No Follow: Do you Follow?

[Chris Boggs]

Friday's SEM Crossfire discusses the Great No Follow Debate of '08 (or at least one of the first ones). This thread will serve as the discussion area for that post.

[AussieWebmaster]

worth a read

[Chris Boggs]

Thanks aussie...I hope we get some conflicting or supporting opinions on this subject, and find out what I missed.

[jimbeetle]

The biggest problem with nofollow is that it's an attribute without a standard. Nobody knows exactly how the search engines treat it and I'm not sure if the folks at the SEs themselves can pin it down. For every Google "This is how we use it...", folks can find a boatload of anomalies that appear to contradict those statements. As for Yahoo and others, well, we're flying totally blind.

Now, as for that unknown Google engineer, I have to preface what I'm going to say next. Matt's a good guy. Over the years he's helped countless web folks make their sites better and more successful. He helped me some years ago by getting a long-term ban on one site lifted. I've watched him at conferences patiently answering questions for hours. He's directly responsible for many, if not all, of Google's outreach to the webmaster community.

But in the nofollow brouhaha I strongly believe he's overstepped his bounds.

I simply cannot understand how one "unknown" engineer, spamfighter though he might be, was able to single-handedly extend a non-standard attribute to even more non-standard uses. From, "this is a user-added link that I can't vouch for," to "I like the way you used nofollow to control the flow of PageRank" (at ?Boston 2006? PubCon), on to "If you want to sell a link, you should at least provide machine-readable disclosure for paid links by making your link in a way that doesn’t affect search engines", and throwing in a reference to the FTC for good measure.

Using nofollow to indicate a paid link is ridiculous on its face. Come on googlebot, tell me, is this nofollowed link a paid advertising link? Or am I using nofollow to sculpt PR? So, where's the disclosure?

And using nofollow (or noindex) to sculpt PR is very dangerous in the wrong hands. Over the past few months I've had to answer a handful of the site dropping in Google inquiries with "Because you shot yourself in the foot."

Mayhap it's time for one of Danny's search engine sitdowns to get this mess straightened out.

Just my 2 cents. ;-)

[Chris Boggs]

great post Jim thanks for the reply. You bring up a great point that I was trying to get across: there is no standard.

Hopefully you would think that Google is smart enough to come up with a rudimentary way for its algo to tell the difference between a link no-followed to ID paid versus one sculpting PR. Just the fact that it is likely going to be an internal link for the latter should make it simple.

The idea of nofollow "in the wrong hands" is very akin to robots.txt. My friend and former co-worker Tobias Klauder always said that when using a ninja sword you have to be skilled, or the results can be messy.

[Jazajay]

Seriously I never get people who don't use the proper tags. If you want to "sculpt" PR use nofollow meta. That way it'll work in the other SE to a degree.

MSN and Yahoo treat it as G treats robots.txt excluded pages
http://www.mattcutts.com/blog/handli...dex-meta-tags/

Actual Jim according to a blog post here at SEW
Google, Yahoo, Msn Unite on the support for nofollow attribute for links

http://blog.searchenginewatch.com/blog/050118-204728

But heres 2 if not 3 reasons why you should dump the rel=nofollow tag on anything but blogs/user generate content.

1. Code bloat.
Say you have 1000 pages you use the rel=nofollow on 10 Support links / page. That = 14 X 10 x 1000 = 140 000 extra characters that the SE has to index, what a waste of time. You want your site indexed as quick as possible G,Y,L,A wont hang around for ever - according to
http://www.beussery.com/blog/
Google uses the same amount of electiricty as a us state-

....electricity used to power the state of Mississippi in 2005.

That blog should allow anchor links to the right blog post. - lol -
But with that much electricity usage it's main goal would be to cut down it's electricity bill how to do that indexing quickly / quoter/site.

Also as Jim said support may be sketchy that blog post, not Bue's, was not definite from what I SKIMMED. So by using it you may not being doing your self favour in the others.

Also <meta name="robots" content="noindex, nofollow" />
Is supported granted it will give "thin" results in yahoo and Live.

Thin results so there is no confusion is a SERP like this

www.example.com
www.example.com Cache page

No description is return just the page title and URL.
Ha I wish I had some one explain that to me in another thread.

2. future Algo changes and all your pages dropping like a stone
So if we take out code bloat which most sites have more than a thousand pages, as you eluded to it may be suceptable to a algo change.

......search engine algorithms updates from now, it will likely have far different value than it holds today

Most sites then would have a huge job changing the tags if the change was negative.

Ok granted it does at the moment "sculpt" PR to lower pages.
IE if you have a 3 tier site
home
category
product
By placing no nofollow on the home page.
category Nofollow on other categories thus "Pushing" it down to the product pages
Product. No follow on everything but it's category does seem to work.

Do I use that technique
No as I know how to do good SEO link building to deep pages,
get pages indexed quickly,
appropriate use of 301's,
good onpage,
good anchor text links.
etc...
etc...
And a site with that blatant use of nofollow would be easy to detect in the future. - future could be as early as tomorrow as SE very rarely give you prior warning to whats going to change - and have 90% of your pages penalized but worse all the pages you want in the SERPS your category and product pages. Hum... if you are an employed SEO/SEM for a big corp and that happened do you think you would work in the industry again?

It's not hard to learn just look at MY posts here at SEW. lol

Any way like always nice article Chris.

Jaza
I'm still thinking about that title change - lol

[jimbeetle]

Yeah, Jaza, I know they united on the word "nofollow," but as evinced by Google's extensions of its meaning there is no standard and we have absolutely no idea exactly how they're implementing it behind the scenes. Do they all treat it the same way, which as a standard they should. We simply do not know.

[Jazajay]

I know I totally agree, I like the title of that diary entry.
Also if they decided on a name it probably means they do take recognition of it, to what degree through..... I totally agree with you.

I wasn't having ago or discrediting you and if fact I did say:

Also as Jim said support may be sketchy....

There you go Jim's just given rule number 3 as why not to use them for anything rather then their intended purpose IMO - blogs / user added content.

jaza

[Webnauts]

- The rel="nofollow" is an attribute, and not a tag.

- I would rather use meta tags or robots.txt if I do not want a page to be followed, and for links I rather would use a server side temporary redirect (302).

- About excluding pages like "Privacy Policy", "TOS", e.t.c just to avoid a PageRank dilution, I do not think that it is a good practice in terms of a site's credibility. I am not sure if that factor might be at some point considered by search engines, but still I would not disallow those pages as they are equally important like all other pages. If they were not, why should I have them there anyway?

[jimbeetle]

About excluding pages like "Privacy Policy", "TOS", e.t.c just to avoid a PageRank dilution, I do not think that it is a good practice in terms of a site's credibility.

Yeah, it especially hurts in a couple of situations I've seen ("What happened to my pages?") where they were using Google for their site search function.

Again, very dangerous in the wrong hands.

[Jazajay]

The rel="nofollow" is an attribute, and not a tag.

Nice catch a honest typo.

They are important for users yes, and that is arguable, my terms and conditions and privacy policy are hardly viewed and in fact if I deleted them but left the link up I would track very few broken links, but not SE I mean who seaches for terms and conditions?
or "I wonder who has the best privacy poilicy".

Apart from people stealing them for themselves no one. So why have them in the results it just means the SE's take longer to get through your site as they are indexing useless pages.

...Google for their site search function.

Fair enough but why use G search anyway unless you are after high abondoment rate? or is it the extra few cents?

Do it you self and you can build it a lot better I know from experince - My PHP's good but I can clear a room if you get me started about IA (with bordom) - Information Arcutecture - as always it's not that hard.

Again, very dangerous in the wrong hands.

Thats why you need hands like mine - lol

At the end of the day that tag - meta noindex - was invented to get rid of pages you don't want in the SERPS, and I dont want them in the SERPS.

Jaza

[AussieWebmaster]

credibility? the link works but do not waste link bleed

[Jazajay]

Originally Posted by AussieWebmaster

credibility? the link works but do not waste link bleed

Sorry aussie I'm being totally thick for a change can you elaborate?

[AussieWebmaster]

sending link juice to wasted pages that are not needed to be added to the index in any significant way

[Jazajay]

Yeah sorry I knew it was just me being thick.
Must be your Austrialian accent.

[Chris Boggs]

FYI just to set the record straight on the use of "Tag" versus "attribute," from an email exchange I had:

I know that nofollow is an attribute. In fact at (more than) one point I refer to it as such (Not surprisingly, the "link condom" attribute -- much like the occasional cheap…)

As you probably know, the alt is also an attribute frequently referred to as a tag “alt tag.” I used to rail against people using the word tag instead of attribute (you can probably find posts I have made saying exactly that) but I finally ended up joining them instead of fighting them.

thanks everyone for the continued conversation maybe among us we'll come up with an entire new use of the thing.

[beu]

Sorry I missed this one, I've been under the weather the past few days but I'm feeling a little better today!

Before being uncovered by Jazajay, I was conducting my own experiment nofollowing ONLY user created content and trying out a plugin Matt suggests.


Here is why I'm experimenting:
- attributes without standards make me nervous

Question I'd like to answer: Why would they use an attribute with no standard use and how are they using it onsite vs. offsite? Are they being used the same?



- nofollow was initially to prevent comment spam.

- nofollow was adapted to prevent paid links passing PageRank and comment spam.

- nofollow was then adapted to prevent paid links passing PR, comment spam and direct PR in a site.

Question I'd like to answer: If you nofollow your own blog index page how do engines find content within your blog other than inbound links possibly from other nofollowed links because xml sitemaps aren't supposed to replace how Googlebot crawls the web?




- Google uses "I can't vouch for this link" in their example.
http://scholar.google.com/webmasters/bot.html

Question I'd like to answer: Why would a webmaster nofollow or not vouch for his/her own link?



- the first page to ever use a nofollow tag was
http://www.cs.unc.edu/~cutts/
but for some reason there are no nofollows at:
http://www.cs.unc.edu/~cutts/face/index.html

Question I'd like to answer: Why doesn't the second URL use nofollows?



- Mattcutts.com employs very liberal use of "nofollow"

No description is return just the page title and URL.
Ha I wish I had some one explain that to me in another thread.

- Mattcutts.com has no description and "no odp" it's homepage

- Google's blogs employ very general use of "nofollow"

Question I'd like to answer: Why not, what are they doing?



Bottom line, like everyone else I'm trying to find some constant, I'd just like to find it first!


Also worth pointing out, a very well know Seattle based SEO company nofollows their "contact us" page and as a result they barely rank top 20 for "seattle seo". I think a lot of thought should go into nofollowing links but hey that's just me.

[Jazajay]

Want to change Google and how they use the meta noindex?
http://forums.searchenginewatch.com/...277#post125277

Vote now and we could change how it is used by Google.

Jaza