PR Sculpting Doesn't Work and Internal NoFollow Can Harm Your Site

[incrediBILL]

In Matt's latest video blog titled Should I use the nofollow attribute on internal links? [youtube.com] he takes 2 minutes to emphatically say "NO! DON'T USE REL=NOFOLLOW ON INTERNAL LINKS!"

I've never personally used it and have never recommended it for internal "page rank sculpting" but some people still promote this crazy trend regardless of the fact that Matt has pretty much emphatically said "don't do it" forever.

I knew someone that attempted to sculpt their internal links once and Google dropped some very odd pages from their site apparently from having the flow of page rank broken throughout the site. Once they removed the NOFOLLOW from the internal links the problem quickly resolved itself.

Again, internal page rank sculpting via REL=NOFOLLOW is a BAD IDEA and can do more harm than good.

[arronsky]

this is good info... I'd be curious if anyone here has any real experience with releasing nofollow and seeing a change for the better. To be honest, we used to nofollow a TREMENDOUS portion of our less interesting internal links (such as user profiles)-- we then relaxed it and had negative results. This was over the last few months, so it's not like it was before the Matt Cutts stance was public.

[tedster]

Last year I was asked to look at a site that had major ranking troubles. They had used nofollow extensively, trying to sculpt PageRank to go only to those pages they wanted to see in search results. Even some main menu items had nofollow!

At the time they were below 10% of total traffic from search. I suggested they remove all of the nofollows and they did. One month later they were up to about 30% traffic from search. I haven't checked in with them lately, but I'd guess they continued to improve - unless they got tricky again.

That said, a year ago Matt did say that he uses an internal nofollow in one situation:

The only place I deliberately add a nofollow is on the link to my feed, because it’s not super-helpful to have RSS/Atom feeds in web search results.

[mattcutts.com...]
The biggest problem with internal nofollow that I see is that so few people really understand what PageRank is even at a basic level. And only Google knows how the granular details work today. Trying to play around with that much unknown on anything but a throw-away domain is asking for trouble.

[CainIV]

Interesting topic. I am surprised there are still people out here doing this after Matt Cutts was so clear about it.

I am wondering how this might affect a redirect on the same website where the nofollow attribute is not used, but where the redirect page is blocked by robots.txt. On one website I use a redirect tracking script as an affiliate, and the url that launches the script and is linked to is setup as disallow in robots.txt

I had heard some rumblings that in those situations Google might treat the link as a nofollow.

[steveb]

"I am surprised there are still people out here doing this after Matt Cutts was so clear about it."


FUD dies hard, though this time he used the phrase "pagerank disappears". he did say that with the examples he used, but if he just would have used those two words, all the sculpting nonsense would never have even started.

[incrediBILL]

The sculpting nonsense started because people needed something new to sell.

I never saw evidence that using internal nofollow links did anything but harm sites.

[tedster]

I had heard some rumblings that in those situations Google might treat the link as a nofollow.

My analysis would be that your redirect script itself can still collect PR -- and that would not be the case with a literal nofollow attribute on the link. So the script could potentially rank as a URL-only listing.

All of that is purely academic, of course. It's very unlikely for the uncrawled redireect script to have relevance for any "real" search query.

But still no PR can flow to the final target page of the redirect because Google has no way to see what that URL is, so in the practical situation it is something like a nofollow. The PR "vote" still gets split among all the links on the page, etc. but the script itself can have a place, however insignificant, on the full webgraph that Google maintains.

[whitenight]

Yay!... my fav discussion topic where the words
"MC says...blah blah blah"
somehow stands as the final verdict on the subject and absolutely NO objective testing or critical analysis is ever used.

(of course, that would involve work and "t-e-s-t-s"... naughty word around here, i know)

And still waiting for someone on this board (who agrees with MC) to actually TEST this THEMSELVES... (except me) and post results?
MC certainly has NO reason to lie, misdirect, or mislead regarding this topic, would he?!

Of course he doesn't. Carry on....

So once again, i'll take MC at his word for the amount of time it takes me to TEST THIS... and then make my own observations. TYVM
(yep, sry, not sharing those results here, good or ill. Not doing others homework for them)


And yes, i DO agree with MC on this fact.
MOST PEOPLE DON'T KNOW HOW TO PAGERANK SCULPT PROPERLY
(or even know how to quantify the results of rel=nofollow sculpting)
and so should not be encouraged to.

But of course, we've had THIS discussion going on 3 years now.
And even tho i've proven it to myself and others until we're blue in the face, there's no fighting it...

"cause MC said...."

<sub>and of course, i have my $5k pagerank sculpting course to sell as well... ;) as opposed to MC who has a multi-billion dollar company that he's getting paid to not get "gamed"

PS. There is no course.. so don't PM me for info...</sub>

[tedster]

Whitenight, did your testing see any change last year in how nofollow affected the division of the PR vote across all the on-page links? I was among the many who did not catch any change - and no one wrote anything online about it until Matt mentioned it.

[arronsky]

extremely well said, whitenight. Also, watching the video, MC wasn't as extreme as to say your site will be punished if you use nofollow internally, and he as usual had some clause of 'for unusual circumstances it's OK' which is the unusual case of the login page.

We don't know what to make of it-- we have not seen positive results from relaxing nofollow use internally, and quite frankly there ARE lots of pages on a large site that aren't that useful to waste not only Pagerank, but googlebot time. Robots.txt is one way to help, and nofollow (and canonical) are other avenues to curtail their indexing.

[whitenight]

Whitenight, did your testing see any change last year in how nofollow affected the division of the PR vote across all the on-page links? I was among the many who did not catch any change - and no one wrote anything online about it until Matt mentioned it.

Hey tedster,

yes, i remember the thread/pronouncement re: that topic and my purposeful restraint from it.

We tried a couple different ways to "decipher" what in the world MC was trying to imply and still couldn't figure out how it affected our pages ranking.

Of course, most of our testing was involved with Caff, so we might not have given the full attention we should have, but you weren't the only who was left wondering "huh...and this effects us how, again?"

Let me re-iterate the HUGE ELEPHANT in the room about this topic.

Gorg is the ONLY one who has changed their policy on rel=nofollow. ...
- They forced it upon the web.
- Noticed smart webmasters were using it for a huge advantage.
- Turned it off.
- Didn't like those results either.
- So they turned it back on.
- And have been making misleading statements about rel=nofollow's "effectiveness" ever since.


Also let me say that as usual, this PRONOUNCEMENT coincides with other interesting rel=nofollow "findings" in the SEO world.

So, i'll quietly test BOTH concepts and see which one is ACTUALLY more truthful...or at least most effective.

[tedster]

My take on this current Matt Cutts video is that he is essentially cautioning the people who don;\'t understand PR very well but have still been playing around with what they think PR sculpting is. I wouldn't be surprised if I could get into Google's data to see a whole bunch of sites who tried to be clever but essentially committed suicide.

This video is aimed at a wide audience, and that's the biggest grain of salt that I take with it.

[arizonadude]

Again, internal page rank sculpting via REL=NOFOLLOW is a BAD IDEA and can do more harm than good.

I clearly remember the Mouth Matt originally telling everybody that sculpting was the in thing, so programmers made their CMS systems and other scripts use the nofollow tag.

Advance forward, and oops! It works to good so we don't want people to use it and no it looks like they now penalize a practice they actually got people using.

Classic Google...........

[tangor]

There's the web (vanity publishing) and there's the commerce. Tough path to travel either way.

[dusky]

Plain and simple, before caffeine, their index was not coping, and needed help from us to minimize on "thin / non essential internal pages" and to use their index space only for the best content any site could offer. Gbot was getting stuck in the loop of the so called duplicate and useless content (printer friendly pages, user profiles, logins, newsletter sub forms, reply to threads links etc, etc..). Now they have caffeine (much larger space to store), faster crawler and indexing system, and a "smarter" algo probably one of its best points is to accurately index the intended page and leave out other dups of it with ease. What else to do except come out with a public announcement on a blog disguised as a user asking a question (as usual from MC).

All of a sudden, now it's okey to let printer friendly forum / news pages for example be indexed and followed and it's fine if they get into the index!

They caused the knee jerk reaction in the first place with all sorts of scare tactics when actually they only wanted us to help them STOP their crawler collecting duplicate and non-essential content, not because of pagerank or any other ranking factor!

Surprise! surprise! the announcement came after an almost complete halt of Gbot activity for two days. Add to that, their hint on site: and inurl: commands and the zillions of phantom backlinks and internal links which are years old, long deleted, dups or blocked by robots.txt file on WMT.
They want us to open every link to every page on our sites for indexing, they want to decide which to index and which to ignore, they even rolled out years old indexed historical data and sent gbot crawling every long dead link just to prove a point, the problem is, that point is top secret!

[dusky]

@whitenight

Gorg is the ONLY one who has changed their policy on rel=nofollow. ...
- They forced it upon the web.
- Noticed smart webmasters were using it for a huge advantage.
- Turned it off.
- Didn't like those results either.
- So they turned it back on.
- And have been making misleading statements about rel=nofollow's "effectiveness" ever since.

Couldn't agree more!

[dusky]

Is G* pulling a fast one on Y! and B*ng, by predicting if most webmasters reversed their nofollow strategy, G* can handle dup and non-essential content now better and they can't, hence a competitive advantage for them and a nightmare for Y and B crawlers and algo!

[tangor]

Don't short Bing (no asterisk). Redmond is turning in better results on average than obese Google who has not YET shed all that OUT-DATED data they cling to. Wonder which of the two philosophies will prevail: KEEP IT ALL DAMN YOU and KEEP IT QUICK NEAT AND TODAY.

Time will tell.

[incrediBILL]

MOST PEOPLE DON'T KNOW HOW TO PAGERANK SCULPT PROPERLY

Yeah, right.

Googlebot has screwed up more than a few sites I know about that appeared to do everything right but it went south in the index and this wasn't just recent, it's happened since rel=nofollow first started best I can remember.

People that knew what they were doing could sculpt a site without using rel=nofollow in the first place, it was just a placebo for those that couldn't think it through and it confused Googlebot in the end.

Bad idea.

Besides, the real elephant in the room was that Google couldn't tell external link relevance so they built this contrivance so you could tell them those links weren't important.

I always said if Google couldn't tell that Granny's Crochet Blog wouldn't normally link out to pharma spam then the Emperor had no RELEVANT Clothes which is why they made rel=nofollow in the first place.

Sad.

[tedster]

sculpt a site without using rel=nofollow

Always the best - and webmasters who put 100 navigation links in their basic template are NOT getting the message about what their visitors want, to say nothing of Google.

But I still think rel=nofollow comes in handy for print pages (no, I don't always want to develop a separate print CSS file or use X-Robots directives for PDF files!), feeds (just as Matt was doing last year) and a few other doo-dads from time to time. Not for PR sculpting purposes, but just for keeping the extraneous site collateral out of the SERPs.

[incrediBILL]

But I still think rel=nofollow comes in handy for print pages

So is robots.txt and it's universally supported!

[tedster]

It depends on the size and complexity of the web development before they ask me to come in with the mop and bucket. Some programmers/developers can weave the most complex technical macrame as they try to show off their chops ;)

[tangor]

I keep it a bit more personal... and not worry about Gorg et al. ...

Do I want to share this or not? More direct: Do I LINK or NOT?

Most times NOT. Hence, no need for nofollow...

Sometimes we create a forest when there's only one tree...

YMMV

[dusky]

Remeber, we are talking about Internal Linking here what MC is advocating / advising, nofollow is a good idea for linking out, as protection from (amongst other situation) comment spam etc.

@tangor

Don't short Bing (no asterisk).

any vested interest there, or is it just a case of let's give those guys a promotional helping hand!

[tangor]

any vested interest there, or is it just a case of let's give those guys a promotional helping hand!

Can also ask if there is any vested interest in obfuscating the rest of the net? :)

Internal, external, what the diff? Link as needed. Period. No rocket science (or Magic Potion) for PR, that ephemeral Carbon Tax and Trade made out of Nothing which is a Google Thingie...which they created, hold, present, and taketh away (consult your local politics to see how it actually works... duh, seeing it on the web already!).

Make your site. Link as appropriate. Done.

[curioustoddler]

I have one experience to share. One of my friends was trying to rank well in a very competitive trade. Site was always ranking on second or third page for most keywords, but very few first page rankings. Then he tried to copy the top ranking site in trade. They had put all the deep internal pages in noindex, nofollow and their links on other pages in nofollow. I tried to stop him because i had the impression that losing thousands of indexed pages will harm his site. He went ahead with the changes and within two months his site is doing very well.

[curioustoddler]

I have a question here.
Suppose you run a real estate site where any user or real estate agent can come and list their properties. Now over a period of time for each city you have hundreds of listings and gradually your site has few hundred city pages which you can manage to be different and rest of thousands of real estate listings which are available on hundreds of other sites. These people who post listings, just copy and paste on each site. They are duplicate pages on hundreds of sites. So whether they should be nofollow( noindex also) or they should be allowed to be indexed and treated as supplemental pages. This is not a hypothetical situation. I am facing this and i believe lots of others will also need to know this.

[Mark_A]

I am not even convinced the SEs comply with no follow on links. One site I track only has one decent inward link, from wikepedia, and it is a no follow, but it seems to be transmitting something that one big link!

[tedster]

These people who post listings, just copy and paste on each site. They are duplicate pages on hundreds of sites. So whether they should be nofollow( noindex also) or they should be allowed to be indexed and treated as supplemental pages.

I'd let them stay indexed and dofollow. I assume there is some internal navigation on these pages, right? If so it sounds like a lot of PR circulation that I wouldn't want to cut off. Have no concern about a duplicate content "penalty" - there really isn't such a thing for a case like this.

[curioustoddler]

Thanks Tedster, I was waiting for your response. I am concerned about two things-
1. These listing pages make over 80% of site's total pages and are typical example of duplicate content. We do allow users to search and navigate to these pages but afraid of google.
2. I assume that google will not index even 100k pages of your site if your site does not have that kind of authority and backlinks. and If we allow these listings to be indexed then google leaves lots of our city pages which are more important and indexes these listings. The same situations being discussed here [youtube.com...]