Relative vs. Absolute Links

[Mac 5]

Does it have any SEO advantage to make links on your site Absolute, give the full path blahblahblah.com/page.html, instead of a Relative link /page.html.

Does a search engine give it any more weight like an outside link on another site?

[StoreFront]

Hi,
One of my co-workers went to a SES conference and she heard that it was always good to list your links as absolute, otherwise Google will not know who that link belongs to if it is just /aboutus.html.

This could be wrong, and Google is changing everyday, so it could be different now!

Hope this helps.

[MtraX]

Hi

"One of my co-workers went to a SES conference and she heard that it was always good to list your links as absolute, otherwise Google will not know who that link belongs to if it is just /aboutus.html."

I don't agree with this, if google's crawling your site they're crawling your domain and a page found on under that domain is yours (in most cases). I don't think Google has a preference on this.

Any external links will have a absolute URL anyway.

MtraX

[sem-seo-pro]

Hi

I believe that what was implied at the SES conference is indeed correct.

Actually the bot cannot think therefore relative urls are given no value.

By using the absolute url the link is given a weight the same as any other internal/external link the crawler finds.

[MtraX]

That's interesting...

Anyone else care to comment on this?

Regards
MtraX

[sem-seo-pro]

Hi MtraX

I should clarify I believe

A relative URL looks like this: "index.html". What this tells the spider is that there is an object called index.html...however what it doesn't say that may be important is what server it's on or what protocol it uses. Therefore the search spiders assumes from the page the link resides on.

An absolute URL looks like this: "http://www.mygreatwebsite.com/index.html". This tells the users web browser what protocol to follow (http/https/ftp) and on what server to find the object on (www.mygreatwebsite.com.

This method ensures the search spiders do not get lost or confused as they spider the page. So it really helps in that aspect.

Would this effect rankings?? I doubt it..only I believe... if a compettors page matched 100% to yours ...then and probably only then would absolute win out over relative.

At most were talking a position over a page in the results.

If that......

[TrafficProducer]

Over time Google Site Map May become more important, for Google than any other way of building links.

Most Spiders should manage relative and absoloute links the same.

You could do one way, on one site/page and another on a different site/page.

Search engine's may always change rules in the future.

[bhartzer]

Does it have any SEO advantage to make links on your site Absolute

I'm not sure if it was me that said that at the SES conference or not. ;)

Anyhow, always always always use absolute links whenever you can. If you use relative links you can suddenly lose all of your rankings in the search engines--it has happened to me and was a nightmare for a few weeks until I got it all straightened out.

There's not actually any rankings boost per se when you use absolute versus relative links. However, when you use absolutely links you make sure the search engines know which "version" of your site to index.

You see, if someone links to your site using domain.com rather than www.domain.com, the search engine will follow the link and start crawling. If you use absolute links then the search engine will crawl your site and only see the pages as www.domain.com/page.html. If you use relative links the the search engine could crawl your site as domain.com/page.html, which is probably not what you want.

Most of your links from other sites will be linking to www.domain.com, not domain.com. If the search engine decides--for whatever reason--that they index your site with domain.com and not www.domain.com, then you could suddenly lose all of your search engine rankings because the non-www version of your site is indexed and the www version is not and you don't have many links to the non-www version of your site.

It's always good to be consistent in your internal linking, and it's good to make sure that all of your internal links specify exactly the domain name of your site so there's no question of which pages to index. There are other reasons, as well.

What happens if you use relative links in your site? What happens if someone decides to download an entire copy of your website and put it up on their domain (trust me, it's happened to me!)? If you use relative links, it would be easy to copy your website and put it up on another domain. If you use absolute links then it's much more difficult--the domain name has to be removed or changed on all the pages of your site. There are people out there who try to copy websites all the time--and making it much more of a chore for them to copy and change decreases your chances of the site being copied. Or, if they do copy even one page of your site and put it up on their domain name then they might not realize that your full URLs are specified--checking your web stats might reveal the fact that they copied your page's content--it's happened to me and I was able to catch the culprit because of my own absolute linking.

I realize that in certain web development environments and the way web designers work on websites it's very difficult to use absolute links. For example, many web developers and designers set up a test area when they're working on a new web site or a new revision of a site. As a result, they must use relative links because the site 'wouldn't work' because it's temporarily in another location. Some developers and designers use clientname.theirdomain.com to test the website. In this case when the site goes live the links should be changed to absolute links.

[bhartzer]

If you'd like to read up more about the subject, there's a good article here.

[sem-seo-pro]

Bill

In the link you send users to there is this mention near the top.

The search engines automatically convert relative links into an absolute links.

Sharris a great author, but wrong on that I am afraid. SE spiders disregard relative URLs.

If that were true then SEs could change sym links to canonical which they cannot.

Bill yours was a better post to be honest.

Clint

[google junky]

bhartzer,

I do have a habit of adding and subtracting the domain from the url if the domain has keywords in it to maintain keyword density within the code.

Would you consider that a factor?

[MtraX]

Hi Google Junky

I don't think it will carry great weight, allthough it can't hurt (same reason it's been recommended to use keywords in file names?). I'm slowly getting convinced that it's good to use absolute URLs, but it is a serious pain since most developers I know want to run a test version first.

I'd much rather just do a permanent redirect to solve the canonical url issues.

MtraX

[TrafficProducer]

SE spiders disregard relative URLs.

Read (For PERL users) (Thier are othe options and languages to do this)

HTML::LinkExtor Extract links from an HTML document HTML-Parser-3.48 - 02 Dec 2005 - Gisle Aas

HTML::SimpleLinkExtor Extract links from HTML HTML-SimpleLinkExtor-1.11 - 23 Aug 2005 - brian d foy

[williamc]

Originally Posted by bhartzer

Anyhow, always always always use absolute links whenever you can. If you use relative links you can suddenly lose all of your rankings in the search engines--it has happened to me and was a nightmare for a few weeks until I got it all straightened out.

I use relative urls more than anything else. The only time it would hurt you any is if google comes in via www.domain.com and lists it as www. yet you are searching for just domain.com. Now in that event, google may devalue or derank pages in the domain.com as duplicate content, which would cause you to see rankings disapear, agreed. However, a simple mod_rewrite that takes anything coming in to domain.com and redirects it using a 301 to www.domain.com, then you will never even see that issue.

Many of us develop sites on a subdomain or in a directory under our development site, then move it to its true home afterwards. In this case, relative is the way to go for ease of install later.

Not to mention doing it that way helps you steer clear of the sandbox later :)

[williamc]

Originally Posted by sem-seo-pro

SE spiders disregard relative URLs.

Step away from the crack pipe slowly with your hands in the air....

Not a single one of the large engines disregard relative URL's. They all handle them just fine and have for over 8 years now.

[TechEvangelist]

I also have to disagree with the statement that "SE spiders disregard relative URLs'. I've been working with SEO since 1997 and have seen only a couple of sites that used absolute URLs. I have never seen a site disappear from the rankings unless there was some other issue with the site.

Just check the sites that rank in the top ten for any search. It is difficult to find a site using absolute URLs, so it would seem logical that SE read relative URLs just fine.

There are some very valid reasons for using absolute URLs in certain situations, but I have never seen any evidence that spiders have a difficult time with relative URLs in a properly coded site.

[incrediblehelp]

I prefer to use absolute, but in the end it is not a big deal to use either or.

The only issue I have with relative URLs is with the home page. I always tell my clients to refer to the home page as http://www.domain.com/ and never as the default document /index.html or http://www.domain.com/index.html. Of course this is even being a little over cautious.

I just like the SE to always think of the home page as the root domain and never the default file. This also avoids potential issues down the road if you want to change you default home page at the server level for some reason. If the SE's always reference the domain they will never be confused.

[TrafficProducer]

Just a note: FrontPage Relative Links ARGH!!!

FrontPage can be a PIG with copying Relative Links from one page to another page/site kept on your Had Drive before uploading.

It tends to add a link to a local file on the Hard Drive e.g. file:///C:...

If these are uploaded they will fail because the Server cannot find the link to your local Hard Drive link.

[Faglork]

Originally Posted by williamc

Step away from the crack pipe slowly with your hands in the air....

ROFL. Good one ...

If statements like "SE spiders disregard relative URLs" are given, they should be backed by evidence. Otherwise, it should be made clear that this is an opinion and not a fact.

faglork

[sem-seo-pro]

Facts I love them too

First it is as much fact as can be.. a relative link produces foo.html which does not tell the search spider to which site specifically foo.html belongs therefore the spider 'assumes' the link is relative to the website.. but since bad things happen when you assume the spider therefore disregards the foo.html link and does not give it weighting.

An absolute link http://www.mywebsite.com/foo.html tells the spider that the link absolutely refers to website and page and therefore the spider is safe in its knowledge and applies weight and score to the link.

Dear ***,, when we assume things it makes an ass out of u and .....

not me... since I do not assume as you do,,,

Faglork wrote

williamc wrote:
sem-seo-pro wrote:
SE spiders disregard relative URLs.

Step away from the crack pipe slowly with your hands in the air....

ROFL. Good one ...

Fact - You and williamc have the same maturity level as a teenage schoolgirl...

And as Bill Murray said in Stripes - That's a fact Jack!

[sem-seo-pro]

Originally Posted by Faglork

ROFL. Good one ...

If statements like "SE spiders disregard relative URLs" are given, they should be backed by evidence. Otherwise, it should be made clear that this is an opinion and not a fact.

faglork

Dear ***

You're a moderator who condones flaming???

Where is Brett in thinking that's acceptable????

Next you do not know relative links have no value to a search spider and you are a moderator ???

LOL

[Feydakin]

Ok, not a crack pipe, needle junkie??

I have always used relative URLs.. Never occured to me to do otherwise.. And it's never had an effect that I can see.. Do you have 'any' statistical data to back this up??

[sem-seo-pro]

No statistical data would be involved in such

Look at a scripts installation instructions for why relative links are disregarded so to say..

We always need the absolute path in order for a script to work properly.

I dont know if it makes any sense and maybe my wording is poorly chosen what I am trying to say is foo.html has no meaning or signifigance to the search spider..

yes the link works and will be followed but it has no value that it might otherwise when in absolute mode.

My drug of choice is life




[/i]

[incrediblehelp]

Originally Posted by sem-seo-pro

No statistical data would be involved in such

huh? Sure it would. Show us a website(s) that the SE's are treating relative links with "no value".

No need to flame you here, but when you say relative links don't have any value you in the search engines what exactly do you mean? Everyone just needs evidence for this statement your making. I sure any of can show you a websites that is using relative links that is doing fine in the search engines.

Do you mean the relative links are not getting spidered?

Do you mean the relative links don't posses IBL value?

Do the search engines have no idea what root domain to match the relative links to?

It is my opinion that all of the statements above are un-true and I have seen websites with relative links do very well in the search engines.

I just prefer to use absolute.

[Feydakin]

Originally Posted by sem-seo-pro

We always need the absolute path in order for a script to work properly.

Really?? I've always installed scripts in "/" or "/gallery" etc.. I don't recall ever needing to add "http://www.domain.com" when installing a script..

As a counter to your argument.. Have you ever been able to navigate away from a domain without a literal URL for it to follow?? I can't think of any way that that would be possible.. So, if you can't navigate from a domain without a literal URL, why would the SEs "assume" that a relative URL would NOT belong to that domain??

Steve - finally got 2 days off in a row since before Thanksgiving!! :)

[williamc]

Originally Posted by sem-seo-pro

Facts I love them too

Great, start using some for a change please.

but since bad things happen when you assume the spider therefore disregards the foo.html link and does not give it weighting.

Funny, seeing as how I have websites ranking just fine and dandy using nothing but relative urls, with the nav system being most of the inbounds the engines see. The engines have no problems at all reading and properly handling relative url's and have not for years before some sales clerk allowed you to purchase a computer.

An absolute link http://www.mywebsite.com/foo.html tells the spider that the link absolutely refers to website and page and therefore the spider is safe in its knowledge and applies weight and score to the link.

Read above.

Dear ***,, when we assume things it makes an ass out of u and .....

not me... since I do not assume as you do,,,

Nope, you don't assume... Sadly, I have to agree with this. You just honestly think that this stuff your saying is true, and not intentionally misleading people. But the fact, jack, is that you are indeed misleading people with utter nonsense.


Oh and for the record, the crack pipe was not a flame. It was just a comical way of saying, what the hell are you smoking????

[williamc]

Originally Posted by sem-seo-pro

Look at a scripts installation instructions for why relative links are disregarded so to say..

We always need the absolute path in order for a script to work properly.

I find it funny you use this to back up your claim, as I run 2 of the larger script sites around and none of the 200+ perl, tcl, python, and php scripts we have written need absolute paths :)

Again, give me some of what yer drinking mate :)

[sem-seo-pro]

absolute URLs. They identify a resource independent of their context.

Relative URLs are a way to identify a resource relative to their context.

Next we have a 10,000 foot ravine to cross.

There are two paths to cross the ravine:

Caution falling will result in death. We're absilutely sure of this.

Path one says: It has been determined this path is the absolute way to cross and not fall to your death.

Path two says: We are relatively sure this path is okay to use, however you have a slightly greater chance we are wrong and you will fall to your death.

Which would you choose?

If you were to bet which would you pick an absolute winner or a relatively good chance of a winner?

Would you want a jury deliberating your guilt in a criminal trial to absolutely sure not guilty or one that is relatively sure your not guilty??

Would you want your lawyer to be absolutely sure you were innocent or relatively sure??

What would the attorneys confidence level be if he was absolutely sure or relatively sure?

Why would I need to study statictics about this?, when I am absoultely sure I prefer always having things absolutely sure as opposed to a relatively good chance.

Since search engine spiders are acripts they cannot assume anything about a document that is not absolute so it does not give it any weight or value.

Also when we are speaking of links which all have a value score according to Google ,

PageRank Explained

PageRank relies on the uniquely democratic nature of the web by using its vast link structure as an indicator of an individual page's value. In essence, Google interprets a link from page A to page B as a vote, by page A, for page B. But, Google looks at more than the sheer volume of votes, or links a page receives; it also analyzes the page that casts the vote. Votes cast by pages that are themselves "important" weigh more heavily and help to make other pages "important."

Important, high-quality sites receive a higher PageRank, which Google remembers each time it conducts a search. Of course, important pages mean nothing to you if they don't match your query. So, Google combines PageRank with sophisticated text-matching techniques to find pages that are both important and relevant to your search. Google goes far beyond the number of times a term appears on a page and examines all aspects of the page's content (and the content of the pages linking to it) to determine if it's a good match for your query

Now given the above don't you want the search spiders to be absolutely sure??

Lastly we know that relative URLS are basically a shortcut and when people take shortcuts this usually leads to an increase in the amount of mistakes made..so the question then becomes,,,which do you want??

Clint

[sem-seo-pro]

williamc said

Funny, seeing as how I have websites ranking just fine and dandy using nothing but relative urls, with the nav system being most of the inbounds the engines see. The engines have no problems at all reading and properly handling relative url's and have not for years before some sales clerk allowed you to purchase a computer.

Fact is there are 100 parts or more to the algorithim and your claim of well rankings does not come from nor is supported by the use of relative links.

My system is a home made dual head 2.8 ghz 1.0 ram 120 gig hard drive asus motherboard wtih two dvd drives and a cd drive no floppy four usb ports with a wireless D link network and two Snycmaster 710MP monitors

It will smoke your peoplepc for breakfast ;->

Peace

[williamc]

Originally Posted by sem-seo-pro

PageRank Explained

I find it extremely funny that you quote an article that was based off Webworkshops PageRank explained whitepaper. My home port is webworkshop, and you sir do not even remotely understand pagerank, let alone any search engines.

As to what system you have, unfortunately sales clerks will sell any PC regardless of configuration to any imbecile that has the cash.

As for it smoking my "peoplepc". It would have to smoke the 6 Dell's and 4 HP Media centers we have on our local network. I just do not see that happening anytime soon with that lower end system you describe :)

aren't 2.8ghz chips obsolete yet anyway?

To be brutally honest, any one of our 4 web server machines would smoke that, and the web servers are not as good as our locals.

Before starting a "my weenie is bigger than yours" fight, you better know who you are dealing with first mate.

[sem-seo-pro]

williamc

You started the weenie toast and your article site wasn't the original author nor said first publisher of any articles I may have cited.

I assure you I spend little if any time on a business you own as in my eyes you are a crass little boy.

Funny... I don't understand search engines yet I wrote an article dated 2-14-05 on seochat.com stating buying links was a bad practice... and here 10 months later we have the Google engineer Matt Cutts telling us the same thing.

I dont understand search engines??

Funny??

I guess...

Maybe you should put the crack pipe down if you think someone with that vision doesnt understand pagerank and search engines.

Peace to you and your weenie.

and I have mates over in the UK

...you are not a mate of mine.. so please refrain from implying such

Clint Dixon

[williamc]

Originally Posted by sem-seo-pro

You started the weenie toast and your article site wasn't the original author nor said first publisher of any articles I may have cited.

Again, you have shown you can not read mate.

I did not say Phil was the "author" of that article. I said it was based off Phils Pagerank Explained whitepaper which came out far earlier.

"open eyes, lest insert foot"

As to the weenie fight, not toast, as that would imply something completely out of sync with the statement: No, actually I just made a statement that stores would sell a PC to any imbecile. I did not say a thing about my system or "weenie" at that time. Only you did with your system stats and the peoplepc comment. It was not my fault you use a low end system, so please don't blame that on me.

Again, please, please, dont let the facts stand in your way, I have yet to see you do that and it would destroy my image of you if you did!

Funny... I don't understand search engines yet I wrote an article dated 2-14-05 on seochat.com stating buying links was a bad practice... and here 10 months later we have the Google engineer Matt Cutts telling us the same thing.

I dont understand search engines??

Maybe you should put the crack pipe down if you think someone with that vision doesnt understand pagerank and search engines.

Vision? I have seen articles out saying the same thing for over 2 years now, yet you somehow think that yours merits any "vision"? Give me a break. So you cried wolf along with 50 others 10 months ago, just like others years ago. Google still can not find most paid links, so they still work just fine. What vision exactly?

Matt says cloaking is a bad idea too, years ago, so did yahoo, msn, altavista, et al. Many cried wolf and wrote articles saying how the engines would find and terminate all cloaking too. Funny thing is, that unless a competitor reports it, it generally still works. FTR: I also say cloaking is a bad idea. But it works regardless.

Search engine "vision"? More like you have blinders on to me.

[williamc]

Okay for all those who really care about the original topic: It has been shown by quite a few people that relative url's are just as good as absolute ones and it really comes down to personal preference. Both have their merits and ways to accomplish the same things.


For those who are laughingly following the sem-seo-poop stuff, it continues here with him showing his masculinity and ability to not grasp simple concepts some more: http://www.webproworld.com/viewtopic.php?t=58581

Anything said above is my own opinion only and not based on any knowledge of metally challenged handicaps at all.

[SPC]

I see you've dropped yourself in it again, Clint (seo-sem-pro). Why doesn't that surprise me?

Here's the definitive answer to the original question:-

The URLresolver reads the anchors file and converts relative URLs into absolute URLs and in turn into docIDs.

The URLresolver is part of the Google engine. That's an extract from Brin & Page's own document, and can be found in section 4.1.

Honestly Clint - you really should stay away from forums.

[williamc]

Bah, now theres no debate. :P

[SPC]

Sorry :(

[SPC]

Ok. Just for you WilliamC - some more debate...

Originally Posted by sem-seo-pro

No statistical data would be involved in such

Look at a scripts installation instructions for why relative links are disregarded so to say..

We always need the absolute path in order for a script to work properly.

I dont know if it makes any sense and maybe my wording is poorly chosen what I am trying to say is foo.html has no meaning or signifigance to the search spider..

Clint is partially correct. Some serverside scripts do need absolute paths - but they are usually absolute server paths and not website paths, are they not? The path to sendmail is one such path that Perl scripts need. Shame on you WilliamC for not bringing that to mind ;)

However, what a serverside or clientside programme (script) needs, and what the requestor of a file needs (browser, search engine spider, etc.) are totally different things. The two are not even remotely related, and cannot be compared with each other.

So I am afraid you are wrong to use scripts (programmes) in your argument, Clint.

[williamc]

Originally Posted by SPC

Clint is partially correct. Some serverside scripts do need absolute paths.

Shame on you WilliamC for not bringing that to mind ;)

Bah, I did not even catch that. Must have still been hungover from new years parties :P

Good catch SPC, and you are correct. Clint just used it wrongly to attempt to support his false claim.

[brian.mark]

Originally Posted by sem-seo-pro

Dear ***

You're a moderator who condones flaming???

Where is Brett in thinking that's acceptable????

Next you do not know relative links have no value to a search spider and you are a moderator ???

LOL

He was laughing at a comment. He's participating, not flaming. Please don't confuse the two.

Also, as is shown later in this thread, relative links do have value to a spider, so if anyone were agreeing with you they'd be the ones that lacked knowledge.

As mods, we wear 2 hats - members and moderators. He's wearing his member hat on this post. The moderator hat comes when threads are locked / quarantined / otherwise modded. We aren't iEntry employees, just volunteers with an interest in the community and the topic.

As for Brett... where do you think you are? In that "I'm too good to be crawled by Google" forum? We've got Mike, Rich and Chris here. No Brett. Sorry.

Brian.

[sem-seo-pro]

Here is the final word on the subject for those still confused!!

http://www.mattcutts.com/blog/seo-ad...onicalization/

[Faglork]

Now YOU are confusing "relative / absolute URLs" and "canonical URLs".

These are not even closely related (apart form being URLs).

faglork

[brian.mark]

Originally Posted by sem-seo-pro

Here is the final word on the subject for those still confused!!

http://www.mattcutts.com/blog/seo-ad...onicalization/

I think you must be referring to "Spike" and his comment on the base tag. That's not really relevant either, nor do I know who he is or why you think he's an authority.

Brian.

[Faglork]

Originally Posted by brian.mark

That's not really relevant either,

... apart from being incorrect.

faglork

[williamc]

So much for his attempt at the last word eh?

[lostboy669]

Hi,

I have been hearing this crap about absolute paths, but I have always used relative becouse of simplicity when developing sites, and I have never come across any search engine problem, but now I'm developing a site that needs switching between http and https and theres no way some elements in the http could be secure, so it's an absolute need to use absolute paths.

So for my point of view use them at your own discretion also eat fruits and vegetables.

lostboy

[incrediblehelp]

What the SE's say is it is best to be consistent. Use one or the other, not both at the same time. Myself I feel better about absolute. An old thread on this:

hardcoded url versus relative path

[Ne0]

ABSOLUTE! not only is it a GREAT Vodka... it's just a good practice...


.02

[idansh]

LOL that was funny