www.leliathomas.com Lelia Katherine Thomas

    I used a plugin on WordPress to get rid of “nofollow.” It is a joke and is counter intuitive to the blogging experience. I think the only reason it was implemented is because there is still a small number of people who are too lazy to moderate their comments or too ignorant to know how to install spam blockers if they’re not going to moderate.

    Micah, you make good points, but who enjoys visiting a site where comment spam is all over the place, be it with or without “nofollow?” In my opinion, seeing a bunch of comment spam discredits the site and makes it appear that the owner doesn’t care enough.

  • http://tinywords.com dylan tweney

    Micah– I have an easier solution to the abandoned blog spam problem. Google should check how recently a blog has been updated, and discount links in comments if the blog hasn’t had any new posts recently.

    For that matter, Google itself could do spam filtering and discount links in spammy comments. After all, they’re the ones who are benefitting, by getting a better index, and they’ve got access to tons more data than I do. Nofollow fixes Google’s problem (pagerank pollution), not my problem (actual spam appearing on my blog).

    Blog data is structured enough that Google and other search engines could probably do a pretty good job of parsing something like that, don’t you think?

    But you are right, I don’t know of any definitive proof of nofollow working or not working across the web. The test would be fairly easy though: Set up 2 identical blogs, one using nofollow and one not, and measure how much comment spam each one receives over a period of several months.

  • http://www.hudin.com Michael

    My problem is less with comment spam (no comments on my blog, but maybe someday…) and more with crappy backlinks that show up in my logs. I’ve never understood the point of that spam. It’s pretty easy to cut off with a properly worded .htaccess file, but it just sucks overall. Anyone know the reason for this crap?

  • http://www.ragepank.com Harvey

    It’s certainly not as effective as we had hoped at removing comment spam – as you say, it comes thick and fast even when you announce in bold that all links will be nofollow.

    As you say, this could be because of human click traffic, but it could also be for other search engines. Does Yahoo or MSN use nofollow? Comments I have posted on nofollow blogs appear in a Yahoo backlink check, so maybe the spammers are doing it for other engines.

    What nofollow does do is ensure that Google doesn’t associate you with the bad neighborhoods that will hurt your ranking.

  • http://randomthoughts.vandorp.ca Darryl

    Michael,

    The backlinks problem is two-fold.
    A) *you* might click on the backlinks in your blog.
    B) lots of places publish their logs/backlinks (or at least used to)

  • Jen King

    It’s possible that “nofollow” has done little to reduce comment spam, but you are missing the benefit that adopting this tag has for Google – less pollution of their search rankings. From their perspective, it’s likely a success. That it hasn’t had much of a noticable impact for bloggers likely concerns them little as long as blogging platforms keep using the tag.

  • http://dylan.tweney.com Dylan

    Jen — exactly right. As I mentioned above, Google benefits from my using “nofollow” far more than I do. So if Google benefits, why shouldn’t the burden be on them to clean up their pagerank algorithm? Their computer is far more powerful than my web host’s, after all.

  • http://biocontact.org Jon Dugan

    Humans needs to be in the loop for deciding what is “spam” and what is not. There’s no way around this it until computers can much more effectively simulate conscious decisions. Give it about 8 years.

    What would be really useful now is for Google or other search engines to organize feedback on web junk and take it seriously: down ranking sites that many different, verified users say are junk.

  • http://dylan.tweney.com Dylan

    Jon, excellent point.

    Akismet, which seems to work very well, does a great job of aggregating human decisions and then applying them programmatically. Seems like a natural thing for something like Google to do.

  • Pingback: Mike Linksvayer » Suppressing spam in search results

  • Jeffers

    I think this blog post was your embarrassing mistake, given that you dissemble your own shakey argument in the comments. Thanks for playing though.

  • http://ebiquity.umbc.edu/blogger/ mayfield

    The logic of spam comments with nofollow links might parallel that of spam comments with no links at all. Peter Kaminsky has posted some interesting observations about such comments at . The comments to that post contain many interesting hypotheses about the spammers’ motivation.

  • http://ebiquity.umbc.edu/blogger/ mayfield

    Oops, I guess I was supposed to enter html. The link to Kaminsky’s post is here.

  • Pingback: The NoFollow Failure, NoFollow = NoBias at MaxPower

  • Pingback: Apogee Search Marketing Blog » Blog Archive » Zawodny on NoFollow

  • http://www.michael-martinez.com/ Michael Martinez

    Google crossed the line with REL=NOFOLLOW. No search engine should be dictating to Webmasters what they can and cannot put on their pages. Your listings at Yahoo!, Google’s listings, and ASK, MSN, and all other search engine listings constitute a social contract between you and the Webmasters: they put up what they want, you index what you want.

    I agree with you whole-heartedly on this issue. REL=NOFOLLOW has already been abused by sneaky Webmasters seeking link juice in reciprocation. It just needs to go away.

    The sooner the better.

  • http://blog.searchenginewatch.com Danny Sullivan

    Embarrassing mistake? Heck, nofollow was a great success, once you remember the real reason for it.

    Before nofollow, people like Dave Winer and others blamed Google for “creating” the trackback spam problem and not helping to control it.

    Sure, Google uses links. So do all the other search engines. But even without them, people would have spammed just for the link traffic below. It was never reasonable to blame the search engines for the mess caused by leaving open, unmoderated systems to accept links from anyone or anything.

    So finally Google trots out nofollow, and lots of bloggers finally think, “At last!.” But the SEO community said from the beginning it wouldn’t stop trackback and comment spam.

    As I summarized back then in my on nofollow:

    “The new attribute won’t stop link spamming. Many people may still spam simply because they hope human beings will see the links, click through and perhaps convert. As with email spam, maybe only an incredibly tiny number will do so. But since there’s no heavy cost to the spamming, that might still be enough.”

    So nofollow a failure in stopping trackback/comment spam? Sure. But anyone who understands marketing behavior never expected it to work.

    A mistake from from a PR front? Nah, it’s a big success in getting the search engines off the hook, as I wrote in a follow-up shortly after nofollow launched:

    “What is clear is that nofollow will NOT stop blog comment spam. Not at all. Don’t believe it? Then right now, all bloggers can stop making use of blacklists, registration schemes and other tactics used before emerged. Sit back and see if the spam goes away. It won’t. Nofollow is a nice new tool that we can use, one that as I’ve said many times before is welcomed for giving us choice and more options, but it’s not a magic bullet. Well, it’s a magic bullet for one thing. It now lets the search engines say to bloggers, we gave you want you wanted, stop blaming us for the problem!”

  • http://dylan.tweney.com Dylan

    Danny, thanks for your intelligent comments. You’re right of course.

    However I think Google may have fanned the flames of ignorance by calling
    nofollow a spam solution … and coming from them, it’s not surprising
    that many people took it that way.

    In retrospect, the discussion engendered by my post has convinced me that
    it’s not so much an embarrassing mistake, as it is a misguided effort by
    Google to enlist webmasters’ help in solving a problem (PageRank
    pollution) that they should have been able to solve on their own.

    –dylan.

  • Pingback: Fuzzuck.com · rel=”nofollow”

  • Pingback: Keentent » Breaking the Internet

  • http://blog.outer-court.com Philipp Lenssen

    Two thoughts:

    – while nofollow may or may not have had an effect on spam volume, it might have had a larger effect on search engine rankings. Spam sites may not spam themselves up the rankings as easily anymore, which makes their business harder. In the long run, this helps fight them. (However, I don\’t think we\’ll ever precisely know how much spam we\’d get these days without nofollow… we\’d have to know precise growth rates and assume they\’re constant.)

    – do like I do with legitimate comments; after 5 days, a comment that was not removed turns to an \”approved\” one, with all initial nofollows being removed. The cool thing is that even when you\’re not able to moderate for some hours during the day (even when you\’re a full-time blogger you need sleep), spammers won\’t get a tiny weeny bit of PageRank, as long as you\’re able to remove spam within N days.

  • Pingback: duncanriley.com » Sharing the link love, or hpw has failed

  • ef

    Add two or three checkboxes with the prescription: please check the first and the third and uncheck the second (which are random, of course).

    The bots wouldn’t get past that for a while, I guess.

  • http://www.info-pack.com/ Mal

    It would be interesting to know how much traffic they get from the blog spamming alone, without any effect from SEO value on the links themselves.

  • http://www.rcassicura.com rc

    I find the rel=nofollow attribute quite useful for controlling the flow of PR within my own sites, but I agree that it’s not at all effective in preventing blog spam.

    Personally, I think it’s only fair that someone leaving a legitimate comment gets some PR recognition for a minute or two of their time.

    As for spam, we just have to deal with it as it comes. There’s no “easy” way to get rid of it.

  • http://www.harveykane.com Harvey Kane

    Agree with the comment above about the human factor – there are so many instances of SEO where a human can pick spam in a second, whereas a computer will struggle. Take the display:none CSS value – could be used for hiding spammy phrases, or could also be used for rollover effects, menus, accessibility etc etc. Ultimately very hard for google to tell.
    So I’m happy to play my part with adding nofollow to links anyway.

  • http://www.pythian.com Paul Vallee

    I know I’m joining this conversation quite late. Sorry about that. I have a couple comments.

    First, my feeling is that pagerank flow is the primary objective of blog spamming. I don’t think any spammer realistically thinks that actual traffic (potential customers) can realistically click through their blogspam. The reality is that it’s a lot easier to send millions and millions of emails via … you guessed it good old fashioned spam.

    Secondly, I think that rel=nofollow is not likely to be anything other than a small parry in the battle against blogspam long-term. More important weapons are bayesian filtering and the creation of a cooperative spamdex. I also think the future of the user-created net has a lot more in common with reddit or digg. Think of the old-fashioned slashdot moderation system, improved. You can still post any nonsense to /. you want, but no-one will see it or read it. So big deal.

    Cheers,
    Paul

  • Pingback: the tweney review » Blog Archive » How to strike back at Wikipedia’s silly policy.

  • Pingback: on, spam and plugins / ajalapus.com / web log

  • Pingback: Rookie Nerd » Blog Archive » DoFollow: What’s all the buzz about?

  • Pingback: From BitchSlap to SmackDown - Why Google is just being DUMB! | InvestorBlogger

  • Pingback: Dylan, Jeremy and