How Does Google Treat Nofollow?

In January 2006 Google went widespread with the Nofollow tag designed to give the impression (to search engines that respect it), ‘I do not vouch for this link’. All outgoing links now added the following to ensure they would not be penalised for chance of linking to a bad web site (that may hurt you rankings) or simply to be a blog snob.

EDIT: We’ve added some new information as of August 8th, 2007, so check it out.

<a href="https://link.com" rel="nofollow"</a>


The question perplexing many SEOs was, ‘what does Google REALLY think about nofollowed links? Well I posed this question and was convinced at one point that Google DID NOT EVEN CONSIDER nofollowed links in response to a post from Adam Lasnik (Google engineer). Link in this post here.

So I go on purporting that nofollow means literally, “I won’t follow”. That means if you’re link is nowhere online but through someone’s (or your own I guess) nofollowed link Google wouldn’t see it. People denied this in a spirited discussion over at SEOChat. I held my ground, but people were showing results where they could SEE the nofollowed links (i.e. Wiki links) in the SERPS (when using link: commands or other tools.)

I have found Wiki links in the SERPs so the whole ‘we don’t follow’ may not be entirely true. If it is, it has been a slow process in refining. What we can be pretty sure about is the value of those links is zero (PR juice wise).

Recently, Eric Enge gets Adam on the line again and asks him the same question about nofollow treatment. The response?

– NoFollow links do not pass PageRank.
– If a page is referred to via a nofollow link, we won’t include that page in our index unless it’s linked (without nofollow) from elsewhere.

Therefore, as you might guess, if there’s a page you want to have kept secret and absolutely kept out of Google, you’d be better off:
– putting it behind an authenticated login (best)
– using robots.txt directives or noindex meta tags

Ok fine, I’ll put it to rest now, it appears through the piles of rhetoric Google will SEE the link but won’t count it towards pageRank.

However, other news and thoughts suggest that Google may in fact check it out (it’s always back and forth :) (this time in reference to what shows up in Webmaster Tools.)

Yes. All links show up in Google Webmaster Tools, but Google Webmaster Tools does not make any distinctions between incoming links that pass link popularity and incoming links that don’t.

I see. But that’s not all.

Can Internal Nofollow Hurt Rankings?

For a while I’ve wondered and posed the question, “if nofollow means you do not vouch for a link, then would that mean placing nofollow on internal links means you don’t trust/vouch for your own pages?”

I’ve thought about this and on one hand it does make sense, however, that’s not what Google is saying. It’s as if they almost treat nofollow and exclusions from the robots.txt file as the same command (in the context of excluding pages from the index.)

This thread from HighRankings pits Matt Cutts essentially saying it’s OK to put nofollow on pages you do’nt want spidered. Apparently it’s an suitable method of managing your duplicate content (i.e. regular pages v. printer friendly pages.)

I’ve run some tests and everything is inconclusive. I see a number of forums using nofollow internally, and I’ve seen Googlebot spider nofollowed links (so evidently it takes time to process as with everything.)

Vanessa Fox when she worked with Google commented on a direct question about internal use of nofollow on an official Google blog. Her comments:

The number of nofollow links on a site won’t raise any red flags, but that is probably not the best method of blocking the search engines from crawling duplicate pages, as other sites may link to those pages. A better method may be to block pages you don’t want crawled with a robots.txt file.

I’m not entirely convinced that nofollow should be treated synonymously with a robots.txt exclusion. Perhaps deep within the Google code/algo/happymonkeysworkingatgoogle there is a difference between the way nofollow links are treated, and pages that are simply excluded with a robots.txt file. My question is: does nofollow provide any detraction from the ‘Trust’ or ‘Authority’ of a web site when used internally?