homepage Welcome to WebmasterWorld Guest from 54.227.20.250
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google SEO News and Discussion
Forum Library, Charter, Moderators: Robert Charlton & aakk9999 & brotherhood of lan & goodroi

Google SEO News and Discussion Forum

This 35 message thread spans 2 pages: 35 ( [1] 2 > >     
Nofollow on internal links - reasons to be careful?
santapaws

5+ Year Member



 
Msg#: 3694262 posted 11:51 am on Jul 9, 2008 (gmt 0)

when some guys first started using nofollow on internal links there were reports that for some it had coincided with a drop in rankings. I wonder whats the current thinking for using this tag on unimportant internal links like privacy statements.

 

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 12:24 pm on Jul 9, 2008 (gmt 0)

I wonder whats the current thinking for using this tag on unimportant internal links like privacy statements.

I just drop a meta robots tag on those pages and be done with it...

<meta name="robots" content="none">

...or the longhand version...

<meta name="robots" content="noindex, nofollow">

That nofollow attribute has never "felt right" to me and I've never used it, ever. If I have to think about nofollowing anything, then something isn't right within my structure. ;)

Also, what happens when your content is scraped and regurgitated elsewhere minus your nofollow attributes? Happens every single day and I would think that just negates whatever nofollowing one may be doing. Same goes for my method but it adds another layer of "basic" prevention that sits behind what can be scraped "visibly".

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 12:46 pm on Jul 9, 2008 (gmt 0)

If we take Google at their word, then the rel="nofollow" attribute is simply not followed at all, not even for url discovery. That means the target url cannot play a part in circulating PageRank through the site. So indiscriminate use of the tag might choke off parts of the site from their rightful prominence.

in contrast, the robots meta tag will NOT stop the url from accumulating PageRank. If you use "noindex,follow" as the attribute value, then the page will still help to circulate PageRank, too.

I've been with pageoneresults on the internal rel="nofollow" issue for quite a while - it just felt wrong to me, so I didn't use it. However, I have been hearing about some successes from various SEO folks, so I'm curious. Googlers like Matt Cutts keep talking about it in a cautious but postive light, and that heightens my curiosity.

I'm currently working with one team to develop a large site High 6 figuires of urls at launch) thats due to go live in early fall. There are some flows in the site wireframe that look to me to be wasting PR circulation in a major way, especially since there are so many deep pages that we want to get cooking.

So I'm thinking of launching the site with a few rel="nofollow" attributes just to see what happens. For one thing, those urls will not be available through any other links anywhere, so I'll get to test if Google indexes those urls or not, even as url-only. If anything interesting shows up, I'll be sire to report on it.

[edited by: tedster at 1:31 pm (utc) on July 9, 2008]

Lord Majestic

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 12:51 pm on Jul 9, 2008 (gmt 0)

nofollow in META and tag have 2 very different purposes - the first one is to actually prevent bots from following those links (ie crawling), where as the second does not hold such prohibitions and used to reduce value of backlinks for PR-like calculations.

Consequently usage of nofollow on internal links shows a very different intention from nofollow in META, if anything this should act as a red flag for an SE to do deeper analysis and if lots of such internal links are marked with rel=nofollow then clearly some kind of SEO games are going on, so the question is whether a particular SE likes being gamed like this or not.

rainborick

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 1:24 pm on Jul 9, 2008 (gmt 0)

Heretic that I am, I usually take Googlers' word when they make public statements. I never sold advertising on my sites and had only one blog to worry about, so from that aspect 'nofollow' was a non-issue. But when Matt essentially said it would be acceptable to use nofollow for PageRank sculpting, I tried it on the main page of one of my most valuable sites where some internal pages were under-performing to add some oomph to the links to those pages. I nofollowed the links to pages like the About Us, Contact Us, Terms of Use, etc. whose rankings were immaterial anyway. My results were very positive. I got a nice little uptick in the rankings for those internal pages over the next couple of weeks. Being a lazy clod, I hadn't done any link building for this site for a couple of years, and it has only rarely attracted links on its own in the 8+ years of its existence. So as a test, I think it bodes well for both the safety and utility of nofollow.

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 1:36 pm on Jul 9, 2008 (gmt 0)

One of my motives for experimenting with rel="nofollow" is that I don;t like seeing "Contact Us" as a Sitelink. In almost every case, I would much rather see something else!

When I speak of indiscriminate use, I'm thinking of some cases I know of where the main menu links were nofollowed, trying to drive more traffic to the sub-category pages. At least one case resulted in completely hosing the search traffic for the site. Some of their key anchor text was taken right out of the algo calculations!

But using rel="nofollow" for those utility pages is not nearly such a risk I think, unless you like seeing them in your Sitelinks ;)

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 1:52 pm on Jul 9, 2008 (gmt 0)

Google Sitemaps allows you to block links from showing up in Sitelinks.

I only use rel="nofollow" for url drops in forums and for spider traps. On "Contact Us" type pages, I use the meta noindex,nofollow tag and never had a one show up in serps, Sitelinks or otherwise.

[edited by: Key_Master at 1:53 pm (utc) on July 9, 2008]

santapaws

5+ Year Member



 
Msg#: 3694262 posted 2:42 pm on Jul 9, 2008 (gmt 0)

why would you make it nofollow in the meta?, just because you dont want the contact us page indexed doesnt mean you don't want the links on that page followed.

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 2:51 pm on Jul 9, 2008 (gmt 0)

why would you make it nofollow in the meta?, just because you dont want the contact us page indexed doesnt mean you don't want the links on that page followed.

Ya know, that's actually a good question. My idea was to remove the page entirely from the graph as it serves no real purpose in the overall scheme of things. I'd rather have the bot spend its time following links on other pages that are of higher importance.

But, I do understand where you are coming from, you could just noindex. Mind you, these are the "poor man's" method of having some control over indexing routines. There are of course other more efficient ways to do this.

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 3:00 pm on Jul 9, 2008 (gmt 0)

Google Sitemaps allows you to block links from showing up in Sitelinks.

True, but you can't choose a replacement, you can only delete them.

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 3:05 pm on Jul 9, 2008 (gmt 0)

In my case, my service type pages contain links to other terms related/contact pages that I wouldn't want followed or indexed.

Besides, with some rare exceptions (ie, sitemaps), I think that if the content on the page isn't fit to be indexed, why should the links? From a programmers prospective, it just seems like a fishy red flag to me.

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 3:21 pm on Jul 9, 2008 (gmt 0)

True, but you can't choose a replacement, you can only delete them.

Maybe Google selects another replacement link. Don't know never had to block one.

I would think Sitelinks don't show up very often in searches or for that matter even clicked on often enough to justify retooling all your menu links with rel="nofollow", when a single meta tag on your "Contact Us" page would have the same effect.

Just my opinion. :)

rocco

10+ Year Member



 
Msg#: 3694262 posted 3:49 pm on Jul 9, 2008 (gmt 0)

I have started to use rel=nofollow on internal links and do not have anything bad to tell yet. I use it 1000-fold on pages with additional info eg some pages of mine provide a stats page for every page with just a flash graph, or a contact form with the id of the current page. It would make no sense at all giving these thousands of page to the googlebot then these pages are sort of stubs and getting them served from the serps would make no sense at all (in contrast to some about or contact pages).

jimbeetle

WebmasterWorld Senior Member jimbeetle us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 3:50 pm on Jul 9, 2008 (gmt 0)

On "Contact Us" type pages, I use the meta noindex,nofollow tag and never had a one show up in serps

Noindexing these type of overhead pages is something I'm very hesitant to do as I don't know what signals are used in part to establish "trust." Might Google or other SEs look to the presence of contact or privacy pages as such a signal? I don't know, but as long as they could, I want them to know that they're there.

tedster

WebmasterWorld Senior Member tedster us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 4:02 pm on Jul 9, 2008 (gmt 0)

One approach I've used, instead of nofollow, is to place all the site's "utility" links in an iframe. That way, the urls are in the index, but there's only one url collecting and circulating PageRank, rather than having those links appear somewhere in every url on the website. This seems to me a good resolution, and I have seen no downside. It's just that the page layout needs to accomodate this arrangement.

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 4:29 pm on Jul 9, 2008 (gmt 0)

Noindexing these type of overhead pages is something I'm very hesitant to do as I don't know what signals are used in part to establish "trust." Might Google or other SEs look to the presence of contact or privacy pages as such a signal? I don't know, but as long as they could, I want them to know that they're there.

I have a p3p privacy policy that Google and other search engines do crawl (and probably understand better than the made for humans version). Even without one, I doubt there would be a trust issue. They know the pages are there and they do crawl them- I'm just politely asking them not to include them in search results. That doesn't mean they can't be used for internal purposes.

ichthyous

10+ Year Member



 
Msg#: 3694262 posted 10:15 pm on Jul 9, 2008 (gmt 0)

I made a big mistake last year and nofollowed quite a few of my internal pages...it really hurt my site when the PR of those pages vanished. I removed all of them except for contact us and search pages and everything returned to normal within two weeks or so. I would use nofollow tags very sparingly and only on things that really don't have any inbound links or pass much PR to other portions of your site.

pageoneresults

WebmasterWorld Senior Member pageoneresults us a WebmasterWorld Top Contributor of All Time 10+ Year Member



 
Msg#: 3694262 posted 11:44 pm on Jul 9, 2008 (gmt 0)

If we take Google at their word, then the rel="nofollow" attribute is simply not followed at all, not even for url discovery.

Ya see, that is the part I could never really understand. If a scraper gets your content and then regurgitates it without all the nofollow stuff, what does that do?

I shouldn't knock the nofollow, all three of the majors worked on bringing that one to market, kudos. You probably won't find many "old, old school" SEOs using those things. If you are a publishing platform, then by all means, nofollow all of the comment links until such time that you've verified their quality and that "you trust them". That is what nofollow was meant for, to combat comment spam, bottom line.

The original protocol got somewhat warped shortly after its release and became a micro-tool for managing links. I'm sure Google saw this huge change in the link graph and they said to themselves, "hmmm, now looky there would ya, I do believe we have someone attemptin' to manipulate our algo. Would ya look at em all Pa!?"

nofollow was an "easy" way for Google to find what they were looking for. It served its purpose originally and still serves its purpose today. nofollow = micro-link manipulation. A clear signal in many instances. ;)

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 12:28 am on Jul 10, 2008 (gmt 0)

Ya see, that is the part I could never really understand. If a scraper gets your content and then regurgitates it without all the nofollow stuff, what does that do?

Google will follow the link. So if you're going to use rel="nofollow" for internal links, don't rely on it as a tool to prevent discovery of pages you don't want crawled. Somebody else can link to those pages and they will be crawled. It's best to cover your bases with a meta "noindex,nofollow" tag on those pages.

Lord Majestic

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 12:40 am on Jul 10, 2008 (gmt 0)

don't rely on it as a tool to prevent discovery of pages you don't want crawled

robots.txt should be used to prevent crawling of pages - some people use nofollow META tags and then get very angry if their pages still get crawled - those nofollow's apply to urls discovered on that page, so if they are found elsewhere they will still be crawled regardless of META nofollow on some other page.

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 1:09 am on Jul 10, 2008 (gmt 0)

robots.txt is an excellent tool to prevent crawling of pages. Unfortunately, those urls will still show up in search results. Like I said previously, by using meta "noindex,nofollow" the page can be crawled, but it won't be indexed or show up in serps.

The nofollow is necessary in this example because internal site links are using rel="nofollow" to prevent the passage of PR to the page. I don't think anybody would want to pass low, and possibly untrusted PR from the links on that page to other links on their site.

Lord Majestic

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 1:13 am on Jul 10, 2008 (gmt 0)

Unfortunately, those urls will still show up in search results.

Well, if this page shows up in results then maybe it is a relevant page and it should not even be disallowed in robots.txt?

Key_Master

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 1:22 am on Jul 10, 2008 (gmt 0)

Well, if this page shows up in results then maybe it is a relevant page and it should not even be disallowed in robots.txt?

Maybe this link will help describe the issue better. And yes, I agree, robots.txt is a poor choice in this particular example.

[webmasterworld.com...]

freelistfool

5+ Year Member



 
Msg#: 3694262 posted 2:22 am on Jul 10, 2008 (gmt 0)

I've had good success using noindex, nofollow to block contact us, privacy policy, etc. pages that are in my footer. In addition, I use it for all pages in a pagination or sorting sequence. I suffered no loss of rankings...in fact I got a bump for the landing pages in each category.

ecmedia

5+ Year Member



 
Msg#: 3694262 posted 2:43 pm on Jul 10, 2008 (gmt 0)

I have found that nofollow is a great way to link to pages without inviting an overlinking penalty. I have one personal blog where I often point to all the articles that I have written on all the other blogs that I write on. Nofollow comes in handy because I can link without the fear that somehow I am trying to point too many links. I have used it very effectively for several months with no penalty or problems at all.

Lord Majestic

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 3:39 pm on Jul 10, 2008 (gmt 0)

I was just thinking that nofollow on internal urls can have completely different interpretation than nofollow on external urls (that may imply spam).

santapaws

5+ Year Member



 
Msg#: 3694262 posted 4:39 pm on Jul 10, 2008 (gmt 0)

yes the thread is specific to internal links.

Robert Charlton

WebmasterWorld Administrator robert_charlton us a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month



 
Msg#: 3694262 posted 5:15 am on Jul 11, 2008 (gmt 0)

Noindexing these type of overhead pages is something I'm very hesitant to do as I don't know what signals are used in part to establish "trust." Might Google or other SEs look to the presence of contact or privacy pages as such a signal? I don't know, but as long as they could, I want them to know that they're there.

One approach to this I'm considering would be to use the rel nofollow attribute on these service links from the home page, but not use them on links from some inner pages. Seems this might help channel home page PR to where I want it to go, but would assure that these pages are indexed via links from lesser pages. I don't know what kinds of quality signals, if any, this might send.

I've got to confess that up till now I've been very wary about using the nofollow attribute for PR sculpting, because it might send signals of micro-manipulation (assuming that Google wanted to bother with micro-analysis of micro-manipulation).

fishfinger

5+ Year Member



 
Msg#: 3694262 posted 8:26 am on Jul 17, 2008 (gmt 0)

use the rel nofollow attribute on these service links from the home page, but not use them on links from some inner pages

Moi aussi - it's not that I want to orphan these pages, but if I can send them less PR and others more then great. The clients want these links prominent so removing them from the template is not an option.

Has anyone from an SE said anything even slightly detrimental about use of nofollow internally, even after the PR sculpting experiment was published? I don't follow these things daily, but I've yet to read anything.

I like Tedster's iframe idea too and have used that - sort of like an invisible include file. But doing this (or using Javascript to import links) means that in Analytics the site overlay showing click data doesn't work because the links are no longer in the code file - and that is handy information. So I only use iframes for boilerplate text or for search functions to minimise code bloat.

jaffstar

WebmasterWorld Senior Member 10+ Year Member



 
Msg#: 3694262 posted 4:01 pm on Jul 24, 2008 (gmt 0)

I was just thinking that nofollow on internal urls can have completely different interpretation than nofollow on external urls (that may imply spam).

Many sites use the no follow to hoard pagerank. They try use it on pages such as contact us, about us etc.

Wondering how G feels about this... Is this considered spam or an unethical policy.

This 35 message thread spans 2 pages: 35 ( [1] 2 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google SEO News and Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved