Forum Moderators: Robert Charlton & goodroi
Will someone please "citate" :) ANY REAL PROOF that linking out helps SERPS?
Wikipedia has over a million internal links and a million incoming links. That's why they rank high.
Pick a REAL example.
Of course, I'll pick two sites off the top of my head that do NOT link out and rank just fine. (since we're using absurd examples of why linking out works)
Apple.com
Adobe.com
(those evil PR hoarders)
[edited by: whitenight at 12:58 pm (utc) on Nov. 18, 2006]
That makes sense.
Europe
I am not an SEO type. I am no different than my trainees and millions of others who do not have sufficient background or basis to make a considered judgement as to what to do.
There are many people who come to this board to try to learn something and are probably turned away because they see the snide responses all too often given to simple questions asked by people who are struggling to get their sites straight.
A professional answer is an answer given professionally and not like one would expect from a clerk at the motor vehicle bureaus of old.
If the answer is that there is no definite answer, that is professional.
I applaud Bison for putting his theory in simple terms. Become the center of influence or hub for information and though you drain page rank, you improve ranking.
And it would appear that you have to choose to whom you link carefully based on some unknown and undefined criteria.
Now that make seem basic, but I expect it is like revealing the light to many people.
I find it quite amusing that using the nofollow tag is considered abusing the no follow tag.
I asked a simple question. Should the no follow tag be used or not and why. And because I asked a question, even after specifically stating that I link out without no follow, I am accused of abusing or advocate abusing it - whatever abusing the nofollow tag means.
I will not pursue that question because it does not fit this thread.
So as it seems:
Linking out bleeds PR.
Linking out, though it bleeds pr, improves ranking.
Taking an approach like Wikipedia where every link out is directly on target, whatever that means, for the subject in the page may be a good guideline to test.
Thank you again Bison.
Trinorth, since you're the only one here going on SOME type of actual testing, I'd like to hear further expounding of your testing and results. If you don't mind.
We have also just set up some links to some wiki definitions as well last week. Since the orgainzations have high PR and were not competition, google seemed to like it.
The best thing you can do is to experiment. Google will not penalize you for tossing up relevant links on your sites. Toss a few up and sit back a few weeks and see if your serps change for variations of the keyword anchor text.
Just make sure the links are relevant and ask yourself if your visitors will find it useful and it will be a benefit to your site.
White - I will both post and sticky you.
Thank you, Gentlemen
I will now go back to fixing the mess I created, otherwise known as an attempted website.
1. If you sold widgets, and wanted to game Google (spam), you could create a bunch of sites eg www.cheap-widgets-in-new-york-buy-from-me.com with garbage/scraped content all linking back to www.fredswidgets.com. Google, as most users would, recognise this as spam.
2. On the other hand, if you had a site www.fredwidgets.com and you had a page www.fredswidgets.com/pink-purple-widgets.html, which was well written and linked to say the pink/purple widgets page on Wikipedia, a news article from the BBC about pink/purple widgets and a .edu article, then Google, like most users, will recognise this as quality content.
The lesson is:
quality content + quality, trusted, relevant links = good.
link pages and anything that resembles spam = bad.
Of course, I'll pick two sites off the top of my head that do NOT link out and rank just fine.
I doubt if many of us here can compete with Apple or Adobe on the "authority" front. In any case, it's pointless to consider factors such as linking in isolation, because Google doesn't consider them in isolation. (Google has said publicly that "more than 100 factors" go into determining a page's ranking.)
Of course, I'll pick two sites off the top of my head that do NOT link out and rank just fine.
F(ranking) = a1f(A1)+a2f(A2)+a3f(A3)+....+anf(An)
Outbound links is just one factor which most of us feel adds on to the ranking as it adds on to the visitor experience. Even if the outbound part is 0, ie ayf(Ay) = 0 you can rank well. You do not need all parts just few like inbound links, content, visitor experience which plays major role.
AjiNIMC
Yes, outbound links help your rankings. No, don't use nofollow on them. Google does pretty good job in finding naturally built sites with good content. Even spam site becomes less spammy if there are a few quality links added. And recognising quality outbound links is one of the easiest things to implement in the algo. Using outbound links I have achieved high rankings many times.
It's self evident it matters, even if it can never help, then having a certain score could hurt.
They don't just do things for no reason.
You are obviously an expert so you are sure to be able to give an answer to the question. I asked if one should not use nofollow, "why not?"
I am looking for a sound reason.
I am sure that you being an expert you read in this thread that I am not using it in my outbound links.
My reason for not using no follow is simple. I just have a gut feeling that no one understands the effects nofollow fully, and that includes google, and there may be some eventual kickback.
So far this thread has firmed my opinion that there is a lack of understanding. I see a lot of generalized statements but no substance. Only one man in this thread gave some advice based on testing. And that is important.
Now when I started programming in 1963 I learned the importance of testing a hypothesis. It has been many years since I have let others do that work while I do other things. But I still look for the test results.
I think this thread has been productive since by asking some very simple questions a very simple hypothesis about how to use outbound links has been developed that can be tested. An people don't have to dig and search to get information to "use their own judgement".
So I go to the heart of it. If one should not use no follow, why not?
Now when I started programming in 1963 I learned the importance of testing a hypothesis. It has been many years since I have let others do that work while I do other things. But I still look for the test results.
Testing the use of the "nofollow" link is a whole different kettle of fish from testing your own program code, because you can't control the many other variables that go into determining how a page ranks at any given point in time.
If one should not use the no follow, why not?
If someone knows, it should be easy to answer.
But judging from the reactions, people are afraid to commit. That is understandable, especially when you are an expert.
So it seems to me that until someone comes up with some good information rather than opinions, when using an outbound link it is safest to not use the nofollow tag, if you can avoid using it, because not even the experts know if it will have side effects.
So outbound links bleed page rank
Outbound links improve page position
using a linking strategy something like wikipedia may be a guideline to use
If one should not use the no follow, why not?
Use nofollow for the links which are not approved by you and for rest do not use it.
I shall move on to doing some testing and then post an answer to why or why not based on results.
That is a guideline and not a why or why not.
Ok here is why or why not.
Google treats each link as a vote and it's guidelines [google.com]says,"In particular, avoid links to web spammers or "bad neighborhoods" on the web, as your own ranking may be affected adversely by those links.".
If you have blogs, forums or wikis where people can add links without your approval. Every content written on your site is under your responsibility so all these links are counted as recommendation by your site. Now if your site links to a lot of bad websites then Google can devalue your site. So to be safe either have a strict moderation or add a nofollow to all the links added by users. Phpbb has facility to exempt few sites while adding nofollow at runtime so you can remove nofollow for approved sites.
When you are adding links yourself you certainly want to claim the responsibility and points so do not add nofolow. Linking to good site will benefit the linked site for sure and it may also help your ranking.
Hope this helps.
AjiNIMC
Also, make sure the site has no hidden text! I have been seeing that a lot in google lately.
Gimp, also our outbound links do not use the no follow tag. We really do not care if google follows the links or not and we do not worry about passing on pr either.
If one should not use the no follow, why not?
You do not want to use the no-follow attribute in the href tag when you want googlebot to crawl through to a resource you think is the correct blend for your situation...in that it adds real value to your resource...and compliments or completes your dialogue with your target user base..(this is the basis of the hypertext linking philosophy)
Of course, if you base your link building (inbound and outbound) on your user (human) experience ... then the bots are likely to approve as well...(translated: Your user base will benefit from the links as will the destination sites you link to ... as your user base moves through the information and finds as complete a set of answers to their questions as possible...)
Well, I have a 2 year old directory with over 80k outbound links. To start with, NOTHING shady... What I mean by that is no 302ing, no nofollow tags, no ANYTHING negative. The links are all real links, directly to the sites in the directory. PR is passed to those who choose to be in the directory.
It is PR7 many sub pages PR6 and tons and tons of PR5s.
A legit white hat directory that is on-topic, human reviewed. I would also add it has a spider that only crawls what is in it's directory, and it deletes any link that consistantly returns a 404 or 302 for a week.
The PR has consistantly climbed through the past years and never suffered any sort of penalty.
I would also like to add, it's not king of the serps either. It does not get a ton of traffic. I guess because it ranks number one on a gazillion obscure search words and phrases, but that's the magic word "obscure"... Just not a lot of people searching for strange keywords or phrases.
I am not wild about the directory from a money making point of view, but I personally like it and use it myself, since I know none of the links are busted and can count on getting what I look for, when I look for that sort of thing. (spiritual themed) But it is no "road to riches".
The point being, it DOES mean that outbound links DO help you in the serps and positively does NOT drain ranking. Too many number one slots in the serps, and the PR is very high for it's genre.
By looking at Google's probable motives, one can possibly get an answer. Even though the answer is not exact, it might help now, or in the long term.
Votes for quality sites (by linking) would certainly be in Googles's best interest. Google also seems to reduce spam as much as possible, so anything that looks strange, like too many "no follow" links, may now, or in the future be detrimental.
It is also suggested that outgoing links that cannot be trusted as quality may be detrimental. Therefore, if you are vulnerable in this regard, "no follow" may help.
Lastly, by giving a decent amount of examples where sites have linked out naturally to quality sites, and not been hurt, one can make an semi-educated guess as to whether outgoing links are harmful.
As a reformed link-hoarder myself, I have not seen any indication that my rankings have declined as I have added outgoing links. I have actually seen a significant rise in referrals from Google, mostly long tail. I am not suggesting that the outgoing links were the cause for the improvements, as I have no way of testing.
Also there seems to be a lack of clarity here, beginning with the Title of this thread. "Page Rank" and "ranking" are two very different things. Page Rank is one specific measure of the interconnectedness of a URL with the rest of the URLs on the web. That's it -- a formula based on links between URLs and nothing else.
If PR were the only thing that mattered for ranking, then the idea of "draining PR" away from your own pages with outbound links might have some merit. But as many threads here will attest, PR is only one of many factors that affect ranking, and efforts to hoard PR can tend to harm other important areas that can be very important to ranking.
wouldn't it seem wise to make "natural" linking your default unless there's proof that an "artificial" linking pattern will benefit you now and in the future?
For most of the sites I handle all links are without nofollow except
1) Blog comment links
2) Forum links
3) Wiki links e.tc
I have added it because I am not moderating it to that level of accuracy. Even if there are some spams left to some of the old pages it should not worry the SEs.
Nofollow was designed to help webmasters with spam links (or user added links without approvals).
-----
MThiessen,
All your site proves is that a high PR site can easily rank for long tail phrases.
I can get a PR5 site with no outgoing links to rank just fine for 1000s of longtail phrases as well.
We're (at least I'm) not asking if linking out hurts or is neutral for a site, I'm asking for proof that it helps rankings.
-----
wouldn't it seem wise to make "natural" linking your default unless there's proof that an "artificial" linking pattern will benefit you now and in the future
EFV,
who cares what's "natural" and what isn't. What works?!
Can we go to a popular blog with 100s of comments to a single post and see hundreds of "nofollow" links and make any sort of deduction? induction?
Come 'on man, of all the people who I thought had thrown away their chicken bones, I assumed you were one of them. :)
--------
Lastly, don't take my comments personally. I'm just tired of seeing conjecture, hypothesis, and postulating going on SEO boards when someone asks a definitive question that can be tested (as has been by Trinorth)
Like Trinorth, I would like to see(hear) of people who didn't link out from a certain page(s) and then began to link out to certain pages and started gaining rankings:
for that page's content,
the anchor text they used in the outgoing link,
or whatever.
And of course, assuming they isolated as many variables as possible ie. the site didn't link back, they weren't actively gaining other backlinks for the page/site, etc.
I can wish for the world, can't I? ;)
And I would hope others would want this type of info as well
For those who thinbk that there are too many variables, please note that there are always too many variables. To get around them one sets up a control and a test. And the results are compared.
I have set up five controls on low PR pages and created five test pages.
I will let them fly and see what happens.
Trynor, as far as I am concerned, is on firm ground. He has results and is not giving opinions based on his reading of green tea leaves or a cast of the chicken bones.
My test might flop. But then it might not.
If it does, I will test again.
But until my results show something diferent than Trynor, I am using his approach.
We're (at least I'm) not asking if linking out hurts or is neutral for a site, I'm asking for proof that it helps rankings.
Everyone knows inbound links give rank, but do outbound ones drain it away?
To get around them one sets up a control and a test.
The algo is not constant. How does one set up a control over the fact that Google can make adjustmests at any given moment? It's like someone coming in during the night and rearranging your lab mice.
Not that one could not learn from specific tests, but one can also look at the previous behavior of Google, and it's probable motives, and come up with a approximate answer.
Neither method is fool-proof, neither method is superstition.
Try it out, if it does not work, try another one.
Google will never penalize you for good links.
Realize it might take a few weeks to show up.
As far as the variables goes, we did our experiment on an eccomerce site where neither the content nor html has not changed in six months. So, the only variables that might have changed were aging and may be the google algo. The site does not get many inbound links because it only sells a few products.
