Groovers, Because of the e-commerce system we use every page can be accessed through two URL's http://ecomsystem.com/myid/productpage1 and http://oursite.com/myid/productpage1 In general the 'ecom' pages rank a lot higher which, I assume, has to do with the thousands of pages on that server (100+ shops) and lots of traffic. Those pages get spidered a lot more often too (nearly daily now) whereas on 'oursite' it can be a week for new pages to appear. To the point: What I was wondering about is this. For the sake of backlinks, is it better to use absolute href's (<a href='http://oursite.com/myid/productpage2'> or can I leave it relative (<a href='/myid/productpage2'>) just so the spider can follow the link on whichever server it happened to be? What's puzzeling me is that for instance both http://ecomsystem.com/myid/productpage2 and http://oursite.com/myid/productpage2 have exact same content and nearly identical backlinks yet different PR and SERP's Any suggestions what would be the best linking practise?
you can use either. google...for example...will always store the link as an absolute. however....always refer to the index page as "http://www.oursite.com/" ...and never leave out the "www" as in your example. how ever subtle the differences may be...they are enough to cause the variation in pr and serps. google sees them as independent pages.
I'm not sure I completely understand your example, but if you really have two domains or two sites with identical information you are running a risk. Google does not like identical sites and what they normally do is decide to completely ignore one of them. Why bother with two domains? Why not select one or the other and spend all your efforts optimizing and promoting that one, rather than dividing your effort between two domains?
Here is an example showing Google doesn't seem to care, or doesn't notice: http://www.google.co.uk/search?sour...F-8&q=blata+ninja&meta=cr=countryUK|countryGB Notice #2 & #3 vs. #4 & #5 They are all our pages and two identical sites. It's not a matter of having to divide efforts; you log into the ecommerce system, update your content, publish it and tadaaa, both sites are updated. They are actually both hosted on the same server though I'm not sure whether they have the same IP address or not. The pages accessed on the domain starting with the ecommerce packages url tend to rank a lot higher than the exact same ones within the domain we chose. This must be down to the 'pages in URL' with G preferring higher numbers though it doesn't seem fair to me. I use in 99% of the cases relative backlinks and most incoming external links refer to our domain as oppose to the same page on the ecom domain. So I thought when G follows inbound links and starts spidering from there, it will follow the relative links, adding our domain in front of it, this making them possibly rank better yet the opposite seems to be the norm. On the same token I find it strange that G shows just 111 pages in URL on our domain yet 255 pages in URL on the ecommerce domain /our id. After a month still a mystery to me.
Would that not depend on how one has the site set up? By which I mean that if one has, as I do, mod_rewrite set to 301-redirect any inquiry to a "www."-preceded URL to a call to the form without the "www.", then it shouldn't matter how I refer to the site--or how anyone links to it--though for consistency it makes sense to use the redirect-to form when referring to the site. Over on SEO Guy's forum, someone was questioning his new host's setup which, as best I could understand the question as posed, was forcing calls, regardless of how made, to the "www." form, and which (if the fellow did it right) he could not affect with a per-site .htaccess file. That seems rather bizarre: I am under the impression that most host servers will indiscriminately accept and refer calls with and without the "www." to the same page, and--if one has not done anything via .htaccess--it is the form used by the caller that will show in the caller's browser URL bar (which is why Google and the others can see a page as two distinct pages, and why one should always have a mod_rewrite for "www.", whether to force it in or out being a matter of taste, but one or the other needs to be the sole acceptable form). Yes? No?
I actually do always refer all absolute links with http://www. but I sparfuchsed it out of my post out of laziness. It happens more often than I'd like that people on the phone or face to face spell out their whole website address including the doooobaaaia doooobaaaia doooobaaaia dot etc.. Just bores me but (shame shame) reading this and testing it on our own domain I see a not-so-nice Cannot find server error! Better ring our hosting provider first thing in the morning and get him to do a mod rewrite. But then again, it only happens when you manually type it in... How many people do that?
Sorry, I'm not getting that. Are you saying that there is a difference in how a URL call is handled depending on whether it is a click-on link or whether it was typed into a URL bar? Or is it just that if it is a click-on link you feel that you have sure knowledge of what that link was set to? The first seems impossible, but the second not far from it: linkers can, and do, supply whatever they choose to (especially amateur webmasters): URLs with and without "www.", bizarre anchor text (I think my worst link anchor text is "2"), and so on. Can you clarify?
You are right, it obviously isn't just when they type it in. It is however they write link in the '<a href=' tag. That indeed makes it more likely to happen that not-so-sharp webmasters might link to it without www. Thanks for pointing it out to me.