Forum Moderators: Robert Charlton & goodroi
TECHNIQUE:
Add rel="nofollow" to pages which don't need to be in the SERPs such as About Us, Contact Us, Shopping Cart, etc. I've had good results with this.
Plus, if you want to remove a page from the serps, you're probably well advised to use robots.txt.
rel="nofollow"Add
rel="nofollow"Add
rel="nofollow"I prefer not to use it on a single reference outgoing link but will use it on multiple references to the same destination. For instance, if I'm linking to Fragment IDs, I'll
rel="nofollow"
It's not the only method I use to keep spammers at bay, but every little helps.
...
Add rel="nofollow" to pages which don't need to be in the SERPs...
I think that this is a bad way of stating the best strategy. I would put it instead as a way of shaping flow of PageRank where it's needed most.
So you may actually want Contact Us, Privacy, etc to be included in the index. You may just not want them to receive more PageRank, say, than 1,000 of your product pages. I think this is a more helpful way of viewing things.
PRINCIPLE: Chrome Diet
Template 'chrome' boiler-plate links do not deserve PR contibutions from every page on your site.
TECHNIQUE:
Currently (subject to revision), NF all EXCEPT Homepage links.
VARIATIONS SLATED FOR "CHROME DIET" TEST CYCLE
- NF all EXCEPT Homepage and Products (2-3 tiers skipped)
- NF all EXCEPT Homepage and 'pillar' pages
- NF all EXCEPT Homepage and from chrome pages themselves
- NF all EXCEPT Products
[edited by: Shaddows at 8:58 am (utc) on Feb. 18, 2009]
If I'm hijacking your thread, just have a mod delete...but contact us? I *frequently* search on phone numbers or 'widgets houston texas' or zip codes or whatever when I'm looking for a business. Remove the contact us page and I won't be able to find you.
Plus, if you want to remove a page from the serps, you're probably well advised to use robots.txt.
[edited by: Tonearm at 2:18 pm (utc) on Feb. 18, 2009]
So you may actually want Contact Us, Privacy, etc to be included in the index.
That could be, but I've kept them out with the idea that it would open "spots" in the non-supplemental index for more important pages.
There are situations where the Contact page should be part of the flow. In fact, I've left those as dofollow since we feel that is an integral part of the structure that needs to be intact.
I agree that Privacy Policy, Terms, etc. can be taken out of the equation.
There are not many instances where I've come across contact pages in me searches. The times that I did, they were part of a "section" and provided a bit more information than your standard contact form. There was location information, hours, contacts, etc. Those types of contact pages can be "power pages" if structured properly.
It's great using FireFox plugins that allow me to highlight nofollow links. I'm a visual person, I need to "see" things before they click 100%. After implementing my first nofollow routine, I clicked that NF feature for highlighting and BAM, I had an instant picture of flow. Of course I had to turn off images so I could see the highlighting on linked images too. Once you get that visual, it all clicks, or it did for me. :)
Add rel="nofollow" to any unmoderated and/or moderated UGC.
Home->Category->Sub-Category->Product
I'm not sure of how to handle pagination of Category and Sub-Category pages. Maybe page 1 should nofollow all page # links except page 2, page 2 should nofollow all but page 3, etc?
Is 50 dofollow links per page still the effective maximum?
EDIT: Should PR "die" with Product pages in the above example? It's the bottom page in my structure, but should it be dofollow linking back up to Sub-Category or over to other products? It seems like that could go against the strict structure concept for PR.
[edited by: Tonearm at 3:27 pm (utc) on Feb. 18, 2009]
If I'm hijacking your thread, just have a mod delete...but contact us? I *frequently* search on phone numbers or 'widgets houston texas' or zip codes or whatever when I'm looking for a business. Remove the contact us page and I won't be able to find you.
Plus, if you want to remove a page from the serps, you're probably well advised to use robots.txt
Its well documented on how to get into googles local business results. you dont need a contact us page to get that listing
It's the idea that lots of pages internally linking to lots of other pages effectively creates some PR of its own which boosts more pages than would be boosted by a strict structural link graph.
Heh, déjà vu! I remember a topic similar to this not so long ago. :)
You're still going to have lots of other pages linking internally. All you're going to do is "focus" the flow of those links. Take a page and start adding the
rel="nofollow"NF = NoFollow
DF = DoFollow
After I saw that visual, it really started to make sense. I look at that visual and think to myself, "okay, anything that is not yellow (my highlight color), is part of the equation. Based on where those non highlighted spots are, is that the best use of NF? Did I accidentally remove something that I shouldn't have? Do I have another duplicate that I've missed? It's an interesting process and I learned quite a bit using this visual approach. I learned all that from that one damn topic. Kept me awake for days. It still does. I browse quite a bit now with that NoFollow plugin active. Opens your eyes quite a bit.
Note: I am not even close to being an expert on the use of NoFollow. But I think I learned the intricacies of its usage by being actively involved in that one topic. tootricky and quite a few others were kind enough to walk us through the process without crossing their frustration threshold. :)
Now, does it work? I don't know...
Now, does it work? I don't know...
Will Google reward adherence to this concept?
No, no and no.
YOU will have control over your PR flow. The Ranking algo will run on the modified values, giving you uplift as appropriate.
UNLESS. And heres the thing. UNLESS they decide, on a whim, that PR sculpting IS manipulative and start whacking you for it.
The PR hoarders are distorting WebSpace as much as the link traders do, so I'd be surprised if NF doesn't get a good looking at soon.
YOU will have control over your PR flow. The Ranking algo will run on the modified values, giving you uplift as appropriate.
So goes the theory, anyway.
I'd be very surprised to see Google change on PR sculpting with nofollow
My uninformed prediction is that first phase would be non-disruptive- G would only look at proportion of NF links relative to all outbounds. Second phase would be disruptive, as G looked at sculpting 'intent' in funneling PR away from outbound pages (for example, replacing a NF OBL with an NF internal to a 'references' page, with DF links)
Of course, the tinfoil brigade would happily point out that MC encouraging SEO use of NF is a bit like Big Brother inventing the Brotherhood to get potential dissidents to out themselves.
internally generated PR
I'd say there may be such a thing - the math in the original PR formula certainly implies it, and I've often noticed that sites with many urls seem to have more PR for their urls than an external backlink report suggests.
This doesn't mean that throwing up a bunch of junk pages boosts PR - but in the early days of Google it seemed to and this was a spammer trick. There are clearly other factors in the algo now to guard against that.
The PR hoarders are distorting WebSpace as much as the link traders do, so I'd be surprised if NF doesn't get a good looking at soon.
Tinfoil-helmet logic: Why stop at "nofollow"? Why not be suspicious of sites that exclude directories or files with robots.txt, too?
Common-sense logic: Since Google actively participated in creating the "nofollow" attribute, why should Google think "nofollow" is something bad?
Why should Google think "nofollow" is something bad?
I too felt that way up until 2008 December. Not that it was bad, but that it was a "clear signal" that someone with SEO in mind is involved. Just one signal, but a clear one.
Now, take that one clear signal, add it to other signals that can be used to determine if there is an abusive manipulation taking place and BAM, you've crossed the threshold. I think that is the main question here, where is that threshold and "exactly" what is best practice for its use in this scenario. I've read "ALL" of the documents on this since that previous discussion and others over the years. I even have a few of my own that I wrote for posterity. :)
My thinking now is that it can be used effectively to micro-manage a variety of structural flow challenges that many face due to limitations of their knowledge, the platform that supports their work, etc. I was definitely late to the game on this one but I typically play it safe. :)
I've always believed that a page could gain (some) rank strength through OUTBOUND links; possibly enough to offset lack of inbounds. There are necessary conditions: the OB's must be 100% on target for the page theme, and the OB neighborhoods have to be A+ authorities (federal & state gov and edu in my case).
I built a whole site around this theme; its pretty much a directory to hard-to-find articles on certain socio-medical topics. After 4 years, it has a few good inbounds, but only a very few. I've felt it proves the theory about some value to good OB's. A typical page has 25 outbounds, 3 cross link internals, and perhaps one InBound.
Pages rank usually between 8-20 for popular 2-word terms; could be better, but I get thousands of long tails every day so I'm okay.
Now comes the wrinkle: I'm updating, and I thought 'why not make all those OB's nofollow, save what little juice I've got?' Maybe move up from the 2nd page to 1st in the Gserps?
But what about the theory that OB's can be good for ranking? Would I be throwing away one good thing for another, resulting in no net gain?
It obvious you don't want to pass anything in links to a privacy page, a junky form, etc. but what would you think about using nofollow on links to pure golden references?
Now comes the wrinkle: I'm updating, and I thought 'why not make all those OB's nofollow, save what little juice I've got?' Maybe move up from the 2nd page to 1st in the Gserps?
This is the part I don't necessarily agree with. There needs to be outbound flow to bring everything together. It's all about sharing. This is where the PR hoarders come into the picture and I don't think you want to be labeled as such.
I'd leave the OBLs alone. You've had them for years, apparently they are quality OBLs or they probably wouldn't still be there.
I might look at narrowing down which of those OBLs are being replicated on a single page. If they are, take the duplicates and NoFollow those. Keep the one that is in context as DoFollow. That is usually going to be one that is inline. Not always, but most of the time. I look at a combination of context and also where in the source the link appears.
I've always believed that a page could gain (some) rank strength through OUTBOUND links
Yes, that's a possibility - but it would mostly be ranking strength from the other query-dependent factors and not from PR.
However, if you are linking into a niche that interlinks relatively tightly, then even the PR equations show that "what goes around comes around." That is, the PR calculation is not just one step, it is iterated through the entire webgraph until the calculated PR value approaches a limit.
Those quality outbounds have served me well, and contributed mightly to my income for years. Why wouldn't I do the right thing to say thank you, and if I can pass along a little juice, then good.
I apologize for considering the greedy road.
But what about the theory that OB's can be good for ranking? Would I be throwing away one good thing for another, resulting in no net gain?
I've always felt that relevant outbounds to good quality pages were good for ranking. Brett in fact, in his classic Successful Site in 12 Months [webmasterworld.com] thread, has this to say....
G) Outbound Links:
From every page, link to one or two high ranking sites under that particular keyword. Use your keyword in the link text (this is ultra important for the future). As you can see from the WebmasterWorld google forum topics, Google highly values links. Inbound links are what people say who you are, and outbound links are what you say you are. Google will clearly use both in a the algo somewhere.
Maybe that's extreme for some sites. I've got to confess I don't always do it or even recommend doing it, but I most definitely consider it.
Also, from the above series of questions...
It obvious you don't want to pass anything in links to a privacy page, a junky form, etc. but what would you think about using nofollow on links to pure golden references?
It's not obvious to me that you don't want to pass anything in links to a privacy page. It may well be that having a privacy page is an indication of site quality. It's very likely that you shouldn't pass too much PageRank on to your privacy page, but you probably do want to make sure it's indexed.
One of the big concerns I had when rel="nofollow" was announced was that it was going to mess up the natural ecology of the web. That's something to continue thinking about. The considerations there aren't quite as simple as they might seem.
Tinfoil-helmet logic: Why stop at "nofollow"? Why not be suspicious of sites that exclude directories or files with robots.txt, too?Common-sense logic: Since Google actively participated in creating the "nofollow" attribute, why should Google think "nofollow" is something bad?
I feel misunderstood.
The (retracted) technique from dibbern2 is the problem I have with NoFollow.
Internal PR sculpting is fine.
Blocking your PR from 'escaping' your site is where the problem lies. If a significant proportion of sites NF'd all OBLs, Google would HAVE to take action, or the basis of their algorithm would no longer work. This is simply not the same as robots.txt or other technical deployments.
I would strongly encourage Google to penalise sites that make heavy handed use of NF in removing OBLs from the link graph. Assuming they do at some point, subdomain interlinking may well be affected.
Savvy hoarders would respond by making outbounds as DF, but would ensure the page they are on get minimal PR by NFing internal links to that page.
Google, as ever, would then seek to mitigate this deception, which would BY NECESSITY go to intent of NF implementation. Thus, the guessing game would begin about how Google would see THIS implementation over THAT implementation.
This is not tinfoil thinking. This is a reasonable analysis of the state of play. Dispute it if you like, but at least provide an opposing synopsis rather than accusing me of being reactively anti-google (which I'm not).
@Tedster
It is interesting that suspicion comes up so easily on this.
Abuse of WebSpace typically draws a response from Google. Penalisation typically produces a work-around, which results in an arms race, which means the "nice and clear" situation we currenltly enjoy becomes murky.
[edited by: Shaddows at 10:13 am (utc) on Feb. 20, 2009]
