Just another of my mad ramblings, appologies to those who get bored of me ...
My question is this: If the rel="nofollow" process is becoming a widespread and Google supported method of safe linking to 3rd party websites, how does Google intend to guage to the relative worth of sites in the future. I presume, as they have backed this process, that incoming links will become less and less important in the future. In fact, why bother with them at all ... surely it's about time that they developed a new and far more practical method of page ranking.
p.s. even I've made my directory links at www.norfolkdays.co.uk No Follow ... If my incoming links are not valid anymore, I'm blowed if I'm going to give any to anyone else ;-)
Here here Colin, I totally agree - read my comments posted on http://www.utheguru.com/a-plague-of-de-indexation-and-supplementals-g... - they are too long to write here, but yes, I think there is a certain amount of link greed around the place - people are reticent to link to others lest they bleed PR. This leads to hysterisis in the results.
I'm with you - Looking into my crystal ball, I think the nofollow tag is either going to totally deprecate backlinks in the future or result in some serious algo changes.
> Just another of my mad ramblings, appologies to those who get bored of > me ...
> My question is this: If the rel="nofollow" process is becoming a > widespread and Google supported method of safe linking to 3rd party > websites, how does Google intend to guage to the relative worth of > sites in the future. I presume, as they have backed this process, that > incoming links will become less and less important in the future. In > fact, why bother with them at all ... surely it's about time that they > developed a new and far more practical method of page ranking.
> p.s. even I've made my directory links atwww.norfolkdays.co.ukNo > Follow ... If my incoming links are not valid anymore, I'm blowed if > I'm going to give any to anyone else ;-)
Yeah Adam, I agree, but we're concerned about the implications from the fact that:-
a) A lot of webmasters use CMS systems that now nofollow by default and.. b) Some webmasters are scared about PR bleeding and might deliberately nofollow all links to avoid that happening, thus breaking the system.
I'm more concerned that for the last 18 months I have been struggling with listings and blaming Google updates and data pushes, when it actually seems that when the likes of Wiki etc withdraw their Follow warrant from our sites ... we drop like a stone.
I remember discussing whether or not it would be possible for a site with a high PR to virtually assassinate lesser sites and I think the answer is now yes. This at least goes some way to explaining Google's apparent surprise that there is any great change in the SERPS, as it would appear just the recalculation of links is all that is to blame. Personally, I will be using NOFOLLOW with all my external links in the future ... If it's good enough for Wiki then it's good enough for me. Let's face it ... Incoming links will never again hold the clout that they used to in PR calculation ... in fact the only reason for using a follow to another site must be to speed up search engine recognition and therefore the practice should be regarded as blatant spamming (Ban them all ;-))
Adam wrote "nofollow should be used when the Webmaster can't necessarily "vouch for" a link ........"
Thanks for replying Adam. As with any method of PR calculation there will always be those out there that use the process to adversly effect the SERPS. I'm sure that Wiki decided to use the NOFOLLOW for honest reasons, but it seems obvious that it has effected a whole load of sites adversly with the result that they now seem unable to recover. What is to stop other, less honest, high PR sites from doing the same to get rid of competition. In fact, in this highly competitive world what is the sense in helping others ... sure it's all very nice, but not very logical.
IMHO Incoming Links no longer truly reflect the quality of sites and it should be scrapped as a contributing factor when calculating relevance. Maybe DMOZ's manual checking is back in fashion ... less sites listed, but the human checking process at least ensures good quality results.
I don't think links from wiki were that powerful to begin with. But I agree with the problem of people, some large, using the nofollow, incorrectly, and its going to hurt the we. I see it often on sites' own INTERNAL linking!! You're telling me you don't trust your own pages? If you don't want pages indexed there are other methods for that and NOFOLLOW isn't one of them. If I were google that would be a signal, and not a good one.
Wiki basically made a statement that it's a weak organization based on a flawwed premise that is not scalable. They say they are trying to figure out a way to remove the NOFOLLOW from trusted sites but just did the global change now until they do. Any site that is too big too manage and too full of untrusted links is the definition of SPAM in my opinion, but then what do I know wiki is #1 search result for just about every word in the english language.
My opinion is that NOFOLLOW should automatically added to unmoderated areas, like forums and blog comments, guest books etc, but then after you've evaluated them the good ones remove nofollow and the bad ones delete. As far as your own publishing goes, if you don't trust the site, or don't want to give it link juice why would you want to give them traffic? Then only list the site name without an active link. I'm working on implementing that right now on my sites.
Oh yeah, and all links to wiki should also be nofollow as its not a trusted resource as none of their sources can be trusted as they are all nofollowed.
> Adam wrote "nofollow should be used when the Webmaster can't > necessarily "vouch > for" a link ........"
> Thanks for replying Adam. As with any method of PR calculation there > will always be those out there that use the process to adversly effect > the SERPS. I'm sure that Wiki decided to use the NOFOLLOW for honest > reasons, but it seems obvious that it has effected a whole load of > sites adversly with the result that they now seem unable to recover. > What is to stop other, less honest, high PR sites from doing the same > to get rid of competition. In fact, in this highly competitive world > what is the sense in helping others ... sure it's all very nice, but > not very logical.
> IMHO Incoming Links no longer truly reflect the quality of sites and it > should be scrapped as a contributing factor when calculating relevance. > Maybe DMOZ's manual checking is back in fashion ... less sites listed, > but the human checking process at least ensures good quality results.
hehe... yes it annoys the living daylights out of me that wiki pages are so often at the top of the results JLH - in my opinion, alot of the information on the wikis is bordering on rubbish anyway..
> I don't think links from wiki were that powerful to begin with. But I > agree with the problem of people, some large, using the nofollow, > incorrectly, and its going to hurt the we. I see it often on sites' > own INTERNAL linking!! You're telling me you don't trust your own > pages? If you don't want pages indexed there are other methods for that > and NOFOLLOW isn't one of them. If I were google that would be a > signal, and not a good one.
> Wiki basically made a statement that it's a weak organization based on > a flawwed premise that is not scalable. They say they are trying to > figure out a way to remove the NOFOLLOW from trusted sites but just did > the global change now until they do. Any site that is too big too > manage and too full of untrusted links is the definition of SPAM in my > opinion, but then what do I know wiki is #1 search result for just > about every word in the english language.
> My opinion is that NOFOLLOW should automatically added to unmoderated > areas, like forums and blog comments, guest books etc, but then after > you've evaluated them the good ones remove nofollow and the bad ones > delete. As far as your own publishing goes, if you don't trust the > site, or don't want to give it link juice why would you want to give > them traffic? Then only list the site name without an active link. > I'm working on implementing that right now on my sites.
> Oh yeah, and all links to wiki should also be nofollow as its not a > trusted resource as none of their sources can be trusted as they are > all nofollowed.
> > Adam wrote "nofollow should be used when the Webmaster can't > > necessarily "vouch > > for" a link ........"
> > Thanks for replying Adam. As with any method of PR calculation there > > will always be those out there that use the process to adversly effect > > the SERPS. I'm sure that Wiki decided to use the NOFOLLOW for honest > > reasons, but it seems obvious that it has effected a whole load of > > sites adversly with the result that they now seem unable to recover. > > What is to stop other, less honest, high PR sites from doing the same > > to get rid of competition. In fact, in this highly competitive world > > what is the sense in helping others ... sure it's all very nice, but > > not very logical.
> > IMHO Incoming Links no longer truly reflect the quality of sites and it > > should be scrapped as a contributing factor when calculating relevance. > > Maybe DMOZ's manual checking is back in fashion ... less sites listed, > > but the human checking process at least ensures good quality results.
> > All the Best
> > Col :-)- Hide quoted text -- Show quoted text -
"As far as your own publishing goes, if you don't trust the site, or don't want to give it link juice why would you want to give them traffic?"
Hi JLH,
>From my point of view if I NOFOLLOW all my external links then Google
can't accuse me of being a link farm or selling links. Then the only benefit that a site can get from my links are direct visitor links. I have a sneaky feeling that Google think we all make millions selling links and I think that this process at least lets them see that I couldn't care about that sort of revenue.
As for internal links, I do not NOFOLLOW them, also I interlink my own sites, as they advertise my different services, which in turn compliment each other. I agree with your point about it looking foolish to not trust your own sites ;-)
There is no doubt in my mind that pagerank leak is real.
In that case, Game Theory suggests each user's (web site administrator's) dominant strategy is to use 'nofollow' for all the links originating from the site.
And guess what the game equilibrium is then?
Nobody links to nobody (except with nofollow links).
Does anyone advise that I should change my "Nofollow" links back again ?
I'd be especially interested to hear from a Google representative on the subject. I'd obviously like to think that my considering someone elses site good enough to link to was going to result in something positive for the site I'm linking to. But obviously I wouldn't want to be penalised for being a link farm by a search bot.
It's brand new and I may change it, but I think if we could all get on the same page it would be a benefit. First thing they should do is change it to "notrust" as they do follow the links.
> Does anyone advise that I should change my "Nofollow" links back > again ?
> I'd be especially interested to hear from a Google representative on > the subject. I'd obviously like to think that my considering someone > elses site good enough to link to was going to result in something > positive for the site I'm linking to. But obviously I wouldn't want to > be penalised for being a link farm by a search bot.
What do you mean by page rank leak? Are you implying that if you link to a site you leak PR on the page you linked from? If so, you are wrong. If you mean you lose some benefit to internal pages of that pages PR by linking out to other sites instead of internal pages you are right. You don't lose PR value on the page you are linking from though.
> There is no doubt in my mind that pagerank leak is real.
> In that case, Game Theory suggests each user's (web site > administrator's) dominant strategy is to use 'nofollow' for all the > links originating from the site.
> And guess what the game equilibrium is then?
> Nobody links to nobody (except with nofollow links).
'in my opinion, alot of the information on the wikis is bordering on rubbish anyway.. '
totally agree. For too long now misleading and wholly inaccurate articles appear in Wiki written purely as a vehicle for promoting the authors commercial site. 'The nofollow tag for all policy will at least combat some of the spamming issues with Wiki and hopefully improve the quality of its articles.
As for PR leakage i have seen quite a few sites using the nofollow attribute for internal links to pages such as 'Privacy policy' or 'terms and conditions' (i.e. pages which although are of importance to a user have no benefit from being given internal PR). I have also seen it used on paypal buttons where a nofollow attribute has been used to the paypal links. I think it stems from a misconception that a link off a page will diminish its pager-rank. Even if that were true the use of robots tags or txt would be the better solution although as regards paypal buttons i assume that 'block' links of that nature are already excluded from a search engines pager rank computations. For me the only use i can see for a nofollow attribute is on a blog or website that invites comments.
On an off topic I wanted to mention that I appreiciate the frankness in which you discuss issues. Listening to you at PubCon in Vegas after the monotone generic speaches from the other 'unnamed' people on the panel was refreshing to say the least. Thanks!
> It's brand new and I may change it, but I think if we could all get on > the same page it would be a benefit. First thing they should do is > change it to "notrust" as they do follow the links.
> > It's brand new and I may change it, but I think if we could all get on > > the same page it would be a benefit. First thing they should do is > > change it to "notrust" as they do follow the links.- Hide quoted text -
That's what I mean - the page itself doesn't lose any pagerank directly, but the internal pages it links to will. As a consequence, if these internal pages link back to the original page, even the original page may suffer some pagerank loss. This is if ranking works as described in the original paper by Google founders.
So the way I see it, you are better off as a whole keeping the links within your site, and never linking out.
This means that right now we have a game that's acceptable only because of user ignorance. On the other hand, ignorance should not be underestimated.
> What do you mean by page rank leak? Are you implying that if you link > to a site you leak PR on the page you linked from? If so, you are > wrong. If you mean you lose some benefit to internal pages of that > pages PR by linking out to other sites instead of internal pages you > are right. You don't lose PR value on the page you are linking from > though.
> On Feb 4, 3:28 pm, Mario_V wrote:
> > There is no doubt in my mind that pagerank leak is real.
> > In that case, Game Theory suggests each user's (web site > > administrator's) dominant strategy is to use 'nofollow' for all the > > links originating from the site.
> > And guess what the game equilibrium is then?
> > Nobody links to nobody (except with nofollow links).- Hide quoted text -
> This is if ranking works as described in the original paper by Google founders.
That's a big if :). Our algorithms have substantially evolved, and are based upon FAR more than just PageRank. PR just happens to be the most (greenly) visible.
Overall, I'd suggest that it's very counterproductive to plan internal linking strategies based upon perceived PR leaks. As you might have noticed, many of the world's most successful sites link liberally to other sites, and this sort of thing is often appreciated by and rewarded by visitors. And if you're editorially linking to sites you can personally vouch for, I can't see a reason to no-follow those.
News of the death of the Web / linking / etc at 11^H^H^H^H^ is greatly exaggerated ;)
<< As you might have noticed, many of the world's most successful sites link liberally to other sites, and this sort of thing is often appreciated by and rewarded by visitors. >>
I think the thing that I trying to get to is why it would make a blind bit of difference to my visitors Adam. They will still see a useful link and use it, my site will still be helpful and the linked to site still receives a hit from me. Once again the webmaster seems to have to decide whether the Search Engine is an important factor in the building of their page.
IMHO I think it shows that I'm no link farmer and that can't be a bad thing.
All the Best
Col :-)
P.S. I agree with Jeff Lawrence on the subject of Adam's easy way with words. It's refreshing not to be given the run around.
I agree that PR hoarding is a plain stupid practice. However, there are many use cases where a redesign of the misleading and pretty hapless attribute *value* 'nofollow' would make sound sense.
"Real" nofollow: suitable in links to printer friendly pages, privacy policies, TOS ... "nofollow" should be interpreted as a search engine crawler directive meaning "Dear Googlebot, please do NOT fetch the linked resource".
Ignore link: suitable for example in crawlable horizontal popup menues where every page links to all pages (of a section or even the site) giving search engines a hard time in figuring out which link should pass which weight. Without an "ignore link" spider directive the only workaround is cloaking, that is suppressing the horizontal menu on crawler requests. Not nice, and by the way there are correlations with Google's section targeting (using HTML comments). Thinking a tiny step forwards, normed "ignore" directives, implemented by all engines, would be a great thing to have on HTML element level at all.
Unapproved: suitable to mark links in user generated content, the primordial intention of "nofollow" before its semantics morphed to a bunch of meanings, misuse included.
Advertising, affiliated, biased, example, norm, inventor, worstenemy, untrusted, hownotto, navigational ...: just examples of more or less useful values of crawler directives on link/element level. Unfortunately Technorati's microformats votelinks and nofollow don't cover those. It would be fun to work out the RFC :)
Meanwhile, gazillions of non-geeky site owners, publishers and editors make use of rel=nofollow for the wrong reasons, or don't apply it to links where they should, just because they do not understand its functionality, intent, and ongoing semantic morphing. That's why I say no to nofollow as is, although I admit that I do use it every now and then.
> I agree that PR hoarding is a plain stupid practice. However, there > are many use cases where a redesign of the misleading and pretty > hapless attribute *value* 'nofollow' would make sound sense.
> "Real" nofollow: suitable in links to printer friendly pages, privacy > policies, TOS ... "nofollow" should be interpreted as a search engine > crawler directive meaning "Dear Googlebot, please do NOT fetch the > linked resource".
> Ignore link: suitable for example in crawlable horizontal popup menues > where every page links to all pages (of a section or even the site) > giving search engines a hard time in figuring out which link should > pass which weight. Without an "ignore link" spider directive the only > workaround is cloaking, that is suppressing the horizontal menu on > crawler requests. Not nice, and by the way there are correlations with > Google's section targeting (using HTML comments). Thinking a tiny step > forwards, normed "ignore" directives, implemented by all engines, > would be a great thing to have on HTML element level at all.
> Unapproved: suitable to mark links in user generated content, the > primordial intention of "nofollow" before its semantics morphed to a > bunch of meanings, misuse included.
> Advertising, affiliated, biased, example, norm, inventor, worstenemy, > untrusted, hownotto, navigational ...: just examples of more or less > useful values of crawler directives on link/element level. > Unfortunately Technorati's microformats votelinks and nofollow don't > cover those. It would be fun to work out the RFC :)
> Meanwhile, gazillions of non-geeky site owners, publishers and editors > make use of rel=nofollow for the wrong reasons, or don't apply it to > links where they should, just because they do not understand its > functionality, intent, and ongoing semantic morphing. That's why I say > no to nofollow as is, although I admit that I do use it every now and > then.
Thanks for the warm welcome :) I'll do my very best to satisfy your needs ;) However, still not recovered I may waive rants in favour of honest opinions ...
> Welcome back Sebastian! Pull up a chair and see how the group has > changed! I look forward to some heated rants and honest opinions.
> On Feb 6, 4:25 am, Sebastian wrote:
> > I agree that PR hoarding is a plain stupid practice. However, there > > are many use cases where a redesign of the misleading and pretty > > hapless attribute *value* 'nofollow' would make sound sense.
> > "Real" nofollow: suitable in links to printer friendly pages, privacy > > policies, TOS ... "nofollow" should be interpreted as a search engine > > crawler directive meaning "Dear Googlebot, please do NOT fetch the > > linked resource".
> > Ignore link: suitable for example in crawlable horizontal popup menues > > where every page links to all pages (of a section or even the site) > > giving search engines a hard time in figuring out which link should > > pass which weight. Without an "ignore link" spider directive the only > > workaround is cloaking, that is suppressing the horizontal menu on > > crawler requests. Not nice, and by the way there are correlations with > > Google's section targeting (using HTML comments). Thinking a tiny step > > forwards, normed "ignore" directives, implemented by all engines, > > would be a great thing to have on HTML element level at all.
> > Unapproved: suitable to mark links in user generated content, the > > primordial intention of "nofollow" before its semantics morphed to a > > bunch of meanings, misuse included.
> > Advertising, affiliated, biased, example, norm, inventor, worstenemy, > > untrusted, hownotto, navigational ...: just examples of more or less > > useful values of crawler directives on link/element level. > > Unfortunately Technorati's microformats votelinks and nofollow don't > > cover those. It would be fun to work out the RFC :)
> > Meanwhile, gazillions of non-geeky site owners, publishers and editors > > make use of rel=nofollow for the wrong reasons, or don't apply it to > > links where they should, just because they do not understand its > > functionality, intent, and ongoing semantic morphing. That's why I say > > no to nofollow as is, although I admit that I do use it every now and > > then.
Would nofollow tags explain the paucity of the link: advanced search results? My link: search has absolutely no relationship to the sites I get my referral traffic from.
