Nofollow and How Google Really Verifies Links

Neil Patel has an interesting case study on how links with “Nofollow” may still pass some Google Juice.

Here’s what Nofollow Should be. It should be a way for publishers to tell the search engines that they have not audited the link and can therefore not vouch for it’s integrity. It should not be a way for publishers to tell the search engines that compensation was received for the link nor should it be a snarky inside joke among search engine geeks (ie, “I linked to you but nofollowed – chuckle, chuckle”).

If you have verified the link and you are putting it on your site, do not put a nofollow on it. It’s bad form.

Now, on to what Google is doing to Verify links. Because Google has so much market penetration with Analytics and the Google Toolbar, they can look at a link and verify it’s integrity (to some extent) by how many users actually click the link. If a link is relevant, it will have a higher clickthrough rate than one that is not. If a link is hidden or obscured, it will have a lower or almost non existent click through rate. In this manner, Google is using user behavior to verify the links.

And when you think about it, this really makes more sense. If a link is tagged “nofollow” but people keep clicking it and continue to surf when they get to the destination site, then who are you gonna trust? The webmaster who may not even know that his CMS put up a nofollow? Or the millions of users who click the link and seemed, at least algorithmically, to enjoy the content?

The answer is pretty obvious.

Both comments and pings are currently closed.

17 Responses to “Nofollow and How Google Really Verifies Links”

  1. [...] So there I am, trolling through my blogs, behind as always, when I come across this line from SEO Black Hat: [...]

  2. [...] Quadzilla followed up on Neil’s post with some ideas on what Google can do to verify nofollow links. Because Google has so much market penetration with Analytics and the Google Toolbar, they can look at a link and verify it’s integrity (to some extent) by how many users actually click the link. [...]

  3. [...] ..NoRanking. Davon bin ich überzeugt (ich glaube..), dass auch Nofollow Links beim Ranking in den Serps eine Rolle Spielen. Aber da ich ja weder ein SEO noch ein Dr. bin, kann ich das natürlich nicht unbedingt “wissenschaftlich” beweisen… Heute habe ich aber im Seo Black Hat Blog einen netten Artikel gefunden, dem ich nur zustimmen kann: Now, on to what Google is doing to Verify links. Because Google has so much market penetration with Analytics and the Google Toolbar, they can look at a link and verify it’s integrity (to some extent) by how many users actually click the link. If a link is relevant, it will have a higher clickthrough rate than one that is not. If a link is hidden or obscured, it will have a lower or almost non existent click through rate. In this manner, Google is using user behavior to verify the links. [...]

  4. DesiNet says:

    I think this theory is half wrong. No one clicks on spam, yet they still rank. The only packet the google toolbar send is the pr, which has no reference to the site the vistors came from aka clicked on.

  5. [...] There are now discussions at SEO Black Hat and The V7N Network intimating that Google may not be following its own rules with NoFollow. The theory, originally proposed by Link Building Blog, is that in some cases, NoFollow’ed links may still pass link juice. [...]

  6. aftermarket says:

    I agree with Desinet, and its all about links and links to those links, Google can’t track clicks on links unless they have some gestapo power I am unaware of. I had also posted this info on syndk8 about 2 weeks ago……Coincidence or not? -UnitedCrown

  7. QuadsZilla says:

    Google can, from the toolbar, determine when someone goes from one page to another. Sure, i suppose that the person could be typing in the url that happens to be a link on the page, but 98 times out of a hundred or more it will just be the person clicking the link.

  8. m1t0s1s says:

    I should mention the rel=’prefetch’ attribute.

  9. SixSigma says:

    I blogged about this last year when I did an experiment on a forum I frequent. The forum doesn’t require registration and it allows your name to link to your site, in the same way blog comments do. It fights spam pretty effectively by using a robust filter, moderators, and rel=nofollow for all links.

    The test was simple, I changed by display name to a unique term not used in any other links to my site and waited a few weeks. Then I checked Google’s webmaster tools (the sitemaps one), and that unique term was showing up as a back link keyword.

    Does it mean that it passes value? I don’t know. But it does mean that Google doesn’t ignore it.

  10. m1t0s1s says:

    To SixSigma: The general consensus is that google does index what is passed those links, but it doesn’t pass any value to them, unless the above post is true -).

    This is pure speculation, but it may take a certain number of nofollow links to actually be indexed. I will test this on my site, theycann.com

    This comes from google’s charter to “index all the worlds information”, no matter how valueless.

  11. m1t0s1s says:

    Quads, could you add this to my comment:

    Google for baconpolenta31337 and nofollownoindexnocache to see the results of my nofollow experiments.

  12. m1t0s1s says:

    sorry, I should of said

    google does index what is past those links.

  13. Josh says:

    I’m not sure if it was because Google followed the rel=nofollow. If he ranked #14 for ’spiderman 3 xbox review’, it is not surprising that he is #8 for ‘piderman 3 xbox review’. If you search Google for the latter term, Google will ask, “Did you mean: spiderman 3 xbox review?” Google guesses that “piderman” is a mispelling of “spiderman”.

    A more accurate experiment would be to try to rank for a nonsense word that has no results in the SERPs like “asdfjieakfnvd” — and only use that word one time in link text that is rel=nofollowed…

  14. adfunk says:

    I agree with much of what has been said.

    You may get some trackback traffic which may be worst more than any increase in your backlinks.

    Its a case of more of everything the better.

  15. [...] There are now discussions at SEO Black Hat and The V7N Network intimating that Google may not be following its own rules with NoFollow. The theory, originally proposed by Link Building Blog, is that in some cases, NoFollow’ed links may still pass link juice. [...]

  16. The nofollow attribute needs to be taken care of – and for the user data, temporal analysis must be a central ingredient to Google’s algorithms. Without the study of user data, the Google engines wouldn’t be capable of generating nice SERPs. I agree full-heartedly with “It should be a way for publishers to tell the search engines that they have not audited the link and can therefore not vouch for it’s integrity.”. I feel as though, if Google looked on the NoFollows in this way, then the ability for more quality content would increase. There needs to just be even more social congruency with the SERPs – Web 2.0 (People), Web 3.0 (Perfect data and Semantics), Web 4.0 (The WebOS/Flawless AI). That’s what I believe.