Some of my links cause clicks on them to be recorded, and the click counts were way higher than human users could have been responsible for. I added rel="nofollow" to the links, and the counts seem to have dropped considerably. It will take a few more weeks to be sure, but it seems to have worked. Don't know that it was specifically Google that was following them, but someone was.
As explained by Tim Converse (Yahoo) and Brian White (Google) in one of the 2006 PubCon Las Vegas conferences, the name of that directive is unfortunate and deceptive.
nofollow does not actually mean "do not continue your crawling any further from here" but, rather "do not count the links included in this page as *votes* for the linked pages for the effect of link popularity".
Tim Converse "translated" nofollow in "untrusted/non-endorsed links on this page".
Therefore, the crawlers will surely follow the links included in this page.
Well, this is is what I would guess nofollow means...
Ignore this link. Do not index the url unless a followable link to it is found elsewhere.
If Google treats nofollow in a way that is significantly different to this, then they are nuts.
They indeed follow that link from that URL.
Guaranteed by the guys I mentioned above (and it's not just Google: Yahoo as well does it).
It's an unfortunate naming of that directive.
nofollow = untrusted
nofollow = "noindex what this url is linking"
I agree, it is unfortunate naming, but ever since the press release about it, they have been consistent in declaring its purpose as not counting as a vote, and I believe that all the partners maintained that they might indeed follow the links.
kaled, as a programmer, you should know by now that the name of a function may make it "obvious" what it does, but you better read its documentation to find out what it is really supposed to do.
I always assumed that Google would follow any link they find anywhere.
I did say it was a guess, and I have many times come to the conclusion that Google are nuts - I stand by that assessment.
In the context of static html, it doesn't really matter if the link is followed or not (except that Google may be stealing bandwidth) but where pages are created by php, etc. spiders could end up in infinite loops and duplicate content penalties could be (wrongly) issued.
When rel=nofollow was introduced, I seem to recall that it was compared in functionality to the robots nofollow meta tag. Of course, many people have commented that Google does not always obey either the robots.txt file or robots meta tags. Is this by design or is it a fault or is it simply misreported by observers? I don't know the answer to this, but as a programmer I am confident in my logical assessment of how Google should treat rel=nofollow.
However, given that Google is more interested in indexing everything it can find rather than everything it has permission for, I suppose calling them nuts may be wrong, perhaps they could more accurately be called liars, cheats and thieves, etc.
In answer to the original question, Google believes itself to be omnipotent, therefore it will do whatever it wants and its wants may be entirely different tomorrow or next week, etc. That may not be helpful but it's about as close to the truth as anyone is going to get.
I've noticed something interesting in Webmaster Tools that relates to this.
I have a certain page on my website. Let's say it's about cheddar cheese. Nowhere on the page itself does it mention the word "dairy" or the word "creamy".
I have two external links pointing to this page. Both have nofollow tags. The links are something like this:
(anchor:)cheddar cheese(/): it's creamy!
(anchor:)cheddar cheese(/): it's dairy!
Now, Webmaster Tools lists the words "creamy" and "dairy" as being common words in external links to my site. Not only that, but here's what's really interesting:
Google searches: (quotes for clarity - actual searches not done with quotes)
"cheddar cheese" ranks me about no.8
"dairy cheddar cheese" ranks me no.1
"creamy cheddar cheese" ranks me about no.5
once again, the words "dairy" and "creamy" do not appear anywhere on my page, only in the nofollow links. The only way Google would associate me with those terms is by getting them from the nofollow links, and it's ranking me for them.
So if for example we are doing our daily promotion and happen accross a great blog or forum of whatever. You feel you can add some value but you check the source and spot the "link condom" do you just pass it up or take a punt that Google does in fact credit your efforts... As things stand I spot the objectionable source code and would rather work with sites that share the love so I move on. Am I missing something here? As a webmaster do I realy benefit if I put rain coats on my outbound links? The logic is lost on me - If google does in fact follow the link then they see the link just like a user would see and as such should determine that it is part of the site.
justbrowsing, that is very interesting indeed - thanks for sharing.
The anecdotes so far seem to indicate that even though PR may not pass through the rel="nofollow" link, other kinds of backlink influence can. I do think that the attribute helps protect the linking website from a "linking to bad neighborhood" penalty -- which was the original idea.
|the attribute helps protect the linking website from a "linking to bad neighborhood" penalty |
or duplicate content filter/penalty?
That would be the nofollow meta tag in the <head> section, Whitey, not the rel="nofollow" attribute inside an anchor element. We already said it was unforunately named - and that potential confusion is another part of the trouble.
whatever else the "nofollow" attribute in anchor may have been originally designed to do ..I beleive it has very quickly become something which the search engines use to flag SEO on pages ..what they do to the pages that are using it then depends upon other factors ..
but only SEO uses this "nofollow" ..
mom and pop and joe sixpack have no idea it exists and so would never use it on their sites ..
it's existance on any page is a marker ..G threw out as bait to catch SEO practitioners ..( and Y caught on very quickly )..
it's like if you had to determine who were the catholic priests in a group of 6 billion people ( the population of the planet is co-incedentally near the number of pages that G claims or claimed to be indexing ) ..
you could discount all the females , all the children , all the bakers , all the tree surgeons all the etc etc ..but it's so much simpler just to spot the "dog collar" ...specially when worn in church ..
then they can look for any other common SEO traits visble in the same individual pages or site ..and other related pages that may have been touched by the same hand ( helps to identify bishops , archbishops , cardinals )use of g analytics and their sitemaps tells them you are a firm beleiver ..and that you know the prayer book ..
and we all know what G and Y and MSN think about SEO trying to manipulate their serps ..they may not penalise now ..but if you use "nofollow" ..they know who you are ..and they have long memories ..and they can and will mine their data to see what else you touch ..which other sites ..
G is an advertising agency ..run by mathematicians and psychologists with the aid of marketroids ..and which happens to do search ..
and "nofollow" was one very effective bait to get you into the open ..
I forgot to mention the PR people ( some of whom have other skills but their primary use to G is their PR skill ) that G uses to communicate with the rest of the world ..including this forum .
|but only SEO uses this "nofollow" .. |
Not quite true.
For example, wordpress comes with nofollow on comments by default.
That doesn't make the site SEOed ($DIETY no!)
It might be one more weight, but its rarely a clear indicator
*but only SEO uses this "nofollow" *
That just helped me decide between using JS and nofollow for a particular link ;-)
Mind you, the nofollow is suggested in the Webmaster Help Center, which G seems to be turning into an SEO guide for the "good guys".
It gives them one more factor to tweak - for good or evil! So from what I am reading here it's best not to fly any flags.
If you have the firefox plugins that highlight nofollow links, you would realize just how prevalent it has become.
Some sites use it on all off-site links. Others use it only on user submitted comments. Some use it on all links in the content section, but none in the site navigation.
Then you get sites like Matt Cutts' where all comments get nofollow, and he will use it on links in his content to spammer sites, but not to other sites.
I have it set up so that new users have nofollow on any content they add, but after a certain number of unmoderated posts, it removes it. I believe in trusting established contributors. Sort of like the way new users can't post their URL on WW.
There are so many indicators that I know something about SEO, that there is no point in trying to hide it. But I believe that the search engines do not want to penalize you for simply knowing what you are doing. Their goal is to keep their SERPs returning good results. They may ding you for doing bad things, but they will not ding you for knowing what you are doing.
I think there are a lot of people here who "know what they are doing" and have been hit by Google.
|But I believe that the search engines do not want to penalize you for simply knowing what you are doing. Their goal is to keep their SERPs returning good results. They may ding you for doing bad things, but they will not ding you for knowing what you are doing. |
GoogleGuy pretty much said exactly that a couple of years ago. In fact, I think he went as far as to say that, within reason, SEO can be a good thing because it makes the search engine's job easier!
This is covered in the Google help pages for those interested...
Anyone interested in matt cutts newest comments on this.....
"The specific guideline for doing paid reviews or paid links is to make sure that your links don't affect search engines. You can do that via a nofollow attribute on your links, or an internal redirect which goes through a page which is robot.txt'ed out, or several other methods."
He left that as a comment in Danny Sullivans searchengine land
[edited by: encyclo at 2:10 am (utc) on Dec. 31, 2006]
|The specific guideline for doing paid reviews or paid links is to make sure that your links don't affect search engines. You can do that via a nofollow attribute on your links, or an internal redirect which goes through a page which is robot.txt'ed out, or several other methods. |
Actually this raised a question for me which I couldn't find answered in any of the threads. I understand that rel=nofollow is intended for certain external links and should not be used for other purposes (specifically not for *internal* links, and not even for keeping the SE's away from duplicate content).
On the other hand, the nofollow attribute is recommended for paid links. All fine and dandy. But what about paid redirect links, like www.mysite.com/adscript.php?redirect=advertisersite.com? Judging from the quote above, this should be a situation where rel=nofollow is endorsed. But strictly speaking, it is also an internal link, so from what I have read so far about using rel=nofollow on internal links, I should be scared using it in this context. Any thoughts on this?