Using Rel="nofollow" To Make Other Links Stronger, Can I use nofollow to sacrifice links? Options

[JeremyH]

I’ve created a web page with about 20 links to .pdf files and a link to another website’s index page.

The .pdf files are graphical and not text based, and I would prefer people to access the .pdf files through my website and not directly. Therefore .pdf ranking is not terrible important to me. However, I do wish the link to the other website to be as meaningful as I can make it.

In my situation, would it make the most sense for me to place a rel="nofollow" attribute in all my .pdf links? That way, Google and other search engines will see only one link on my website, making that one link stronger.

[Scottie]

If it were me, I'd just block the .pdfs using robots.txt.

CODE

User-agent: *
Disallow: /*.pdf

[Jill]

Ditto what Scottie said.

The no follow attribute won't help for your situation, it's purpose is completely different. (It's to stop blog spam, basically.)

[mcanerin]

Actually, the previous robots.txt example won't work, even though it makes perfect sense and *should* work if the creators of the standard had any (IMG:http://www.highrankings.com/forum/style_emoticons/default/censored.gif) sense themselves.

You ARE allowed to use a wildcard "*" character in the "User-Agent" line, but you are NOT allowed to use it in the "Disallow" line. I don't know what was going through the minds of those that wrote this spec when they decided to do that, but that's how it works, much to my frustration.

I recently wrote a robots.txt generator and got quite the introduction to the strengths and weaknesses of robots.txt while researching and coding it.

The best way to disallow PDF's under the outlined circumstance would be to put them all into a directory (ie www.domain.com/pdf/) and then disallow that directory:

CODE

User-agent: *
Disallow: /pdf/

Although technically the rel=nofollow attribute could work, it's clear that at this time the search engines are all handling it differently (if at all) and it can't be relied on right now.

You could also use obfuscated javascript code to create the links to the PDF's and accomplish much the same thing, though some search engines have shown a surprising tenacity in actually figuring out how to follow those, as well.

They all agree that they respect the robots.txt though, so that's currently your safest bet.

Ian

[JeremyH]

Thank you everyone for your help.

Luckily, I had all my .pdf files in a folder anyways, and I’ve been meaning to write my robots.txt generator, so it all worked out.

I was surprised to hear that my idea would not work. Sure “no follow’s” initial purpose was to block spam links, but it prevented spam links by taking the value out of having those links. The search engines did not “see” those links, so they did not follow them or count them as back links. If they can’t see all those links, but they can see THE ONE link-the one I want to matter most, then I thought that would make that link that much more powerful. To be the only link on a website with PR X.

[mcanerin]

I was surprised to hear that my idea would not work. Sure “no follow’s” initial purpose was to block spam links, but it prevented spam links by taking the value out of having those links. The search engines did not “see” those links, so they did not follow them or count them as back links. If they can’t see all those links, but they can see THE ONE link-the one I want to matter most, then I thought that would make that link that much more powerful. To be the only link on a website with PR X.

That's the theory and that's how it was initially presented when it was introduced, however, at the recent NYC SES the search engine reps were questioned about it in depth and the resulting information was decidedly lackluster.

It's better than nothing, but right now I would not decide on any strategy revolving around the attribute, lets put it that way. Once the dust settles, you may be able to rely on it, but for now apparently some engines DO follow, but simply treat the links as "untrusted", some ignore the tag totally, and others don't appear to know what they are planning on doing with it, only that they will take it into account in some manner.

I suspect part of the issue is that a spammer can just as easily use this tag to attempt to hide their tracks as an antispammer could to try to combat spammers. To totally and blindly ignore the context of the nofollow tag would ignore the fact that many people choosing to use it may not have the purest of motives for doing so.

That being the case, since the SE's can't rely on it - neither can users.

Ian

[Alan Perkins]

Sure “no follow’s” initial purpose was to block spam links, but it prevented spam links by taking the value out of having those links.  The search engines did not “see” those links, so they did not follow them or count them as back links.

The idea behind nofollow was not to prevent links being followed, but to prevent any weight being attributed to the links. Links may still be followed.

If they can’t see all those links, but they can see THE ONE link-the one I want to matter most, then I thought that would make that link that much more powerful.  To be the only link on a website with PR X.

You can't second guess how the links will be treated to any degree of accuracy. It may frequently change, anyway.

robots.txt provides the best solution. If the PDFs are all protected by robots.txt, the links to them will be left dangling. According to the literature, dangling links are removed from PageRank calculations. So this probably achieves what you were trying to achieve. (IMG:http://www.highrankings.com/forum/style_emoticons/default/smile.gif)

[Webnauts]

Add the rel="nofollow" tag to all the links that point to pages that are not important for ranking and traffic generation.

Another use is to keep a link from going out to another site and giving away PageRank to other sites, or referring to a site that you don't necessarily need to be helping with their popularity.

[Betty]

Add the rel="nofollow" tag to all the links that point to pages that are not important for ranking and traffic generation. 

Another use is to keep a link from going out to another site and giving away PageRank to other sites, or referring to a site that you don't necessarily need to be helping with their popularity.

How does it "give away" PageRank? Besides Blog spam, I don't understand why you wouldn't want to pass along your page rank.

[Jill]

There's no reason, justagrrl.

Webnauts is not providing very good advice, in my opinion.

[rustybrick]

The idea behind nofollow was not to prevent links being followed, but to prevent any weight being attributed to the links.  Links may still be followed.

Well, actually, I believe Google said that they will simply not follow those links at all, if the nofollow is used. Yahoo! did not comment on this. This is from the SES NYC Indexing Summit session.