MediaWhiz Services

Do Nofollow Links Count?

Generally we say that nofollow links don’t count, but Ben
Fisher actually found that they do help with ranking based on a small experiment he ran. He did an interesting study where he wrote on the SpiderMan
3 game and then commented on other related blogs that linked back to his SpiderMan 3 post. The unique thing about the link was that it contained the anchor text "piderMan 3" and contained a nofollow, but within a short period of time his site ranked on the first page of Google for piderMan 3 Xbox and other terms.

Piderman_3_xbox_google_search_11785

If you look at the image above it shows that there are 224 results for "piderMan 3 Xbox". This means that it should be fairly easy to rank for the term, but the actual blog post that ranks for the term doesn’t have the word piderMan anywhere within the content. The only reason I think it ranks is because the nofollow link contained the anchor text "piderMan".

I am not sure if the way Google looks at nofollows is currently messed up on their end or if they actually do count them, but based on this experiment it seems that they can help with rankings. What do you think?

May 7th, 2007 | Posted by Neil Patel in Uncategorized | 71 Comments

Ben Uy on May 8th, 2007 at 1:09 am

This is interesting info… I never bothered with nofollow tags before

Michel on May 8th, 2007 at 5:54 am

digitalbattle.com have recent comments autor links without nofollow. One of the experiment link was a comment in that blog.

Martin on May 8th, 2007 at 7:50 am

IMHO: If Google knows “piderman 3″ equals “spiderman 3″ (see “Did you mean…” in the screenshot), then it knows that pages containing “spiderman 3″ are also relevant for query “piderman 3″. This experiment isn’t a good proof (althought I think Google counts nofollow links).

Jeremy Luebke on May 8th, 2007 at 10:44 am

Sorry, I have to call BS on this one. The page contains the exact phrase “piderman 3″ and is part of the blog internal navigation. This has zero to do with nofollow links.

Jestep on May 8th, 2007 at 11:12 am

I have speculated for a while now that Google may count some nofollow links. It seems almost like the links do not effect the site that they are on, but they still effect the site they are linking to in some situations. Maybe, nofollow links are associated with some sort of trust ranking. If the site being linked to is trust-able, the links count, if it’s new or questionable they don’t.

I’ve seen this a few times with Wikipedia since the links were nofollowed. There was an immediate boost after a link was added even though it had nofollow attached to it. When the link was removed, the ranking went down.

Neil Patel on May 8th, 2007 at 12:00 pm

Jeremy, thanks for the heads up… I missed the internal navigation. I guess I should never trust the search and find feature in IE 6. ;)

Tom Ferris on May 8th, 2007 at 1:05 pm

I think that this is very interesting, but have to say that I dont think it has anything to do with the comment he made. I think that its just a very easy keyword to rank high for…

Halfdeck on May 8th, 2007 at 2:42 pm

Nofollow leaks anchor text. I tried it months ago by using a nonsensical nick on problogger. A few days later I was ranking first for that term. The effect is probably minor but still, nofollow doesn’t block anchor text completely.

Jeremy Luebke on May 8th, 2007 at 3:40 pm

Now show me a page that ranks for a completely made up word “without” the word being on the page and the only links using that anchor text being nofollowed. That I would be interested in.

WeCanDoATestAndSeeTheResult on May 8th, 2007 at 11:41 pm

I have done all types of testing in the past but might as well do a public one.

AndyBeard on May 8th, 2007 at 11:44 pm

I will also have to do a test somewhere else because you have that horrible Typepad redirect script as well.

Mike Sansone published instructions on how to remove NoFollow and the redirect from comments on Typepad a while ago.

cdc on May 9th, 2007 at 3:01 am

I ran a test back in January:

http://www.wagerank.com/2007/the-nofollow-experiment/

The results are that nofollow links won’t help you get indexed in Google or MSN, but it does help in Yahoo.

Michael Martinez on May 9th, 2007 at 11:43 am

Jeremy is right. The test is invalid because “piderman 3″ occurs on the site itself but at least someone is testing the implementation. People should conduct nofollow tests periodically just to keep the research current.

Michael Martinez on May 9th, 2007 at 11:45 am

Jeremy is right. The test is invalid because of the onsite occurence of ‘piderman 3′. Still, it’s good to see people poking the algorithm. SEOs just need to tighten up their testing tactics and the research will become more reliable.

webee on May 9th, 2007 at 12:03 pm

well… i’ve seen a lot of blogs releasing their dofollow links lately.
so i think this may not be so important to bloggers.
off course if we’re talking about a commercial site… we should aways keep our ears up !

webee
[is a design blog]

Social Dude on May 9th, 2007 at 3:09 pm

Hi all,
Jeremy, where on the page do you see just the word “piderman 3″? I did not type it on the page intentionally and have looked a few times since this post and still do not see it.

Let me know, if that is correct then on to more testing..heh well I will be doing more testing anyway :P

Arcangelion on May 10th, 2007 at 3:48 am

Hola, alguien sabe como es que google calcula el pagerank, es k mis blogs una vez llegaron a tener 5 y 4 ahora tienen 3 y 2, a k se debe esto?
http://transformers-peru.blogspot.com
http.//cursosymanuales.blogspot.com
http://networking-solution.blogspot.com

Att
Arcangelion

Ralf Skirr on May 10th, 2007 at 8:08 am

Interesting experiment Neil.

SEO company on May 10th, 2007 at 10:03 am

Really interesting Blog…
I’m glad i found you!

Unitedcrown on May 10th, 2007 at 9:13 pm

I had posted this info on syndk8 in april.

Comtom on May 10th, 2007 at 11:17 pm

Great info!!!
i just finished uploading my blog..
check it out!

http://modit.com.ar/

tell me qhat you think about it!
wy i still hav a PR 0?

SEO company on May 11th, 2007 at 9:13 am

Working on misspellings and human mistakes is a good way of doing SEO when you realise how much traffic you could get from those misspellings!

Haytham on May 12th, 2007 at 3:17 pm

That is very interesting. I have been avoiding creating no-follow links for the past year or so. Those links can’t be that strong can they?

Blog Guy on May 13th, 2007 at 3:01 pm

Hi there,
thats a quite interesting topic, thanks for informing us!
I had the intuition, that it would be like this, but I never have tested it.

Paul on May 13th, 2007 at 8:34 pm

I had always figured that even with nofollow the search engines may establish a small amount of value to a link depending upon it’s relationship to the content. Great experiment and thank you for sharing!

awflasher on May 14th, 2007 at 10:55 am

So the spammers also rocks.

Phillip on May 15th, 2007 at 1:51 pm

Hi all. From an SEO point of view, i just noticed a really interesting quirk in Google spelling and SERP result with this phrase as well

Spider-Man 3 is how Sony / Marvel spell it hyphen and all, and Google have the whole top 50 just like that except one site and they will be juiced to the hilt,

and yet they offer “spiderman3″ and the correct spelling,,, no hyphen,
interesting stuff really,
i registered http://www.spider-man-3.info last month too, if anyone wants a crack at it,

Philski

COMPAQ on May 16th, 2007 at 4:37 am

It sounds feasible.However ,the amount of the index are only 224 .It’s need a more forceful proof-test.

Cornwall SEO on May 17th, 2007 at 10:18 am

Interesting, I am definately going to run my own tests and publish on my blog.

jigsy on May 21st, 2007 at 3:09 am

It would be interesting to see if a true test one done. I have seen yahoo count no-follow. The code needs to be cracked

michal frackowiak on May 21st, 2007 at 9:39 am

I do not really think this experiment is relevant…

Google does not only search for direct word occurrences (in anchors or page content) but also takes into accounts keywords that are:

* misspelled forms
* have connection to keywords that appear in the page

So if the Google finds a relationship between “piderman3″ and “spiderman3″ (e.g. by finding both of these on a number of pages) then the results of the experiment can be interpreted alternatively.

I would suggest trying keywords that are unrelated, sth like “usjcnjsbhgauxas” ;-)
Good luck! A nice blog anyway!

michal

Jenny on May 27th, 2007 at 1:24 am

Yea, dofollow helps your rankings A LOT. I’ve been using it for a while.

brazz on May 28th, 2007 at 2:20 pm

When wikipedia decided to add nofollow to all their links, Matt Cutts said that he wouldn’t expect big chances in SERPs because of that. Hmmmm…

Find out a semi-obscure page in wiki (one which doesn’t draw much attention); create a page in your site which relates to that page; add some useful content on that page; create an external link to your page (and hope your contribution is meaningful enough so that nobody pulls your link); wait six months and see if your page raises in Google.

I’ve done it. And it works.

PocketSEO on May 28th, 2007 at 5:12 pm

This isn’t an accurate experiment because Google is checking the spelling and the page already ranks high for the correct spelling.

A more accurate experiment would be to test it with a made-up word that doesn’t exist in the serps — something like “reijfsjfisajfis”.

InsideBlog on May 29th, 2007 at 1:18 am

i had a test for a while with a fresh domain (no backlinks) and put some Backlinks on it (all nofollow). some high-pr (6 and 7) also. but google doesn´t indexed the domain for a long time. but after some weeks/month it was indexed. i think its because of yahoo and msn, they get the domain and list it and so some other sites get the domain too (who get their results from yahoo and msn) and list it too and now the domain have some backlinks without no follow.
just my thoughts about that.

Ralph on May 29th, 2007 at 3:02 pm

Thanks for sharing this. It confirms what I was already thinking.

Joe Whyte on May 30th, 2007 at 1:15 am

Looks like it helps determine relevancy, not neccessarily improve ranking. I could fart this phrase to the top of the search engines. Its not hard at all.

Cary on May 31st, 2007 at 11:06 pm

Has anyone tried http://www.adzlive.com yet? It seems new. Looks interesting. They are offering webmasters(publishers) $25 for listing their site.

Posicionamiento Web on Jun 3rd, 2007 at 10:04 am

For me still working, I have good results doing NoFollow, but my results are better when I get more visits from that site.

girl on Jun 4th, 2007 at 11:58 pm

i like it

Realizzazione sito web on Jun 5th, 2007 at 7:31 am

Nice test, but nofollow don’t pass pr but pass trustrank.. is this true?

Text Link Auction on Jun 6th, 2007 at 5:34 pm

This has been my observation as well, although I’ve never performed an SEO experiment to verify it. Hopefully, this finding won’t trigger an avalanche of renewed blog comment spamming.

The Junk Market on Jun 13th, 2007 at 4:20 am

So, should i be using kewords missing their first letters?

prashant on Jun 14th, 2007 at 5:21 am

Nice information for me, I am very new in SEO.

duluoz on Jun 15th, 2007 at 4:13 am

Pagerank is a strange alchemy…

Dubai news on Jun 18th, 2007 at 2:22 am

Text Link Auction said:
“Hopefully, this finding won’t trigger an avalanche of renewed blog comment spamming.”

Nofollow is not an effective defense against comment spam. Spammers are interested in link volume. Most never check anything because they are sending out hundreds of comments per minute.

Nofollow means no reward for commenting, this has helped google deal with the volume of legitimate links in blogs. Like stopping the bank form giving out money because they may be a criminal.

iclic on Jun 20th, 2007 at 9:23 am

Interesting.

niken on Jun 21st, 2007 at 11:15 am
SEO Brasil on Jul 1st, 2007 at 8:51 am

Good and well thought testing, nothing better to make things clearer to us. I always thought so, know I’m pretty sure! Great article, thanks.

suray on Jul 3rd, 2007 at 12:27 am

I think that nofollow keyword is definitely don’t count by Google bots!!

Wahlau.NET on Jul 3rd, 2007 at 4:59 am

wow..this is a good find

James on Jul 11th, 2007 at 5:39 am
seoworkers on Jul 12th, 2007 at 6:34 am

Title: SEO Workers - Search Engine Optimization Consulting Services

Description: SEO company dedicated to furnishing the means necessary to make your web site easily indexed by
search engines, and thus available to the largest possible audience and customer base.

seo on Jul 16th, 2007 at 5:58 am

I too commented in different blogs about that issue even before that experiment. but still there is no clear evidence there are some flaws in that experiment too.

seo blog
http://seoz.wordpress.com

SEO WATCH on Jul 22nd, 2007 at 1:53 am

yeahâ those NOFollow tags!
wonder what happen if we all start using nofollow on our websites?

Have a good one.

Jane on Jul 30th, 2007 at 10:41 pm

I think it is a complete crap shoot.

Lola on Aug 5th, 2007 at 12:44 pm

I think nofollow links are not giving any PR to the site, but the anchor text is still counted. So, basically Google uses nofollow links to “know” what the site is about, but not for rating it. Therefore, i guess no site can gain PR having only nofollow links, but it’s quite possible to raise it’s positions under some keywords that are not too popular, because where the keywords are popular the high PR sites will be on top.

BabyFirstYear on Aug 5th, 2007 at 11:29 pm

I don’t know this information before. I have notice the comment follow site now. But, how about Page rank?is google will consider link from comment for page rank decision?

Marketing Man on Aug 9th, 2007 at 11:26 am

Great article, it started quite a debate.

It is a shame spamming brought about ‘No Follow’, but I guess we have to live with it.

Personally I am sure there is still some weight to No Follow links with regards to anchor text, not sure if there is anything else in terms of Page Rank etc’.

Ashish K Arora on Aug 12th, 2007 at 4:28 am

Interesting info… Is there any way we can check if a site/blog alows nofollow links?

Thanks

Tia on Aug 13th, 2007 at 9:11 am

Is there anyway I can post with a link where it isn’t going to be removed as spam?
No follow links are not being indexed =(

jazz and piano videos on Aug 27th, 2007 at 11:45 am

I think nofollow is just a filter not to pass Page Rank or Trust Rank. But google and the other engines “follow” the links.

Search Engine Optimization on Oct 5th, 2007 at 8:06 am

I think I would rather follow the no-follow ;)

Title: Search Engine Optimization

m1t0s1s on Oct 17th, 2007 at 5:12 pm

Does wordpress’ rss feed have trackback links without nofollow? In earlier wordpress versions the rss feed didn’t contain nofollow on the links, when the normal html version did.

piles on Nov 11th, 2007 at 11:42 pm

PR will come, no need to rush it If you write good copy you’ll have nothing to worry about.

Acne on Nov 29th, 2007 at 1:51 am

I have had this suspicion for a while now, that non-follow links actually count. Good post, it cleared a lot up for me. My experience was with multiple keywords, i had unintentionally created links all non-follow. To my suprise, i was ranking for those keywords within 3 weeks.

Andy

cruzibn on Nov 30th, 2007 at 2:36 am

Wonderful and informative web site.I used information from that site its great.

Your Natural Remedies on Dec 1st, 2007 at 2:35 pm

I don’t get the whole thing.

Joy on Dec 5th, 2007 at 10:43 am

I haven’t encountered no follow link tags yet..^^..lucky me!

Kumdo on Dec 9th, 2007 at 12:48 pm

Pretty nice site, wants to see much more on it!

Guitar Deals on Dec 13th, 2007 at 1:16 pm

WordPress automatically makes comments nofollows but I just have this feeling that it still counts. I’ll have to do more testing