Jump to content

Cre8asiteforums Internet Marketing
and Conversion Web Design


Photo

Blogs, WordPress and the nofollow attribute


  • Please log in to reply
7 replies to this topic

#1 DianeV

DianeV

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 7216 posts

Posted 26 January 2005 - 05:54 PM

Discussion of the new nofollow link attribute arose in the midst of the No sleep 'till DMOZ thread. Since it was off-topic, I've started this thread.

Bill asked about the nofollow attribute; I posted:

As I recall, Google also introduced a nocache tag ... but didn't sites employing the tag get booted from Google?

So. I'll preface this with saying that this is totally my thought, not even a theory but just a concern.  But I'd be concerned that sites employing nofollow might suffer a similar fate: once identified as a page or site employing nofollow -- that is, identified by the site owner as containing untrusted links -- it's possible that the results would be more than just not following and/or giving credit to outbound links.

I've blogged a bit about this, and got into something of an argument with a troll with a vested interest posting anonymously over at the WordPress forums. I can say it clearer here, where there is an understanding regarding PageRank versus link pop versus anything else:

Fact is that Google doesn't "like" link farms. They've been made something of a laughingstock over Googlebombs.  And while blogs are not link farms per se, some of the more popular ones with lots of posters link back and forth and all over the place, giving them all kinds of ranking power based on links alone.  Heck, Anil Dash (SixApart) won a prize in the nigritude ultramarine SEO contest by getting zillions of his closest friends to link.  Think Google likes this?

WordPress is installing nofollow by default without a toggle button to turn it off.  I won't be using that version.


and Stock asked:

Does the nofollow tag only kick in in comments? Or in the main blog post, too? That's the key - the blog posts themselves should be followable, but the links in the comments, should not.


to which I reply here:

~~~~~~~~~~~~~~~


G, that's what one can surmise from the amazing amount of assumption at the WordPress forums: nofollow only in the comments. Not in the rest of the site. All I asked for was an on-off toggle, or a way to undo nofollow if I so chose.

The whole link-pop-in-the-algo thing is a double-edged sword for search engines. On the one hand, links to sites were supposed to be "votes for" rather than just citations of material being discussed. I think when you assume that a citation is a "vote for" or any type of approval, you've assumed far too much.

Example: let's say that in an article/discussion/post you link to a page as an example of the wrong way to do things, poor ideas, etc. That's not a "vote for" although it is a citation.

However, for the sake of argument, let's say that 'way back when there were only X number (however large a number) of news sites, discussion sites, commentary sites, etc. So far so good. With the advent of blogs, however, and free or cheap software, we have a gazillion sites pointing links to anything and everything, and comments and trackbacks doing the same ... which only magnifies the "vote for" assumption. Exponentially. 'Cause, gosh, those bloggers sure like to chit-chat and link to each other. Can't even call it a link farm, or the entire blogosphere might go ballistic and it would probably hit the evening television news.

Heck, if I owned a search engine and I employed link pop, I'd be dying right about now for a simple method of identifying those types of pages and links. I might not even penalize pages employing that method; of course, it would depend upon how much effort/time/expense it cost me or saved me. And on my Good Mood on any given day.

Of course, I might just be clever enough to present it as a "benefit" for those who would use it. Say, to stop comment spam. Although, in that regard, I expect it to be about as effective as anti-virus and anti-spam programs have been in stopping email spam.

Call me wrong. Call me cautious. Or call me bored. <grin> Of course, I was the one who was pooh-poohed for uttering a caution over the BuddyLinks thing 'way back when. And this is going about the same way.

Look: if I link to, for example, a post on Kim's blog, that is both a citation and a "vote for" -- and I want to be able to give Kim the benefit of that link, since the availability of such benefit has been given out to one and all by search engines employing link pop of any kind. It's a realistic view, unless we're being asked to "pretend" it doesn't exist.

I wouldn't mind an attribute that said "hey, I'm linking to this but I don't mean it as a 'vote for'" ... but I'm cautious enough to want to know how that attribute will play out in the long run. And, just as I don't want to use Microsoft-only tags, I'm leery of using an Google/Yahoo/MSN-only attribute.

But, since the real issue, for bloggers, is stopping the sometimes avalanche of bot-driven comment and trackback spam, I think I'll continue to employ methods or to find ways to stop that.

#2 BillSlawski

BillSlawski

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 15644 posts

Posted 27 January 2005 - 01:08 AM

Does the nofollow tag only kick in in comments? Or in the main blog post, too? That's the key - the blog posts themselves should be followable, but the links in the comments, should not.


Looking at it carefully, the new value that has been defined is tied to the use of a link on a page, anywhere on the page.

The nofollow value is one potential value of the "rel" attribute when it is used for the anchor element.

According to the 4.01 specifications for html, the "rel" element is supposed to define the relationship of the present page with the the page pointed to by the link with the href element.

So, in the example <a href="http://www.example.com" rel="nofollow">, the relationship defined between the page the link appears upon, and the page it points to is set to "nofollow".

Logically, the "nofollow" value is questionable in that it doesn't really define a relationship, but rather gives an instruction to an indexing program. But, getting past that, there's nothing that says that a link with that element set to that value has to appear on a certain part of a page, such as the comments section.

I don't see how a spider would know whether the nofollow instruction was in a comment or was in part of a page written by the author of a blog. Especially on pages where comments appear on the same page as the blog post.

The nofollow value could easily be used on pages other than blogs, too. Nothing in the definition of the new value shouts out, "only do this if this page belongs to a blog, and this is in the comments section, and it is written by someone other than one of the writers of the blog."

Is it possible that is something that the search engines adopting the use of this tag will be capable of programming into their indexing software? I don't know.

Will the search engines ignore the nofollow values that have been started to be used by the people at wikipedia for their outside links? Will others start using "nofollow" in other parts of their pages?

The whole link-pop-in-the-algo thing is a double-edged sword for search engines. On the one hand, links to sites were supposed to be "votes for" rather than just citations of material being discussed. I think when you assume that a citation is a "vote for" or any type of approval, you've assumed far too much.


I agree with you.

#3 bwelford

bwelford

    Peacekeeper Administrator

  • Site Administrators
  • 9018 posts

Posted 27 January 2005 - 06:24 AM

In other words, Bill, you're saying that the whole use of PageRank and backlinks by Google in its algorithm is invalid. If you're talking straight numbers, then I'm inclined to agree with you. But it's a huge shocker.

Perhaps by including some stuff via link text and semantic analysis, you can make it a bit more valid, but I'm not sure on that either. :?

#4 Grumpus

Grumpus

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 6327 posts

Posted 27 January 2005 - 08:02 AM

This is the stupidest thing I've ever heard. (Well, no, I have a lot of stupid things - for example, the link exchange request I got in my e-mail this morning asking for an exchange between a site and our Cre8asite Directory.) But, it's definitely stupid.

If every link on a page is going to have the "nofollow" attribute on it, then you are doing nothing but wasting bandwidth. There's a meta tag that keeps spiders from following every link on a page.

This new attribute is for people to take some links on a page and stop the spiders from using it. For example, our "login" links etc. would be good links to have that attribute applied to. And, we could replace our redirect script and all the other rewriting code we've added over the years with a simple line of code in the bbCode parser, here, so that links in posts wouldn't be followed.

It seems to me that, unless the folks at WordPress are complete idiots, that what they are currently offering is merely a quick patch for folks who are being flooded with comment spam. True, the patch won't do much good, but it probably is a good customer service move and it's at least a good marketing ploy to be able to say that you were the first blog to add support for the nofollow attribute.

I can't imagine that this is anything more than a quick fix. Wait for another software update and they must surely be planning to implement a properly (and usefully) functioning version of the tag. They just have to be. I can't believe that people are really that stupid to just add this as they have and leave it this way. It's just not what the attribute was intended for.

G.

#5 DianeV

DianeV

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 7216 posts

Posted 27 January 2005 - 11:50 AM

Alright, I'm lost.

I don't see where Bill implied that "the whole use of PageRank and backlinks by Google in its algorithm is invalid".

And Stock, I'm not sure what you're referring to. At the moment, I've been able to find little information at WordPress beyond that they're going to be utilizing the tag. One assumes that it is going to be applied to links in comments (and maybe trackbacks) only.

#6 bwelford

bwelford

    Peacekeeper Administrator

  • Site Administrators
  • 9018 posts

Posted 27 January 2005 - 12:06 PM

I inferred that message re back-links invalidity from the Quote at the end of Bill's post, to which he said he agreed. Even if he didn't fully agree, I've been thinking that way for some time. I believe that Google loses more in relevancy by using the volume of backlinks in their algorithm than it wins, however they reweight the volume. Hence my comment.

#7 DianeV

DianeV

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 7216 posts

Posted 27 January 2005 - 01:37 PM

Ah, I understand. Well, since that quote was from me, I agree. :-)

It's unfortunate that what started out as a good idea doesn't scale well given the lack of sufficient parameters. PageRank has always been weighted towards older sites (and sites apparently "seeded" with high PR), and links are ... just links. Of which older authority-type sites would generally have more, until the link exchanges, etc. come into the picture.

I don't think that putting the nofollow attribute in the hands of webmasters will change this scenario much. I think Google, Yahoo, et al. need to think further upon the link portions of their algos, and possibly to expand and refine them.

Easy to say, isn't it? Probably not so easy to do.

However, if what they want is to rank pages independently of whatever webmasters do, then relying upon an attribute the use of which is entirely controlled by webmasters would not seem to be a long-term solution.

#8 Ruud

Ruud

    Hall of Fame

  • Hall Of Fame
  • 4887 posts

Posted 28 January 2005 - 12:41 PM

Ammon elsewhere raised the valid question whether or not the use of this tag will prevent PR from passing on.

If not then the tag fails as an anti-spam measure.

If, as one would suspect, it would then I wonder how Google et al will deal with possible abuse of the tag for PR harvesting. One wouldn't need a ?out, CGI or jump link page: simply use this tag everywhere. Unlike the overuse/abuse of a jump tag Google would have a hard time punishing as they suggest using the tag. Nor can they simply follow the links anyway and deliver PR, as they suggest they have done with the PR harvesting pages, as this would defeat the whole purpose of the tag.....



RSS Feed

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users