Video Transcription
Howdy, SEOmoz fans. Welcome to another edition of Whiteboard Friday. This week we're talking about the cross-domain rel=canonical. It's a very interesting attribute that we can now add to web pages.
Last year, Google created the ability for the rel=canonical, which has always enabled us to canonicalize duplicate content or to tell Google, hey, these pages are actually copies of this particular page, and now we can do this across different domains. For example, we can say, oh, the New York Times runs IHT, the International Herald Tribune, and IHT.com features many stories that were originally published on New York Times. The IHT sort of goes, well, we don't really deserve the search traffic for this. It's a great resource for our readers to be able to browse on the IHT site, but what we actually want to do is say, search engines the original article exists here on the New York Times' website, let's point to that. This cross- domain rel=canonical enables webmasters of all stripes, big, small, medium, to be able to do this.
I want to point out Tony Adam, who is a great SEO in Los Angeles and wrote on Visible Factors just this week about the cross-domain rel=canonical. He pointed out that in some testing he had done the cross-domain rel=canonical looked even faster, even faster and more effective than a 301, which has been the traditional way that SEOs and content producers on the Web of all kinds have repointed content.
Now, what a 301 does is it means that not only do search engines, but visitors as well, get pointed to the new version. So, it is a little frustrating if you are, in the example I pointed out earlier with the IHT where the IHT says, well, we still want to keep that visitor on our site reading this story, we just don't want search engines to be confused about which one is the original. So, the New York Times can have that originality through the cross-domain rel=canonical. It looks like search engines are really respecting that.
Props to Tony for doing some experimenting there. I wanted to point out two experiments, interesting ones, that we've actually done here at SEOmoz and the results that we've seen.
A lot of times, you have this call, like I have page A here and I want to redirect it to this other page over here. So maybe this is A-X and I want to redirect it to content A-Y. I can use the 301 or I can use the rel=canonical. I sort of have a choice to make. Now, there are a lot of times when the rel=canonical might make good sense, like that IHT example. What we've been seeing is that the cross domain is a really powerful way to make this work.
I'll show you a first example here. What we've seen is that taking an individual post, so we produced a blog post on SEOmoz a couple weeks ago called "The Story of SEOmoz." It was actually originally written by Robin Good. He transcribed some video of a presentation that I made in Rome, and he posted it on his website. It's under, I think it's called, "Entrepreneurship: The Story of SEOmoz," is the name of his post on his website. It's got video. It has all this great content.
We took that content and we said, man, Robin, we really love what you've done here. This is fantastic transcription work and all the video that you took. This is great. We'd like to reproduce that on SEOmoz, give you credit for it, and we want to make sure the search engines know that you originally made the article as well. So Casey Henry on our team actually implemented a cross-domain rel=canonical to make sure that this version is the one that shows up. If you search for "the story of SEOmoz" or "Entrepreneurship: The Story of SEOmoz," you will find some older blog posts of ours and some different ones, but that one won't show up. Robin's on his Master New Media website will show up. That's pretty cool. Showing the power of that cross-domain rel=canonical in action and really good use of it. Sort of a white hat way of saying, hey, we want to reproduce your content, we want to license that content, but we want to make sure you get the search engine credit because you were the publisher. So, very cool.
The second example is even niftier and suggests some very cool applications as well, and so I want to point this one out. I was frustrated because for the last few years a very old domain that I created, I don't know, back in the late '90s, early 2000s, Randz.net was ranking really well for my name. I think it was ranking number 3 actually for my name, for Rand Fishkin in Google. I was always kind of frustrated because it's an old domain. I haven't updated in forever. I need to do the WordPress reinstall. I don't even know where the server login is. Whatever. It's kind of defunct at this point, and I haven't updated it in years. But I have this new blog, RandFishkin.com/blog. I really wish I could this one ranking because it has some good content on there, a bunch of posts that have been on Hacker News and some interesting things. It's much more current and updated. I do once a month at least put something new on there. So, what I did is I took very page in the header of the WordPress template, I took every page and I put a cross-domain rel=canonical to this URL. So every page at Randz.net now says canonical version is RandFishkin.com/blog. You know what happened? Two days, literally 48 hours, like the next time they crawled Randz.net, bang, RandFishkin.com/blog ranking number 3 for my name. It hadn't even ranked on page 1 or 2. I think it was on page 3 or 4 up until that point. So, just awesome to be able to put this, the page that I really want in the search results and kind of retire my old blog from being searchable.
This suggests that there are lots of good things you can do with this. If you have content and you sort of, oh, I have it available but I can't do the cross domain 301ing, hey, I can use a cross-domain rel=canonical to do that. If I have content that other people are reproducing from my site, I can ask them to put it on there when they license content from me. If I have content that I am producing out, I can make sure that people rel=canonical back to my site. There are just great uses for this cross- domain rel=canonical. Very powerful tool. I am excited because I think over the next few months we're going to see some really creative uses, some good white hat ways to do this. I am excited to see what the SEO world comes up with in terms of strategy, and we'll certainly be reporting on some of that here on Whiteboard Friday.
All right, everyone. Thanks. Take care.
Video transcription by SpeechPad.com
So did I understand correctly, that you are now ranking for your name higher because of the rel canonical tags on the old site, even though the content from the old site is not on your new blog?
Great question, I would have expected this to only bring benefit if you'd duplicated the content and then used rel=canonical to show which should be considered as the original version?
This was what I was going to ask. How similar was the content on the two pages?
If the content is really different on both pages this could be used for quite some pages, maybe even for pages with thin content. I'm testing this right now with some old domains and looking forward to see if google pick it up and will show some ranking changes.
It would be very nice to hear more details of the content on both the domains. Is only the subject the same or also some of the articles?
update: one of the domains was ranking #1 for it's main keyword (exact match domain) but isn't ranking anymore after I implemented the cannonical. The other domain isn't ranking for those keywords either so it doesn't seems a way to just transfer rankings between domains.
Rand Said that Google understand this tag and give priroty & authority to Rel Tag so no dublicate issue arise here
there is already a fair amount of abuse here. Clever affiliates are hosting search shopping sites for big magazines and getting them to link to the "magazines" own ecommerce section. Then they canonicalize to their own tld and just steal the link juice from big magazines
The Big Magazines publishers are just getting pwned by these shopping engines
For example. Vogue's style.com is hosting a widgets that link to stylecom.shopstyle.com but then they cross subdomain cannonicalize to www.shopstyle.com essentially stealing link juice from Conde Nast turning it into one ginormous link farm. Maybe this is why magazines are on their last legs they have nobody technical when inking these 'ecommerce' deals
Check the widget at the bottom of this url:
https://www.style.com/trendsshopping/shopthelook/category/clothing/
Thanks for sharing this. I'm not going to say there aren't very valid and very, let's say, "creative" ways to use this. I did just think it was worth pointing out that Mr. Cutts also covered this quite recently in a Webmaster help video on Youtube when not really answering a question I asked him.
You raise a lot of the times that rel=canonical can be used (and would be better used than a 301) but it is perhaps worth mentioning that Google talks about a number of reasons why (where possible) a 301 redirect would be preferred not the least of which because whilst Google may honour rel=canonical the other engines may not necessarily do so.
Ultimately I think it's about intent. As you say, if you want to share something and keep the user on your site but give credit where due to the creator of the content that's great. If you don't have access to change server headers (because of your CMS) also a great option.
If you wanted to their are also loads of ways that using the rel=canonical can and has been used to benefit a different site/provide a boost as you referenced with the randz.net example but just thought it was worth sharing that as a "best practice" it seems as though Google would rather have you 301 where possible - which likely means the other option would be better for SEO performance.
*Worth mentioning that the question was originally asked when looking at site migrations and having watched the power of the 301 redirect "boost" from old domain to new seemingly wearing off after about 3-4 months and I was trying to figure out if the same would occur over time with the rel=canonical effect. I'm tempted to say it would not (though Matt didn't say either way).
Couple observations/questions...
For the "many to one" on the old domain, how long ago was that?
It would be interesting to see whether the engines choose to disregard over time based on the content not being the same or even similar. And related to that, whether searches for content from the old domain bring up the individual URLs from the old domain or bring up the homepage of the new domain.
The interesting part may be if the engines do disregard (or you remove the canonical link elements) but the new site continues to hold its position and this served as a booster.
In regards to Tony's test and related to the comment I left on his blog...an important element and possible flaw in the test is not accounting for when the content was crawled. To truly test which is faster, then the clock needs to be started upon content discovery/rediscovery, in other words, when it is crawled/recrawled.
For instance, URL "A" & "B" both get updated at the same time and "A" gets the cross-domain canonical and "B" gets the 301 on Monday. Let's say that "A" gets crawled on Monday and the new URL from the canonical overtakes the ranking on Wednesday, which appears to be faster.
But what if "B" doesn't get crawled until Wednesday and the new URL from the 301 overtakes ranking on Thursday. Based on crawl to change time, "B" would actually appear to be faster, even though "A" appears faster based on the initial push.
This is exactly what I needed to get one of my clients to use the rel=canonical tag on a sub site directing to their main site. Great stuff, thanks.
Cross-domain canonicalization sounds great for syndicated content. Is Google treating this different from their original-source and syndication-source attributes that they use in Google News? Or is this a sign that they will begin to phase out the source attribution tags?
Great WBF. I'm new to SEOmoz and must say that I'm already addictd to WBF. Keep up the good work Rand!
Rel=Canonical sounds great and all but only if you can access the header of the page to insert the meta tag. What happens if you are syndicating content to multiple websites but don't have access to the header? This is the situation I am in. I have many related blogs that I synidcate content to but I only have access to a WYSIWYG editor and no way to inject the canonical tag into the header. Anyone know if Google will honor the canonical tag if its not in the header?
Any thoughts on using syndication-source or original-source tags instead of rel canonical?
I've been testing cross-domain canonical in a situation similar to your Randz.com vs. Randfishkin.com scenario, but I have to ask in this case - even though it may be more effective, wouldn't a 301-redirect be more appropriate in this situation? I assume you want visitors who happen across the old blog to go to the new one, too. With a cross-domain canonical, they won't see that intent (they'll still be taken to the old blog).
Of course, if your traffic to the old blog is 99% organic search, then that's pretty moot.
I'm curious if anyone has tried doubling up, to see what happens (use a 301-redirect AND a cross-domain canonical). If the canonical tag kicks in first, the 301-redirect still should be read over time, and then you could remove the canonical later. If the 301 kicks in first, the canonical might not be processed, so it's a bit tricky (but still no-lose). I'm always cautious about doubling-up two tags that do the same thing, but it might be interesting to test.
If you set up 301 and rel=canonical at the same time, then the rel=canonical will not be seen since the response from server will be 301 rather than the page HTML.
That's what should happen, in theory, but it almost seems like rel=canonical is being processed faster by Google and given more priority in some cases. It would be interesting to test them head-to-head on the same page, although I'm not clear on how you'd tell which one kicked in faster (since, as a user, you'd always see the 301, and the end-result for Google would be the same).
Hmm, this is where I am getting confused on what do you mean and wondering if we are talking cross-wired.
If at the exactly the same time I add rel=canonical on page A to point to page B, and (at the same time) also set up 301 redirect from page A to page B, then googlebot request for page A would get 301 response to page B and hence googlebot would not be able to see rel=canonical. This is why this cannot be tested on the same set of pages.
If you want googlebot to see rel=canonical, then you would have to implement rel=canonical first, wait for the page to be crawled and only then set up 301 redirect, but then this would not be the same test.
The only I can see to test something like this is to have two sets of pages (lets say A1 and A2), on A1 set do 301 (to B1) and on A2 use rel=canonical (to B2), and hope that A1 and A2 have been crawled by googlebot on the same day. But in such case the other outside factors may influence the results.
Has anyone looked into how easily it would be to reverse the result of adding this tag?
i.e. You add a tag to a site and then down the road, you want to develop the site again and take the tag off.. I would be curious to see how fast google reacts to taking away the tag. Also, if there is any long term damage done by adding the tag in the first place.
Any thoughts?
Fantastic WBF this week. I always used a canonical as a last resort when a 301 wasn't possible. It definitely makes sense for cross domain canonicals to be used when syndicating content out to other sites. Absolutely love the tests with randz.net, makes this info even more valuable!
Just a quick heads up. I visited randz.net this morning to check out the tags, and Google is throwing a Malware warning suggesting the site is including content from metrics.performancing.com "a site known to distribute malware."
https://twitpic.com/4vf6z8
Hi Rand, great post that will lead to some very interesting tests and applications of the tag.
It would be interesting to see some analytics data for randz.net and the sort of changes in search traffic you've seen since the introduction of the rel=canonical tag. How drastically has the traffic changed?
Thanks
Search traffic to Randz.net is basically down to 0, while RandFishkin.com is getting almost all of those referrals (at least, when it has content for the queries).
Ok, In my Opinion this 301 and Canonical thing should be based on scenarios when it comes to cross domain and in some cases I think people should adopt using re=canonical
Scenario 1: remember the cases discussed on SEOmoz (Youmoz) where the original post is ranking lower than the republished content? In that case I think websites like business Insider and others who republish the content of partnered website and link to the original source should also use the rel canonical in order to let search engine know that where the original copy is!
Because in that particular case Google set business insider as original source (may be because it is more valuable in the eye of Google)
Scenario 2: The example what sir Rand described that his old blog have somewhat similar kind of content and he used rel canonical to the new domain, IMO In this case it is better practice to use 301 (that’s the way I think!)
rel canonical is there to let Google know where the original source of the content is… to derive the traffic to the new domain and get the SERP advantage 301 is the better option.
to conclude may be rel canonical is faster than 301 but we should use 301 and rel canonical where it suits!
I'm really scratching my head here because about a month before this post I ran a test on this with TOTALLY opposite results for a client of mine. The client has three brands with different sites and virtually the same products, this happened via a merger. All 3 sites are very authorititative and have massive organic rankings and traffic. They wanted to consolidate into one brand and I was consulting on this move. I commissioned test cases to see what the various results might be considering the fact that there is a huge amount of traffic at stake here.I setup 3 tests. 301 redirect. Canonical tag. 301 URL, replace URL on original site with a similar (placeholder) URL and add the canonical to that new URL for good measure (the reason for this was to maintain the structure and usability of the original site so that it could be used for paid advertising).The 301 worked the quickest. And the cross domain canonical failed miserably. In fact Google just went and indexed and ranked another similar page on the original domain rather than ranking any page on the site I was working to reference! All pages redirected or being used for canonical reference were very similar to each other so the changes in rankings and authority should have worked out. Overall, each page that I used any form of the canonical tag on totally failed and we monitored things for over 5 weeks and still no results. I'm wondering if anyone else has seen the cross domain canonical fail to work as miserably as I did in my test?
Great post. So how would I go about actually doing the rel canonical? If I have www.mysite.com and www.othersite.com wants to create new pages and use my content, what will the rel canonical tag look like on www.othersite.com? Do I need to also put this tag on www.mysite.com?
Awesome post rand! I have been searching for an effective way to redirect my old domain to my new one without losing my content. This just solved my problem, I'm going to have some fun with this one!
Awesome, Rand! I think this is my favourite WBF yet! I'm going to try doing what you did for randz.net on a couple of my sites.
Well done.
There is a project that I have been working on where I have been stuck on how to redirect all the old posts to the new blog. Putting a rel canonical in the WordPress theme just made my life a whole lot happier.
Hello fellow Webmaster
before i make any stupid decision its safe to ask people at seomoz first right? I have a website and i sell digital goods, i offer paid items and free items. i give free items for the purpose of attracting backlinks from other authority website so i could rank better in serps.
Now i don't want to create more free stuff, why? because this will only hurt my brand too much free stuff makes people not buy from me and instead the'll only wait for new free stuff which i release every week whereas the paid items i only release every month. plus since the number of free stuff increased the sales dropped so i need to come up with a new SEO strategy.
so i came up with a strategy, I need to build a new website for the purpose of frequent release of free items, Build Backlinks and using canonical tags to credit my main website giving all the credit, PR, link Juice etc.. to my main website. i have drawn a simple diagram so you could understand more about whats in my imagination, it took me 30 min to draw this hope you have some time to give me an answer.
See Link Below
Link To Image Diagram
You're comments and feedbacks are welcome. I thank you in advanced for having the time to read this.
Deleted comment. Please do not mind. Sorry and thanks.
Thanks for explaining this! Great help :-)
Sorry, a little slow here. Does cross domain rel = canonical actually cause the page to physically redirect to the canonical page or is this just a signal to Google telling them which domain the canonical page lives at. In other words, when the user clicks on the link in a SERP, will they go to the non-canonical page and NOT be redireted to the canonical page?
Like an internal canonical tag, it's just a signal to Google to credit the source in rankings. Users can still land normally on the non-canonical version.
Wow, that was fast Dr. Pete. Thanks!
Hi,
is there an update on this topic? any more experiments you guys have done?
Thanks
Interesting note: even with the rel=canonical tag linking to Robin's MasterNewMedia original blog post from SEOmoz's copy, the original still ranks below a scraper site and another site that links to the MasterNewMedia original post with the title as the anchor text."Entrepreneurship: The Full Story Of SEOmoz Told By Rand Fishkin" - original blog post ranks #3"Entrepreneurship: The Full Story Of SEOmoz" - not sure where/if original blog post ranks, stopped looking after several pages
In my opinion, the reason is very likely related to Google Panda (which Robin states has impacted his blog)... since Robin's blog posts have a big Adsense block on the top above the fold, which along with the navigation/template take up close to 90% of the space in a 1024x768 browser window.
Just goes to show that sometimes even rel=canonical tags are not enough to boost an original article's ranking if you have a Panda filter applied on your website.
There is a project that I have been working on where I have been stuck on how to redirect all the old posts to the new blog. Putting a rel canonical in the WordPress theme just made my life a whole lot happier.
Interesting update about Rel Canonical. I have a question here Can we use Rel canonical tag for subdomains as well? Just assume I have dev, staging and live most of the time all the three are accesible for Crawlers.
Thanks Randfish always very helpful. I wonder if we will migrate a domain with the same content, would use the very finest Cross Domain of all pages for the home or give each page equivalent, or do the 301? I am always very much in doubt when using canonical and always end up choosing the 301. But as I see that whiteboard is good to think about it right? Thanks. The SEO community in Brazil is happy for you to come here in July
João Vargas
Hey Rand! This is great stuff. I ran into a scenario where I had to implement the cross-domain rel canonical on a site in March 2010 and Google was super fast at picking up the attribution. In fact, it resulted in a 35% lift in listings for a site that was being cannabolized by a naughty dev site gone rouge.
However, back then Bing and Yahoo (this was pre-merger) both did not respond to the cross-domain rel canonical. I know most SEOs are primarily concerned with how Google reacts to major changes like this, but were you at all watching what happened to these sites in the Bing index?
Any insight as to the changes there would be appreciated.
Very useful atrtibute from SEO perspective, we used it for a site where we had three domains(one real and two CMS domains) pointing to same web hosting space and withing a week CMS domains were out of competition to our real domains.
I'm curious if the cross domain canonical can be used to prevent, on some level, duplicate content created by trade partners or "accidental" site scrapers. For example, a shoe company could publish an original product page description, which is then copied verbatim by resellers that are not search savvy. I know that if they steal the code, then the potential exists for them to not remove the canonical tag, but could it have any impact on preventing duplicate content if a reseller copies just the content?
They usuallys scrape the body and maybe the title tag. I don't usually see scrapers taking the entire html header area, where that tag would be located. It would be worth a try though if you're having trouble with scraper sites stealing your content, header and all.
What I want to know is: What's In It for the Re-Publisher? And how can the original source make it easy for a syndicating site to customize their header area to add this tag? I think an easier way of implementing this would be if the search engines said you could put it in the body. Of course, there may be some additional security issues to worry about at that point.
This is a wonderful post, but I wonder what the potential for abuse is here. Say, using throwaway URLs, all canonicalized to the site you're trying to get ranked, stuff like that. Does it only really work if the site giving credit is already better ranked?
Also wondering about using rel=canonical as a prelude to a 301, to better bridge the transition.
Regarding your lst example, did you rel can Randz.net to the index of RandFishkin.com ? (like a bulk redirect)
Even if the content is some how similar is not exactly the same. Will this mean that if I rel can a computer shop website to a cinema website for example will the cinema website get the benefit of the computer shop juice ?
Is this a two ways street ? (hijack threat)
Would this be an appropriate place to ask another site to use the crossdomain rel=canonical?
e.g. We own 5 restaurants in New Jersey, each of which has its own page on our website. Unfortunately, most of our SERP competitors are vertical search engines. To maintain visibility of the brand, we then have to pay the vertical search engines to display the content about our restuarants, that was already on our website, as part of their directories of 'restuarants in new jersey' etc.
Is that an example of where it would be appropriate to ask for a cross domain rel=canonical tag to applied (notwithstanding the likelihood of the vertical SE site declining our request)?
Little bit confusion 301 is for redirect and rel tag is for content rights than how we will differentiate both of these!!!
Dr. Pete actually did a great post comparing rel canonical and 301s here.
Edit (by Dr. Pete): As much as I love taking credit for things, that one was actually Paddy at Distilled. It's a good read, though. Also, I'll miss you all when Rand kicks me out for editing his comment ;)
Going by this, it could be construed that you're abusing the use of the canonical with your randz.net > new blog. It's not the same content, you said in the video it's all old stuff on randz.net. So you're using the cross domain canonical but not pointing to a copy of the same (or very similar) content. That doesn't sound terribly white hat to me.
You've given a great example as to how it should be used with the History of SEOmoz post, but then gone against what is recommended in other posts on SEOmoz as to how canonicals should be used.
That seems like rather inconsistent advice.
You raise an interesting point, which I believe can be seen two ways. One way being how you stated, that there is no matching content on the page targeted by the canonical tag.
The second way has to deal with the "essence" of what rel canonical is about. The tag is supposed to give credit to the originator of the content on any given URL. So what if someone wrote a piece of content for another company's website/userbase, and said company wanted to give that person the credit for doing so? If said person does not want to create the same page of content on their own website, than the company could just use the canonical tag to say, "Hey, we have this awesome article on our website, but we really aren't responsible for its creation. That credit should go to this guy at ...".
I'd argue in that situation the canonical tag is certainly an appropriate and white hat tactic to use.
Food for the brain...
If that was the case, and the content was to remain on your site but the credit was due elsewhere surely you would just link to the guy who wrote the content?
The canonical is clearly about managing duplicate content. It's about attributing ownership (be it on a page-by-page basis or a domain-by-domain basis). If the content only exists in one place, why are canonicals even a serious option? Either 301 your pages to where you want the user to go, or in your scenario, chuck the guy a link.
To me the advice Rand gives in this video goes completely against what rel=canonical is intended for, and goes against advice given in the post Rand links to on this site above.
@bludge, I think you may have missed the point of the test. this whole topic is centered on the "speed" at which Google is picking up rel=canonical tags compared to 301's. Rand's test was a valid test. I assume he'll eventually 301 the old blog to his new one. But if he had done that, he would have missed out on the opportunity to test what someone else was claiming, that rel=canonical works faster than 301. Going by this example it could make sense to implement a rel=canonical FIRST, then once that change has taken effect go back and do your 301. This could speed up the end result you intend.
I've been noticing that rel=canonical can take effect in a matter of hours in some of my tests. Great WBF!
@Bludge, It's a test first of all and second it might not be exact content but it is related, it's just the same as 301 redirecting a 404 page to your homepage to keep the link love.
Cheers Rand great WBF keep more like this coming!
Thanks, Rand now i got the point there is specific situations where we have to think about which one is useful for us.
Rand, similar to your old & new blog, am I right in thinking that this can be applied to businesses that have re-branded or changed their name?
Rather than redirecting their old URL, when my parents re-branded their company, they kept the old site live with a page explaining the re-brand (as not to confuse people who knew the old name but not the new one) and with a link to the new site. Can that page on the old site be cross-domain canonicalised as well?
Is it worth it (especially as it's already linking)? Are there any risks?
This could be awesome for solving problems for people who make uncareful moves from 3rd party websites - say you have a popular blog moved away from Blogger to your own URL. This could fix that potentially lost linkjuice with only page level changes in the Blogger account, to drop in the rel canonical links.
Great Whiteboard Friday Rand! My question is more towards canonicalizing pages on one domains, but maybe apply to cross-domains too...
If a full site migration were to take place, and a change URL structure were to happen, would keeping the old pages up, but adding a rel=canonical to the new page with modified URL structure do the trick to preserve link juice? I'm thinking not...
From what I know, the old URL will be deindexed, and the new one indexed once the page goes live. But I also understand that a 301 tells search engines that a page has been moved (with the same content), thus people who have linked to the old domain, the link juice attached to it is still going to be with the old page URL (so we have to make the old page is still up or redirected). How would rel=canoncial vs. a 301 actually benefit if canonicalizing the page doesn't pass link juice?
Thoughts?
Thanks,
Jackson
My understanding is that implementing rel=canonical on the page WILL pass link juice to canonical page in the same way 301 does, the difference is that 301 is a directive in your control whereas rel=canonical is a "strong hint" which may or may not be honoured by search engines.
Hence the canonical should be "fall back" for when you cannot execute 301 for whatever reason.
I'm working with site that has a lot of video content (flv & mp4). Some of the files are local, but some of them are from partner's projects.
Is there any link checker for such kind of stuff, for remote media files? It's not enough to check that link to the file return 200.
Very often partners are removing files but instead of 404 page their site still sends 200 code and some kind of sorry page. But it kills flash player on my page of course, and as result I have page with dead content.
So it's important to know that for example ******/movie.mp4 is real movie file, but not just live sites page.
Thanks for any help. www.seoresearch.info
Realy i like this article. i have lot of intrested to working on canonical url like 301 and 302 and 404 options.
I'm working with site that has a lot of video content (flv & mp4). Some of the files are local, but some of them are from partner's projects.
Is there any link checker for such kind of stuff, for remote media files? It's not enough to check that link to the file return 200.
Very often partners are removing files but instead of 404 page their site still sends 200 code and some kind of sorry page. But it kills flash player on my page of course, and as result I have page with dead content.
So it's important to know that for example ******/movie.mp4 is real movie file, but not just live sites page.
Thanks for any help.
Good to know that cross domain usage of rel canonical is so powerful - kind of astonishing that it adds a boost in the rankings, too. Does anybody else experienced that?
We have some old but not bad skeletons in the closet - let's see if we can use them the same way.
Thanks Rand for this example. I recently met the same consequences for canonical rule between 2 different domains.
Main page has progressed its position on search results and the duplicated content is gone on page 3 (before that it was on the 10th position on first page).
Canonical is a powerful way, easy to implement :)
I can see how this would be great for liciencing content, but not really sure how great it will be for a true 301.
I have an old Blogspot blog that I don't use anymore - I'm kind of wondering if it will produce the results that Rand menioned if I put that code in for my new blog. Going to try that today.
Thanks Rand!
Done I took two Blogspot blogs and added the rel canonical to my personal webpage. They didn't really have any significant amount of SEO value, but every little bit counts!
This is great. Very helpful for our cms sites. I can see it being abused though.