Same content, different domains? There's a tag for that. Using rel=canonical to tell Google that similar or identical content exists on multiple domains has a number of clever applications. You can cross-post content across several domains that you own, you can benefit from others republishing your own content, rent or purchase content on other sites, and safely use third-party distribution networks like Medium to spread the word. Rand covers all the canonical bases in this not-to-be-missed edition of Whiteboard Friday.
Click on the whiteboard image above to open a high-resolution version in a new tab!
Video Transcription
Howdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're going to chat about the cross-domain rel=canonical tag. So we've talked about rel=canonical a little bit and how it can be used to take care of duplicate content issues, point Google to the right pages from potentially other pages that share similar or exactly the same content. But cross-domain rel=canonical is a unique and uniquely powerful tool that is designed to basically say, "You know what, Google? There is the same content on multiple different domains."
So in this simplistic example, MyFriendSite.com/green-turtles contains this content that I said, "Sure, it's totally fine for you, my friend, to republish, but I know I don't want SEO issues. I know I don't want duplicate content. I know I don't want a problem where my friend's site ends up outranking me, because maybe they have better links or other ranking signals, and I know that I would like any ranking credit, any link or authority signals that they accrue to actually come to my website.
There's a way that you can do this. Google introduced it back in 2009. It is the cross-domain rel=canonical. So essentially, in the header tag of the page, I can add this link, rel=canonical href — it's a link tag, so there's an href — to the place where I want the link or the canonical, in this case, to point to and then close the tag. Google will transfer over, this is an estimate, but roughly in the SEO world, we think it's pretty similar to what you get in a 301 redirect. So something above 90% of the link authority and ranking signals will transfer from FriendSite.com to MySite.com.
So my green turtles page is going to be the one that Google will be more likely to rank. As this one accrues any links or other ranking signals, that authority, those links should transfer over to my page. That's an ideal situation for a bunch of different things. I'll talk about those in a sec.
Multiple domains and pages can point to any URL
Multiple domains and pages are totally cool to point to any URL. I can do this for FriendSite.com. I can also do this for TurtleDudes.com and LeatherbackFriends.net and SeaTees.com and NatureIsLit.com. All of them can contain this cross-domain rel=canonical pointing back to the site or the page that I want it to go to. This is a great way to potentially license content out there, give people republishing permissions without losing any of the SEO value.
A few things need to match:
I. The page content really does need to match
That includes things like text, images, if you've embedded videos, whatever you've got on there.
II. The headline
Ideally, should match. It's a little less crucial than the page content, but probably you want that headline to match.
III. Links (in content)
Those should also match. This is a good way to make sure. You check one, two, three. This is a good way to make sure that Google will count that rel=canonical correctly.
Things that don't need to match:
I. The URL
No, it's fine if the URLs are different. In this case, I've got NatureIsLit.com/turtles/p?id=679. That's okay. It doesn't need to be green-turtles. I can have a different URL structure on my site than they've got on theirs. Google is just fine with that.
II. The title of the piece
Many times the cross-domain rel=canonical is used with different page titles. So if, for example, CTs.com wants to publish the piece with a different title, that's okay. I still generally recommend that the headlines stay the same, but okay to have different titles.
III. The navigation
IV. Site branding
So all the things around the content. If I've got my page here and I have like nav elements over here, nav elements down here, maybe a footer down here, a nice little logo up in the top left, that's fine if those are totally different from the ones that are on these other pages cross-domain canonically. That stuff does not need to match. We're really talking about the content inside the page that Google looks for.
Ways to use this protocol
Some great ways to use the cross-domain rel=canonical.
1. If you run multiple domains and want to cross-post content, choose which one should get the SEO benefits and rankings.
If you run multiple domains, for whatever reason, let's say you've got a set of domains and you would like the benefit of being able to publish a single piece of content, for whatever reason, across multiples of these domains that you own, but you know you don't want to deal with a duplicate content issue and you know you'd prefer for one of these domains to be the one receiving the ranking signals, cross-domain rel=canonical is your friend. You can tell Google that Site A and Site C should not get credit for this content, but Site B should get all the credit.
The issue here is don't try and do this across multiple domains. So don't say, "Oh, Site A, why don't you rel=canonical to B, and Site C, why don't you rel=canonical to D, and I'll try and get two things ranked in the top." Don't do that. Make sure all of them point to one. That is the best way to make sure that Google respects the cross-domain rel=canonical properly.
2. If a publication wants to re-post your content on their domain, ask for it instead of (or in addition to) a link back.
Second, let's say a publication reaches out to you. They're like, "Wow. Hey, we really like this piece." My wife, Geraldine, wrote a piece about Mario Batali's sexual harassment apology letter and the cinnamon rolls recipe that he strangely included in this apology. She baked those and then wrote about it. It went quite viral, got a lot of shares from a ton of powerful and well-networked people and then a bunch of publications. The Guardian reached out. An Australian newspaper reached out, and they said, "Hey, we would like to republish your piece." Geraldine talked to her agent, and they set up a price or whatever.
One of the ways that you can do this and benefit from it, not just from getting a link from The Guardian or some other newspaper, but is to say, "Hey, I will be happy to be included here. You don't even have to give me, necessarily, if you don't want to, author credit or link credit, but I do want that sweet, sweet rel=canonical." This is a great way to maximize the SEO benefit of being posted on someone else's site, because you're not just receiving a single link. You're receiving credit from all the links that that piece might generate.
Oops, I did that backwards. You want it to come from their site to your site. This is how you know Whiteboard Friday is done in one take.
3. Purchase/rent content from other sites without forcing them to remove the content from their domain.
Next, let's say I am in the opposite situation. I'm the publisher. I see a piece of content that I love and I want to get that piece. So I might say, "Wow, that piece of content is terrific. It didn't do as well as I thought it would do. I bet if we put it on our site and broadcast it with our audience, it would do incredibly well. Let's reach out to the author of the piece and see if we can purchase or rent for a time period, say two years, for the next two years we want to put the cross-domain rel=canonical on your site and point it back to us and we want to host that content. After two years, you can have it back. You can own it again."
Without forcing them to remove the content from their site, so saying you, publisher, you author can keep it on your site. We don't mind. We'd just like this tag applied, and we'd like to able to have republishing permissions on our website. Now you can get the SEO benefits of that piece of content, and they can, in exchange, get some money. So your site sending them some dollars, their site sending you the rel=canonical and the ranking authority and the link equity and all those beautiful things.
4. Use Medium as a content distribution network without the drawback of duplicate content.
Number four, Medium. Medium is a great place to publish content. It has a wide network, people who really care about consuming content. Medium is a great distribution network with one challenge. If you post on Medium, people worry that they can't post the same thing on their own site because you'll be competing with Medium.com. It's a very powerful domain. It tends to rank really well. So duplicate content is an issue, and potentially losing the rankings and the traffic that you would get from search and losing that to Medium is no fun.
But Medium has a beautiful thing. The cross-domain rel=canonical is built in to their import tool. So if you go to Medium.com/p/import and you are logged in to your Medium account, you can enter in their URL field the content that you've published on your own site. Medium will republish it on your account, and they will include the cross-domain rel=canonical back to you. Now, you can start thinking of Medium as essentially a distribution network without the penalties or problems of duplicate content issues. Really, really awesome tool. Really awesome that Medium is offering this. I hope it sticks around.
All right, everyone. I think you're going to have some excellent additional ideas for the cross-domain rel=canonical and how you have used it. We would love you to share those in the comments below, and we'll see you again next week for another edition of Whiteboard Friday. Take care.
Hi Rand,
Let's say you wrote something on myblog.com/page and the Guardian wanted to republish it.
Do you think that by using rel="canonical" on the Guardian to point to the original on myblog.com/page that myblog.com/page would inherit the power of the Guardian's domain passing through the canonical link?
It sounds like you are saying yes. And that if guardian.com/myblog earned links, as long as that piece canonicalised to myblog.com/page, those links would pass 90% (or so) equity to myblog.com/page and guardian.com/myblog would get no SEO value from those earned links.
I've never thought of it this way but I would have thought Google would not handle it that way as the intention of a canonical link is obviously not the same as citing a webpage as part of your own, original content.
And if this is the case, why are more people not doing 'canonical link building'. I would imagine it could be gamed as many webmasters without much SEO experience might just think "oh that's nice, they are letting me repost their article and all I have to do is point to the original" - which is why I would have thought Google handles it differently.
But, I'm really not sure of that. Every day's a school day. :)
Interesting stuff!
Cheers,
Gill.
I just want to add a very small point here, and it's that large publishers simply DO NOT use the canonical tag properly. (Almost) all publishers point the canonical to their own pages, no matter who's the source.
It's one of the largest problems I'm personally seeing with our news when these are syndicated to 3rd party sites. I really wish things would be as simple as in your example.
But regarding your question, if I may, in theory the answer is definitely yes, that's just how Google treats a canonical.
Not sure we can say "definitely yes, that's just how Google treats a canonical" on that.
The jury's out over here!
Gillon, I agree with you on the suspicion here. I do not think that the canonical on another site passes SEO benefits back to you. I think it just takes them out of ranking ahead of you for your own content, and that's all.
My thoughts exactly!
To explain my doubt further...
If I were Google, I would build into the algorithm a rule which ignores inbound links on the duplicate page that is canonicalised to the original and only use canonical tags to stop the duplicate ranking ahead of the original as you said.
People will link to The Guardian just because it's The Guardian. People won't likely link to myblog.com in the same quantities as myblog.com just won't get the same exposure (even if myblog.com/page is the one indexed in the SERPs).
So to give myblog.com the link equity from The Guardian is not logical in my mind. It would also mean that entities like The Guardian are unknowingly casting "votes" to the entities they are canonicalising to.
Just because you want to republish something doesn't always mean you trust or want to vouch for that entity (imagine The Guardian republished a hate speech article to highlight how awful it was for example). Adding a "nofollow" to a canonical link defeats the purpose of the canonical so that means ALL canonical links are dofollow which doesn't give webmasters a choice and means it's possible to game the system.
Also need to remember Google says "Note that although we recommend that you specify a canonical page for duplicate URLs or versions, Google might algorithmically choose a different page as canonical for various reasons such as performance or content."
I was using some rel=canonical articles but took them down. I was afraid that Google would devalue or penalize the practice as "another form of linkspam". In the past, google has recommended some practices and even rewarded them... then they changed their mind without telling anybody - after tons of people started going overboard using them.
I think that if a person decides to try this, it is best done where you have some control or trusted relationship. That way you can yank them down fast when Google gets tired of it.
If I may, in nutshell, we are saying duplication is okay as long as we say where did we took it?
Duplication is okay as long:
But I'm not talking about whether duplication is okay or not, I'm talking about how Google handles canonical links from a link authority perspective (not a duplicate content perspective).
We know for sure how it works from a duplicate content perspective, but I personally don't think we (none Googlers) know for sure about the authority perspective. There are lots of estimated opinions on it, but no hard facts :)
Thanks for sharing the tips, Rand.
I would also like to add here that if someone is making use of mobile website version of their desktop version, they need to implement a canonical tag on their mobile website page with an URL of the desktop version.
For example,
Your main domain: iamexample.com
Your mobile version: m.iamexample.com
Then, have this tag in the <head> section of your main domain -
<link rel="alternate" media="only screen and (max-width: 640px)" href="https://m.iamexample.com">
And, have this tag in the <head> section of your mobile version page -
<link rel="canonical" href="https://iamexample.com">
Thanks
Thanks for the helpful tip praveen!
I didn't know about that! Thanks you for the good tip!
Useful tip. A similar case exists if you have an AMP version of your website.
Add the following to the non-AMP page:
<link rel="amphtml" href="https://www.example.com/url/to/amp/document.html">
And the following to the AMP Page:
<link rel="canonical" href="https://www.example.com/url/to/full/document.html">
Thanks Praveen for the shared this helpful tips with us.. Its really helpful for me and others..
Didn't know about the Medium.com cross domain canonical. Just uploaded one of my good articles on there. Thanks for the tip, Rand!
That Medium tip is (insert fire emoji)!!!!
A great video there, Rand. I've only had to do cross-domain canonical once before, but now I feel like i understand it better regarding the matching bits. Thanks also for that Medium tip too - i'd been concerned about the duplicate content issue there for a while, so it's good to know they've addressed it.
Hi!
I have a question. On my educational domain I have tons of content, some of it is very similar.
Let's assume, that I have 5 subpages on same topic (content differs slightly). Would it help if I put cannonical on 4 of them and link to the best one?
I will be very contented with your help :)
Love the tip for using Medium as a content distribution network! We will definitely be utilizing Medium.com/p/import for our content distribution moving forward. Thank you so much for all the great insight on Rel Canonical Rand!
hmmm. I wonder if there is an IFTTT solution for new blog post > medium.com (with rel canonical)
Thanks Rand for sharing this amazing information. Canonical URL we use for disallow duplicate pages from indexing and duplicate content appearing on multiple URLs. then What is the Difference between 301 Redirect and Canonical URL.
A 301 would mean 'example.com/page-one' would redirect to 'example.com/page-two' or 'anotherexample.com/page' (the same domain or a different domain) meaning that 'example.com/page-one' would essentially no longer exist to users or search engines crawling your site and the majority of the link equity pointing to 'example.com/page-one' would be passed to 'example.com/page-two' or 'anotherexample.com/page'.
A canonical doesn't redirect the page. It means both 'example.com/page-one' and 'example.com/page-two' or 'anotherexample.com/page' can exist but the target page *should* be the one which ranks in the SEPRs (not both will rank).
Rand is also saying that the page you are canonicalising to also would benefit from the link equity of 'example.com/page-one' in the same way it does with a 301 redirect. I'm not yet sold on that part though. :)
Also bear in mind a canonical is a hint, NOT a directive, meaning sometimes it may not be honoured (but that should be rare).
Google says"Note that although we recommend that you specify a canonical page for duplicate URLs or versions, Google might algorithmically choose a different page as canonical for various reasons such as performance or content."
Very useful, I have been studying Content SEO and this a really key factor mainly in the distribution process of it. So this article was very useful. Thanks Rand
Great post. I wonder, how closely does that headline need to match the original for the authority to pass along? As an original publisher, I might not have control over that.
Very informational post Rand, Thank you!
I was hoping to get some feedback from you (and your readers) about the application of this technique to a slightly different use-case than discussed above.
We have an eCommerce site on example.com and we also have a blog associated to "example" running on a different subdomain (example.something.somethingelse.com). The idea was to configure the blog as a subdirectory (example.com/blog) but due to a technical challenge we were not able to achieve this. So, for now we add an explicit backlink every time we post content on the blog with the hope that the search engines will see them as one entity. We're not sure how successful we are with the above though.
Having read this post I was wondering if rel-canonical tag can be used to attribute the link equity of the blog to the main site. We also repost the content on Medium &Tumblr so of if this can work, we would do the same for other platforms as well.
Appreciate your views on this. Thanks again.
If you're not able to create a subfolder blog on example.com/blog, there will be no page(s) to add canonical links from. And if all the content is created on subdirectory.example.com and example.com/blog and example.com/blog/articles never existed, there is nothing (in terms of link equity) to pass from there to the subdirectory.
The canonical tags are for when two (or more) pages, which are the same, exist. It sounds like in your case there would be no duplication of content, and if there were, you'd be far better off doing it all on the subfolder, not the subdomain.
My advice would be to try and get it working on example.com/blog. The benefits would, in theory, be far greater.
Good luck!
Thank you Gillon! I understand what you are saying.
Hi Rand, great post. As always.
I have one question though. How do you handle duplicate content that is not fully identical but has duplicate phrases? For example: web page A contains the same paragraph as webpage B. This is not fully duplicate content, but it is of course partially duplicate content. How does this affect your rankings in Google and are there ways to solve this problem?
Like to hear from you!
In principle nothing happens, what google penalizes would be in case of having both identical pages as it can be the paginated of an ecommerce, that each page is the same, here if a canonical would have to be applied.. But if the text has some parts or phrases nothing happens, unless there is a totally equal content.
I hope I've helped
Regards
Thanks Rand, you always comes up with something interesting and thoughtful things. i didn't know that the cross domain usage of canonical, it will really help full for my last project which is for 5 different country but one dataSource. it was travel website and each domain containing thousand of pages generated by programmatically...
By the mean of content does really need to match...what if 90% content is match and few things are changing by variable like main title for Dubai Page, "Cheap Flights to Dubai" and for Los Angeles "Cheap Flights to Los Angeles" will it be fine..and same things for URLs. small words changing...our goal is trying to index as many pages possible but beside we still worried about duplications.
But this article give me more suitable and clean solution.
Hi Rand,
Thanks for sharing once again an interesting article with us. I have little bit confusion...
As you said above if "friendssite.com/green-turtles" re-publish your content ("mysite.com/green-turtles") then friendssite.com should have meta tag <link rel="canonical" href="mysite.com/green-turtles">
Am I right? but
I have checked that Medium's source code they are using <link rel="canonical" href="medium.com/@username/green-turtles"> instead of <link rel="canonical" href="mysite.com/green-turtles">
It's confusing...
Please explain...
Hey Rand, I don't suppose there is any way to do this in LinkedIn Publisher?
Hi Rand,
Great whiteboard Friday as usual. Quick question on the inclusion of the same images on the 'duplicated' page. Do these also need to have the same file name / alt text / etc as the originating piece? Or can these simply be the same image?
Thanks in advance.
Anything change with rel canonical for press releases?
Excellent post! We are starting to use the rel=canonical more and more in order to expand our content audience. I was wondering, in terms of the things that do need to match and the things that don't need to match, I was wondering how you managed to find that out - please let me know if there is some research or a study I could look at.
Thanks!
Interesting Rand Reflection.
I have a question. if an article is written on the same subject matter, but the content is different, for example by incorporating new ratings. Are we dealing with duplicate content? In these cases this operation interprets that I can not be done because it is different content.
Thank you very much
Hi Rand. That is a great article! I always love your white board Friday, sweet & short and great tips. :)
I have a question regarding the cross domain canonical. I have a client who is publishing the same content on multiple website which serves different geographies. In this case the client want their respective website to rank for the respective geography, in this case what needs to be done?
Because, if canonical is added, Google will consider the main content and start ranking that content.
Thanks Rand!
It's great to brush up on some canonical 101.
I especially liked the Medium part - didn't know they had an option to add a cross domain canonical.
Hi Rand,
Thanks for sharing this. Now consider this situation:
I am a publisher and I have 2 different domains for 2 different countries. Now I publish a content @domainA and also @domainB. Now I was domain A to rank for US audience and domain B to rank for India audience. But if I put "rel=canonical" I guess only one of them will rank. What to do in such tricky situations?
Thanks,
Dipankar Biswas
Hi Rand,
Can i post other blog posts (e.g. moz blog post) with rel=canonical to my website?
Thanks
Hello Rand,
I have one question related canonical, If I have one main site (a . com) and also one web 2.0 site (a. wordpress . com). I copied multiple pages (a. com/1, a. com/2, a. com/3 etc..) content of the main site and put in web 2.0 one page (a. wordpress. com/all), then should I add multiple pages URLs with canonical tags of my main site pages in web 2.0 site pages?
How I canonicalization for these?
We have a single domain which is country specific folders. Many of course are in English. The product is for the local market. So it's specific to that market. How do you use rel=canonical in this instance? Or don't you....?
In this case, you would use hreflang tags rather than canonical tags.
https://support.google.com/webmasters/answer/18907...
This is where I get confused. Lets say I've 10 regions, all in folders (ie USA/UK/AU/SG/etc.). All of the 10 are English. The content is the same in all of them as it's English.
For each region their version is the definitive version.
Is this seen as duplicate content or because the language variable would be different, ie En-GB, En-AU, En-SG, etc. different....?
Moz explains this well here: https://moz.com/learn/seo/hreflang-tag
What is Rel="Alternate" Hreflang="x"?
Introduced by Google in December 2011, the hreflang attribute allows you to show search engines what the relationship is between web pages in alternate languages. It's useful when you've created content that's specific to a local audience. The hreflang attribute adds a signal to search engines that a user querying in language "x" will want this result instead of a page with similar content in language "y".
For example, if you create a Spanish-language version of your English-language homepage, you would tag it as "Español" by using
hreflang="es"so that searchers with an IP address that a search engine has reason to believe is in a Spanish-speaking country are served that page in Spanish instead of the English version. This can decrease your bounce rate and increase your conversions by making sure your target audience lands on the version of your page most appropriate for them.Hreflang can also be used to show that you have content targeted toward variants of a single language. If that's the case, you can target your pages even more specifically by extending the hreflang attribute with annotations that indicate which region the content is localized for, e.g. Spain
hreflang="es-es"versus Mexicohreflang="es-mx". This is particularly useful to geotarget users to control for variations in currency, shipping, seasonality, and culture.Hello Rand,
You have cleared the whole rel canonical tag very well. I gave proper attention to everything that you said, I do the same for all white board Friday videos. But at the end of the video you talked about using medium rel canonical tag. Here is my question for you
Why would google index a content which is already available on web. Instead google would like to index, crawl and keep the rel canonical url in their database.
What you would say?
Nice tips Rand!! I´ve never used this cross domain canonical tags before, but looks like a good tool to have in the SEO Pocket knife for future actions or projects ;-)
Thanks so much for sharing your knowledge!
Hello Rand
We are really thankful of you for this post as it is going to help us a lot. Look forward to read your other blogs as well. Thanks
Another great WBF Rand, thanks!Really interesting and informative article and it is useful information. I will share this article to my social media account. Thanks for sharings
As always from rand, very complete and interesting information. Thank you very much for sharing your knowledge Rand.
Best regards
Thank you very much for the information, it is really interesting and very very useful !!
The issue of duplicated content is something really complicated, you have to go with much thought, not only on your part ... that is, not only to take care of your texts and publications, but also where you have to publish!
Another great WBF Rand, thanks!
Rand,
I was using Medium this way for a long time, but in the last weeks something wrong started to happen.
Articles with properly added rel=canonical started to rank higher than my own site. The same article, the same content, etc - and searching for main phrase you've got Medium higher than original content.
Has anything changed last weeks with it?
Hi Rand,
it is a great presentation like always. But you describe a perfect situation in a perfect world, where everyone modify their html code on websites to give us a canonical to our content. In my company it is impossible. We publish a lot of useful content which is in a Press room of our website and we have a problem to "force" people even only to put backlink to us from content which was incurred from us.
What to do in that case? Do you have useful advices how to seduce people to put canonical to our content?
Hi Rand,
Its a really useful Whiteboard Friday in clarifying doubts regarding this area. I have a question regarding publishing translated content which is theoretically duplicate content, except that Site A has it in English while Site B has it in Spanish (or any other language). In this situation would you recommend using the Cross domain canonical or else would you suggest using the href-lang tags on each piece of content? Or the combination of both?
Appreciate your input on the above
Thanks
Hye Rand,
Your concept to provide the information is very clear and impressive. Before this article I am not was completely clear about canonical tag that we can earn money via this concept of selling content and also it is more effective in case of SEO. But I have one question in mind that can we sell same content to two other different websites using this method? I think not possible. If you know about it then please tell me about it?
Thanks for it.
Great WBF Rand,
Yes! Using the Cross Domain Rel=Canonical is always helps gaining link authority, ranking factors and more.
I have a condition here? One of my client wants to gain more value from few of his article which were earlier published in "LinkedIn" now can he just get that content import in Medium? For me this didn't work! Can you help with some other way to re-publish this content on Medium.
I suggested him:
Anything more you would like to suggest?
Cheers
Great tips. Especially with LinkedIn and Medium - 2 platforms I use frequently.
I had no idea Medium automatically put in the cross domain tags. I hadn't seen much use for it until now :)
Brill - the Medium tip is perfect. I've been wondering about just such issues. Kudos!
Always a great source of food for thought, especially in regards to Medium. I am now rethinkign the strategy I have been using with that particular platform.
Intuition Softech, known for custom web application developmentin Australia will take care of the mentioned points while doing ecommerce development.!
Thanks A lot Rand :)
Hy ,Rand Fishkin
Really interesting and informative article and it is useful information. I will share this article to my social media account. Thanks for sharing
Hi Rand,
I always follow your blogs and tips over here and LinkedIn. I always get something inredible from you and yes this time also a great article about rel=canonical tag which will help everyone a lot.
Thanks
Thanks for the helpful post