After I wrote about my intentional experiment in catastrophic canonicalization last year, I started getting a lot of questions about other uses (and abuses) of the canonical tag. In many cases, I couldn't find much data out in the blogosphere, so I decided to put a few of these questions to the test in a series of mini-experiments. Most of these applications are a bit extreme, and you'd probably never try them on a real site, but I think they all help to test the boundaries of the canonical tag and how Google processes it.
(1) Cross-domain Syndication
Rand recently wrote up his experience with a cross-domain use of the canonical tag, and I had an opportunity to try it on 2 of my own sites. The purpose was legitimate – I wrote a post about celebrating 5 years in business, and it made sense to cross-post on both my company (User Effect) and personal (30GO30) blogs. Since my personal blog is relatively new, and I felt the post was more personal than corporate, I wanted it to get credit for being the source of the article.
Of course, my company blog is quite a bit older and stronger on just about every dimension you can think of. I've listed a few metrics below (from the start of the test), for reference:
So, the obvious question was: could the cross-domain canonical tag override all of the other signals suggesting that my company blog was actually more authoritative?
The short answer is: "Yes". I published the post nearly simultaneously on both blogs on May 10th. The next day, Google started indexing the title of the post from the home-pages (the 2 home pages both appeared in SERPs). On May 12th, the full post was indexed and ranking only on 30GO30 (for the post title). Google seemed to have no issue with the cross-domain canonical from a stronger domain to a weaker domain.
(2) Canonical in <BODY> Tag
One common fear about the cross-domain use of the canonical tag is how it might be hijacked. Obviously, someone can hack your server, but what if you allow user-generated content and someone simply drops a canonical tag in the middle of the page?
To test this, I dropped a canonical tag right before the closing </BODY> tag. I referenced a page on the same domain, assuming Google would be more likely to process the internal canonical than a cross-domain (if this worked, I could move on to Phase 2). The misplaced tag seemed to have no effect – I made the change on May 9th and the page was re-cached on May 14th and May 18th with no impact on the SERPs.
After I launched this experiment, Matt Cutts posted about canonical corner cases and addressed this specific issue:
First off, here's a thought exercise: should Google trust rel=canonical if we see it in the body of the HTML? The answer is no, because some websites let people edit content or HTML on pages of the site. If Google trusted rel=canonical in the HTML body, we'd see far more attacks where people would drop a rel=canonical on part of a web page to try to hijack it.
Since I was already mid-experiment, I thought I'd let it ride, but it was nice to see the confirmation.
(3) Canonical in False <HEAD>
Just so I don't get accused of mindlessly sheeple-ing whatever Matt says (which I can almost count on as soon as I link to his blog), I tested a variation of (2). This time, I put the bad canonical tag in a second <HEAD> tag, at the very top of the <BODY>. In other words, my page looked something like this...
<HEAD>
<TITLE>Experiment 3 Page</TITLE>
</HEAD>
<BODY>
<HEAD>
<LINK REL="canonical" HREF="https://www.example.com/bad-page">
</HEAD>
</BODY> The change was made on May 18th and the page re-cached on May 20th, May 22nd, May 26th, and June 4th. It had no measurable impact, consistent with Matt's statements.
(4) Canonical to Fake URL
In parallel with (2), I tested an idea that came out of Q&A. What would happen if you pointed a canonical tag to a URL that doesn't exist? Obviously, you wouldn't generally do that on purpose, but if, for example, you made a major error in your CMS, how damaging would it be?
I introduced the canonical tag (on a different page than (2), of course) on May 9th. It re-cached on May 15th, May 17th, May 21st, and June 1st. It had no apparent impact.
It turns out Matt addressed this one in his post, too (thanks for ruining my research, Matt):
For example, if we think you're shooting yourself in the foot by accident (pointing a rel=canonical toward a non-existent/404 page), we'd reserve the right not to use the destination url you specify with rel=canonical.
The lesson here is pretty simple – the canonical version of the page has to actually exist. While that may seem obvious, I've had people ask about using the canonical tag as a sort of URL rewrite. On the surface, the idea has a certain logic, but in practice, it goes completely against the purpose of the canonical tag.
(5) Crossing the Streams
I asked the SEOmoz staff if they had any extreme canonical experiments to try, and Cyrus suggested pointing the canonical tags of 2 pages at each other. I should've listened to Egon when he said "don't cross the streams", but I'm not a very good listener.
So, on May 18th, I pointed the User Effect "About" page at the "Services" page, and vise-versa. Clearly, no one would ever make that mistake, but this was an exercise in exploring how Google would interpret the suggestion – a peek into the black box.
Re-caching took longer than expected, and at first the results looked pretty dull. On May 28th, the "Services" page apparently re-cached, and by June 3rd that page was showing for searches on "About User Effect". It seemed that either the stronger page had won, or the "Services" page had simply been re-crawled first.
Then, something very strange happened. The "About" page reappeared in searches on June 7th, but a query on "User Effect Services" resulted in this:
Both pages were now appearing in search results, but the "About" page had its title rewritten to match the service-related query. This was not the title of the actual "Services" page, but a complete invention by Google. Clearly, the mixed signals of the 2 canonical tags created a problem.
I think there's an important lesson here – if you send Google mixed signals, there are consequences. I see a subtler example of this all the time – people rel-canonical (or even 301-redirect) to one version of a URL, but then use another version in internal links, inbound links, social media, etc. If you say one URL is canonical but act in every way like another URL really is, Google may choose your actions over your words. Don't mix signals.
(6) Facebook/Twitter Buttons
This last one's not really an experiment, but something interesting I noticed about social media plug-ins while I was stirring up trouble. To make a long story short, my personal blog focuses on 30-day challenges. I'll often have a main post about a challenge and then a number of update posts to tell how it's coming along. Those updates aren't usually core content that I want to rank, so I decided recently to canonicalize the weekly updates in the challenges to the main challenge post.
A Few days later, I was revisiting one of my weekly update posts and was surprised to see this at the bottom:
I had barely mentioned this particular post on social media, and something was clearly out of whack. I realized quickly that these were the numbers from the canonical version of the page. The Facebook and Twitter scripts were actually honoring the canonical tag.
In the intervening couple of weeks, Facebook no longer seems to be reporting numbers from the canonical version, but the Tweet counts still match the post that the canonical tag points to. I'm not entirely sure what to make of this, but it's food for thought – canonicalization may be impacting more than just your SEO.
The Usual Disclaimers
I'm not a real doctor – I just play one on TV. Don't try any of this at home (or at work). Matt Cutts is not the source of all wisdom in the universe, nor is he the antithesis of all wisdom.
Obviously, a couple of these experiments were sillier than others, but I think they all give us some insights into just how seriously Google takes the canonical tag, and how seriously we should probably take it as SEOs. That means using canonicalization to actually point to the canonical versions as we honestly intend them. By playing around the edges of the black box, I'm not trying to crack the Google code, just better understand how we can use these tools effectively and responsibly.
Cool post! Silly expriments can yeild serious results. Next time, try canonicalising to google.com and see if the Internet disappears in a giant vortex? Or maybe it'll just manifest the Stay Puft Marshmallow Man. (Or, more terrifyingly, Clippy.)
1. Why this post is important: Because as use of the rel=canonical tag becomes more widespread, SEOs will have to do more and more troubleshooting to root out potential problems.
2. Dr. Pete - Thank you for performing the experiments us mere mortals or too afraid to try. You are a certified mad scientist.
My favorite "certifed mad scientist!"
This is one of the best summaries (& tests) I've seen to date on the canaonical tag.
yea eBays templates have templates in a iframe in a new page with the content and I tried to use something like this to outrank for simular keywords. It didn't work. :( muh, at least I tried. What kind of SEO would anyone be if they didn't test?
I would just like to say that I admire that you for testing these things out on your business website, and not some test domain with fake content. If you screw something up in these tests (like when you rel=canonical'ed the entire site to the homepage), the effect actually matters.
I'm not sure it's the brightest idea, but honestly, I don't think some of these tests are very meaningful on dummy sites with no authority that Google doesn't care about. What we really want to know is - what happens when you do this on a real site that ranks?
I agree, thanks Dr. Pete for making the "lab rat" a valid one whose results actually teach us something. Much appreciated! :)
LOVE IT, Pete! This is fantastic. The last example (about the title being replaced) is especially interesting to me. Crazy that even though you *tried* to trick them, they still pulled page information and replaced your title tag because it seemed to be more relevant! Maybe Google actually CAN sniff out relevancy more than we give them credit for.
Great post. You and I need to get together sometime and brainstorm some crazy things to test on some sites.
Yeah, love the title overwrite. Very interesting! I've seen some very odd titlte re-writes recently (which reminds me, I must write those up!!)
I must play with canonicalisation w/ the social plugins for just published pages see if there is a way to give a psycological boost to encourage more people to sharie it. One Question about timing though Pete - could you indicate a timeline of events? e.g. when did you canonicalise the post and when are the numbers updated in the twitter button?
Yeah, I find it interesting to see how Google copes with something that's clearly illogical. As Charlie said, I think this fits into some of the broader TITLE tag rewrites we've been seeing over the last year or two. It's a glimpse into how the algorithm thinks.
it seems to me that the title effect gave away the best title you can have on a page that ranks high in SERPs. but what bugs me is WHY they changed that title? Are there more cases of altered snippets?
Anyway good post. You should track experiments live on your blog
It's so nice to read a report on real live testing, and useful testing too! I see clients who have mangled their canonical link from time to time and it's good to have an idea about what kind of repurcussions might show up - or not. Nicely done.
To be honest, I sometimes get frustrated with the forum posts where 10 people argue about whether something will work but no one actually tries it. Granted, some tests are extremely difficult or even destructive, but "Will a canonical tag work in the <BODY>?" is a 5-minute change and 3-4 weeks of waiting.
Hi Peter! I sincerely believe that people like you experimenting and testing things are a true blessing for the SEO community. And if we add to this also you good amount of sane craziness, better.
The last observation, re: effects of on Social, is something that should have to be experimented apart. The first very side thought I have is the confirmation of how much not reliable can be Facebook, as it changes so fast and so "secretly", changing always the cards in the middle of the game.
Finally, just one thing is still not clear to me. Isn't the should to be used - as search engines say - in case of totally identical contents in two or more separate URLs? Because, reading I understood that was the case in some of your experiments, but not so in others, for instance nº 5 and, in the Rand experiment the two domains were not technically duplicate ones (as far as I remember it). May you please clarify this, Peter?
Ciao!!
Test #5 was definitely not advisable or in any way consistent with the intention of the canonical tag. It was an experiment in how Google would interpret a fundamentally illogical request (A->B + B->A). That was purely exploratory on my part.
Bonus thumbs up for the Ghostbusters reference!
Great experiments!
We also ran a few tests on our company's sites. We operate in the U.S. and UK and have a separate website for each country with the identical layout and branding as well as the url strucutre, but different text copy.
Our U.S. site is much stronger in terms of domain authority and inbound links, so if we publish a blog post on this site, it normally gets a good ranking straight away. What we decided to do however is to write a blog post about On-page SEO for Ecommerce and publish it on our UK site first, publisize it via social media sites to show Google that the UK site is the original source of the article, then publish the identical blog post on the U.S. site and put a canonical from the UK site to the U.S. site to take advantage of the U.S. site's stronger authority and get good ranking and finally after keeping the blog post on the U.S. site for 1-2 weeks, put rel canonical from the U.S. site back to the UK site. The whole idea was to promote our UK site. Here is what happened:
1. After the blog post was published on the U.K. site, we did not rank in top 10 Google for "SEO for ecommerce" or "On-page SEO for ecommerce".
2. When we republished the blog post on our U.S. site and put rel canonical from our U.K. site, we immediately appeared as #1 Google SERP for "On-page SEO for ecommerce."
3. In a week when we put rel canonical back to the UK site, we dissapeared from search for a couple of days completely and then the UK site started to rank #1 under "On-page SEO for ecommerce."
I am not sure what Google thought of us when we were switching rel canonicals back and forth, and I was really terrified when we dissapeared from SERPs for a couple of days. Here is the url if you want to look into it yourself - www.artdriver.co.uk/seo-for-ecommerce
Question: There is a WordPress blog that uses a third party SEO plugin that allows to insert canonicals. In the end, each blog post ends up having two rel canonicals in the <head> section: one that is default WP and another one that relates to the third party plugin. How does this affect our site in SERPs?
That's interesting. So, basically, you canonicalized to the US version first to get build up initial authority, and then reversed it? I think the reversal can throw Google and cause them to re-evaluate (short-term) - I've seen similar things happen. I wonder if it would be better to just post it to both sites, let them be duplicates, and then canonical to the UK site once the US version is indexed and ranking.
Of course, now you're tying in international SEO factors, so this gets even trickier than a regular cross-domain situation. Theoretically, Google should allow you to have the two versions in the two countries, but obviously we see many situations where duplicated English content gets weakened (even US/UK).
Hi dr. Pete,
interesting test.
I experienced a slightly different version of test number 5: fake URL that returns a 200 HTTP status code.
As you know, a lot of CMS returns the same page for different URLs (this is not so good for SEO)
For example this page:
asos.com/Men/Watches/Cat/pgecategory.aspx?cid=5034
You can obtain the same content, even if you use these URLs:
- asos.com/Mwerwerwetches/Cat/pgecategory.aspx?cid=5034
- asos.com/letsparty/Cat/pgecategory.aspx?cid=5034
- etc.
If you put rel=canonical to URLs that have no links, but the have a prettier form, Google will index it without any problems :-)
I don't like this solution, but I can say it works :-)
Yeah, if the target URL will resolve, that's going to be a different situation. I do think you have to be careful, though - if you choose a weaker version of the page (say, one that has no internal links) to be the "canonical" version, I suspect you're weakening your rankings in the long-term. In theory, your canonical version should be consistently applied, which includes having strong internal and inbound links.
very interesting post However I am a bit confused. You wrote:"I decided recently to canonicalize the weekly updates in the challenges to the main challenge post." this implies that those are completely different posts, although they do have a common thread. The G guide to canonicalisation (linked from Matt Cutts' most recent post) reads: "Must the content on a set of pages be similar to the content on the canonical version?Yes. The attribute should be used only to specify the preferred version of many pages with identical content (although minor differences, such as sort order, are okay). (...) would not be appropriate if that same site simply wanted a gel insole page to rank higher than the shoe page" I thought the canonical attribute should be used with almost identical content but different urls (for example), but apparently that's not the case, at least that's what happened to you? In which case it would make sense, since it's much easier than a redirect to implement! I'm afraid to risk it :) Keep up the good work!
You're technically correct - this was not a sanctioned use of the canonical tag. I'm going to argue that my usage in #1, while violating the letter of the law is consistent with the spirit of the law.
Essentially, these weekly updates aren't good stand-alone content. If someone searches for one of my challenge posts and gets "Week 2 Update", it isn't going to make much sense - in other words, I think it's a bad search user experience. These updates should remain available, though, in an archival sense, so I don't think a 301 is appropriate either. So, the canonical tag is my best alternative.
I also think these posts are conceptual duplicates, although that's a weaker argument. Any given challenge has a set of 5 posts, but only 1 of those posts really stands alone.
Of course, the classic approach would be just to link each update back to the main post, which I also do.
Yes I agree with you that it makes sense to use the canonical in that case! The user will find the main post and then browse all the updates straight from it. The reason why I wanted to try the canonical as well is because we have many old pages in our website that outrank (or somehow compete in the serps with) our new (much better, more informative) ones. This is a common problem, I know... however, we wouldn't do it just for seo purposes: the old pages are bad quality, sometimes google translated, low-quality design, full of typos etc (website localisation is the main culprit). The content isn't identical, although it's the same topic. It might be worth to try and experiment with a few of them and see how it goes. Thanks for your reply, looking forward to a new post ps forgot to say, redirect is not feasible as IT is too busy. Same old story :)
Yeah, the implementation angle can be tough. All else being equal, I'd say a 301-redirect may be a better fit there (the content isn't useful to anyone, it sounds like), but I understand the dilemma.
Thanks for the useful information. But there's still question. I wonder if other search engines (like Bing or Yahoo) support Cross-Domain Canonical tag or not.
If the answer is not, what can we do without 301 redirect (cause the policy of Google Adwords doesn't allow 301 redirect for landing page).
I love these kinds of tests. Thanks for sharing your findings!
Interesting points you raise, I like the one about the social based buttons very interesting stuff.
I have been testing a few canonical tricks on ezine based site will be interesting to see the results.
There's definitely more investigation needed on the social sharing aspects. Presumably if I put a canonical link in to make a page appear more popular then any likes would then go to the destination of the canonical URL; but what if I then remove the canonical? Or if the canonical destination disappears?
Facebook's use is inconsistent, and I suspect they're more savvy than to just take the tag at face value. I think this is a case where you also have to consider the user experience of your social network. If they Like one post and their Like "lands" on a completely different post, it's going to feel like a bait-and-switch.
Wow, wouldn't have thought to test Twitter or FaceBook's respect of the canonical attribute. Good job!
if both twitter and facebook honor the canonical tags... i can already see some abuses in the horizon to get a very high number on both of them >:))
Great tests, I just launced my own test using a cross domain canonicals. I feel like I'm checking Raven every morning to see if the traffic/rankings changed at all but I can't always expect results in the first 24 hours.
Great post Pete. Sometimes a bit messing arround (e.g: with rel canonical) helps to clear the clouds. Liked #6 though ;)
Cheers!!
Helpful post, thanks for sharing. I love the canonical tag to be sure and it's potential to sort out issues of authority, but it seems as if it could be abused to sculpt page rank in the way that nofollow links did a few years ago. I wonder if anyone knows of Google's approach and/or prevention of this. Have not seen anything authoritative on the web.
Awesome post Dr.Pete. It is really infomative post. This post helped me to make my collegues easily understand about Canonical tags. Keep on sharing post.
Its useless to write Canonical tag in head tag as google will read only Head tag for such Tags.
Cool story, very informative. Btw I can't really see trick here - just tests and half of them are not giving results cuz Google knows how to deal with rel=canonical.
"6 Informative Canonical Tests, Only Half of Which Worked" didn't seem like a particularly catchy title :)
Running this in your own website it make sound like a bad idea, but if you really want to get real data, that's the best way of getting it. Great post!
in a endomorphism which is nearest to Google (query / search results) the basic concept or reference material, a reduction in the endomorphism is easier if we diagonalize the matrix by causing appear up to zero. This final matrix call'd a matrix in a specific canonical base is a base that contains all the basics. By analogy, it is easier for a search engine to find pages in its index simply by following certain components (provide data (pages) linked to a standard that contains allthe other standards), where the importance of canonicalization pages. You need to know in terms of optimization it is preferable to add on pages that will form the basis or the graduation site.
I've seen number 4 in action - the syntax of the canonical wasn't correct and resulted in a non existant URL being indexed, as URL issued a 200 response instead of a 404. the interesting bit about this was how long it took google to update the indexed page once the change had been made, it took a few weeks.
and i'm not 100% certain but i'm pretty sure i've seen number 5 applied by accident too, and that got pretty painful...!
The feedback on the share buttons recognizing the canonical counts is very interesting. Great post!
very interesting subject, I waz testing this some days ago and still looking to see the results on Google.
Late to this blog post... sorry.
"(6) Facebook/Twitter Buttons"
Does this only happen with the facebook/twitter buttons? What about the +1?
Not sure, honestly - didn't have any sites with [+1] buttons running when I was collecting data. If anyone else knows, please leave a comment.
I aksed in Google Webmaster forum if we can duplicate the home page with different phone numbers and street addresses and then add a canonical tag on each copy toward a central page. I think that a centralized contact info is not always the way to go, so it makes sense that each local stores re-use the home page with their own contact info. In any case, it is interesting to know if Google will accept to merge pages with different phone numbers and street addresses under a same canonical. Some "top contributors" said that, because phone numbers and street addresses are important, it was very unlikely that Google will follow the canonical. So, I decided to make my little experiment on the canonical tag Vs local contact info. The results is that Google follows the canonical tag. After the experiment, another "top contributor" said that he never saw a case where Google did not follow a canonical tag because of page differences. I felt that this was interesting, but would it always be like that? If Google starts to check for page differences, which is certainly possible, what will be the criteria? So, I keep my little experiment running.
Very interesting, Dominic - thanks for sharing. My experiences are similar - Google trusts an internal canonical a bit too much, even allowing it for wildly different pages. It's tough to say whether they pass link-juice the same way between different pages, but they certainly will honor the indexing signal.
If they do not pass the link juice, it means that the canonical is not followed. One "top contributor" mentioned to me that she lobbyed to have Google sends a signal to the webmaster when a canonical was followed. I largely support this proposal. The webmaster should at the least know whether or not the canonical was succesful. They have this kind of signals for cross-domain merging of URLs (see https://googlewebmastercentral.blogspot.com/2011/10/raising-awareness-of-cross-domain-url.html ). They should have a similar signal for the link canonical, even within a same domain.
I think it depends what you mean by "followed". I see it as a two-stage process: (1) Google honors the indexation cue and de-indexes the non-canonical page in favor of the canonical version, and then (2) Google passes the link-juice from the non-canonical to the canonical. If they do (1) but not (2), I'd still say they "followed" the canonical, but obviously the end result isn't as desirable. Unfortunately, while (1) is pretty easy to measure, (2) is extremely difficult.
At the least for me, the most natural intepretation of "followed" is that the canonical link is managed as when we follow a 301 redirect. In what way the link would be "followed" if nothing is passed to the target? Irrespective of terminology, my mainpoint was that, as they already have for cross-domain merging, we should have a way to measure (2), i.e., whether the link juice is passed. Now, after your comment, I must add that we should have a non ambiguous term to mean that the link was (fully) respected. I am very surprised that "followed" is not that term. Are you sure that Google uses it with a different meaning. Wow ! Added to the fact that we cannot even measure it, that would be so confusing.
I know it's late buuuuut:
If you canonical a landing page to itself?
What could happen there?
Bad or Good? Or nothing?
Having a page canonical to itself is perfectly fine, in theory, but it depends a lot on what "itself" means - remember, that canonicalization is about URLs, not pages/files on the server. So, it's really tough to say without details.
@Dr. Pete - this article is fantastic. I have a question. I work with a lot of small businesses and most of them have 2 URLS for the same page. They have a profile page set up on the corporate website (let's say corporatewebsite.com/bobsmith) but then they buy a domain name to advertise themselves (let's say bobsmith.com) and when they're creating the Facebook Page for their business and putting themselves on yellowpages directories they always list bobsmith.com. Essentially not much links to the corporatewebsite.com/bobsmith URL.
The corporate office actually contracts with the company that these franchisees buy their domains from and has gone and put canonical tags on all the domains (including bobsmith.com) saying that corporatewebsite.com/bobsmith is the original. They also put canonical tags on the corporatewebsite.com/bobsmith copy as well saying corporatewebsite.com/bobsmith is the original (redundant but as far as I can tell, doesn't do any harm).
However, all links out there on any site for the most part are pointing to bobsmith.com. So I was wondering if this falls under what you said in point 5 where it can confuse the search engines. What I've been discovering so far is that what page Google has indexed changes pretty frequently. I'll look one day and the page ranking for keywords we'd want it to is the bobsmith.com version. Then next week I'll look and it's the companywebsite.com/bobsmith version. I haven't seen too many instances where they both rank since the canonical tags were added.
The possible negatives I see in this scenario is that it's annoying because it's hard to track ranking unless you're tracking both domains but other than that are there any negative outcomes? I've found that since one of the 2 always seems to rank it doesn't matter too much but what I was more afraid of is if this "confusion" could result in both pages having a weaker ranking for a particular keyword. Is that possible? If the keyword I'm trying to rank for is "Bob Smith is cool", I've found that sometimes bobsmith.com is there on the first page, sometimes it's companywebsite.com/bobsmith. I just wanted to see if this would result in eventually neither ranking high for it.
Hopefully that wasn't too confusing :p
It is a mixed signal, in that you're linking consistently to the bobsmith.com domain, but I think you're probably doing it right. One domain is for branding/simplicity, but you still want the corporate site to get the link equity and be the destination in most cases. You could also just 301-redirect to the corporate site and only use the vanity domain as a placeholder.
Thank you Pete! The only reason why they shouldn't/don't do the 301 redirect is because a lot of the small businesses do Adwords and domains that are 301 redirects don't work on Adwords. They don't use the companydomain.com/bobsmith for Adwords either because the corporate office sometimes has ads and Google wouldn't show both due to their double-serving policy.
Thanks man the tips and tricks work great for me
I learned a lot about that "canonical" thing today! Thanks to this post! :D
Good testing. Very interesting
Thanks
Great article Pete! I could actually share a canonical experience I have with one of my clients. You would have thought that people can't get it that worng but hey...they can:)
This website had an issue with duplicate pages ( domain.com/x and domain.com/x/y present the same content) so they placed canolical links on them but instead of pointing to 1 preffered version of the URL they did the following:
on domain.com/x <link href="https://www.domain.com/x"/>on domain.com/x/y <link href="https://www.domain.com/x/y"/>The amount of internal links pointing to those pages is exactly the same - links to them are in the site's main navigation only.Re external links, /x version has slightly more (nothing major though)Both pages are indexed and ranking for various keywords relevant to the content of the page however the page /x/y (which isn't preffered version) seems to get more exposure on brand related keywords and a few important generics.It seems like google is completely confused in this case! I'm currently waiting for the tech team to implement the changesso let's see how the situation change after that:)
Ouch. You know that someone said "Use the canonical tag", and so they did, and to someone it made perfect sense. That's why I hesitate every time I give out that kind of advice. We forget that these are complex topics and most people don't really understand the tools that we use every day.
Totally agree, every recommendation should come with 'how to implement it' guide and we shouldn't just assume that the tech team will know what and how to do. Although the conversations we have in nowadays with them are much better and they do understand and appreciate the SEO aspects more, it's still our responsibility to guide them through the process:)
Does anyone have any thoughts on what would happen if a cross domain canonical pointed at a page that was 301 redirected? I recently implemented a bunch of cross domain canonicals which worked great. There is now pressure to change the url structure on the new site (where the canonicals are pointing) and 301 everything to the newly structured urls.
Even chaining 301s isn't the greatest idea, especially long-term. I'd either change the canonical tags to the new URLs or do one set of 301 redirects. At any rate, try to avoid having two directives for any given situation (if some people get canonicals and some get 301s, that's ok, but not 2 "hops" for 1 visitor, so to speak).
Dear Dr. Pate
Your experiment is really excellent. 301 redirects is not a right solution in ling run. I always use 302 to resolve canonical issue. Dear pete please tell something about 302 redirect..............
The crossing the streams test was pretty interesting!
Thanks,
Storwell
a good collection of canonical tests, thenks Pete, I personally found your post very useful
How can you link from a .blogspot or .blogger to new site not on a third party subdomain?
This is great - I like it when people dare to do things we don't dare to do ourselves with our own sites and this is really useful, so thanks.
(P.S. I much prefer this comments box integrated into the page than a separate popup - thanks seoMoz!)
HI Pete... Thanks for experimenting... I really liked the 5th point. I shoudl admit it was a smart experiment ;-)
After this post Pete, now i will deeply consider the Canonical tag , this thing will realy makes the difference thanks for shared a lovely post man love it !!!!
Indeed sharing those kind of results based on real tests is really interesting and useful
Thanks a lot
I really enjoyed this post Dr. Pete. Canonical tags are pretty scary, so it's good to see some tests and results, and it's also reassuring that Google have 'idiot proofed' them to some extent. Would love to see more of these types of posts on SEOmoz, as I don't think these areas/specifics get enough attention. - Jenni
Nice Dr Pete,
I like the cross stream double SERP action. Be interesting to scale this over 100 pages and see if it gives you more click throughs. Great way to grab more of the natural listing real estate.
I am NOT recomending people do this BTW, just be interesting to do the test.
Kind Regards,
Allan
Yeah, I'm NOT going to be running that experiment :)
These kinds of results are always good to see - both in that they support what Matt Cutts says (so that those of us who do folow him like a sheepdog can breath a sigh of relief) and in that they show that while Google gives some leeway in making sure that we've employed the tag correctly it's not a good idea to do something completely stupid.
I need to start doing this kind of experiment - damn work getting in the way!
Really great experiments Pete, I hope you'll continue your tests and share with us.
Great job!
Jon
Cool insights!
Calming that 3) and 4) are not working ... otherwise Google could stop supporting the canonical tag...
Makes me want to go back through and think up a few tests. Maybe even start a novelty site where I test each of Matt's claims like you accidentally did. Funny how you paralleled him for a few of those.
Awesome post. Thank you.
So looking at experiment #1 if I have a blog at blog.mysite.com could I do a canonical to mysite.com/blog and would this help DA?
Sweet post Dr. Pete! I particularly liked your social media test & insights. Still chuckling over "Don't Cross The Streams" too :)
because "that would be bad" :)
Very cool Dr. Pete. Thanks.
Google creating their own Title tags? Possibly preparing a SERP that poses a pic, title, description, and website URL that they create from website information?
Great test theories and results. Thanks for sharing.
I'm sure there's a post about Google displaying their own versions of title tags on SEOmoz somewhere. It happens with some of our rankings, for example if you search for our brand and then a type of hosting, the same page of the same site for each set of results can show different titles. - Jenni
Not sure why my comment was voted down; if it was you, could you give some feedback on that please? :) - Jenni
Pete thank you for this article. (sorry for English, I'm French ...).I have a question about duplicate content.I am a brand site and have product descriptions. The problem ismany sites (sites of e-commerce, shopbots, for example) copy my content on their site.What would be the best solution for not having duplicate content?Thank you very much for your answers.A.
Unfortunately, unless they're stealing your pages wholesale and picking up your canonical tags (in other words, you use a canonical tag and they copy your header), there's not a lot that you can about scraping from a purely technical standpoint. If you have any control over how the content is syndicated, you could have them link back to you or use the canonical tag, but in these situation, most sites want to rank for the product pages as well, so they aren't going to want to do that.
So awesome. It is really interesting how Facebook and Twitter are recognizing the tag. I would have never thought that they would have. Thansk for the great post.
I never thought about it either but it does make a lot of sense, since they are basically each trying to base their numbers off of a database associated with raw URLs, and a lot of URLs can have extra things on the end depending on how people arrived that the page that have nothing to do with the valuable content on the page, and it's clearly better for everyone to see the numbers aggregated as single pages.
(So I guess that means if you wanted to share a competitor's link but add a ¶meter=blahblahblah to your link so it wouldn't count towards their real URL's OpenGraph total... well maybe it actually will, mwahahahaha.) Never would have thought about it, but it's awesome that it does!
It kind of makes sense, but I wonder if Facebook reversed it after "Like" became more like "Share". With my example in #1, you could end up Liking a post that you didn't even read, in theory, which is kind of weird. Of course, my usage in #1 isn't 100% kosher.
Hey Pete, how about adding this one to your next play in the world of canonicals? ;)
https://www.seomoz.org/q/effect-of-rel-canonical-on-links
Hadn't thought about the impact on internal links. I have to admit that one of the things I like about testing the canonical tag is that it usually either works or doesn't work and is really easy to measure. Link impact is so tough to wrap a number around. I guess you'd have to cheat - create one link on the non-canonical page that was unique (that nothing else linked to) and see if the page it linked to got indexed. Or, take the classic approach and use unique anchor text on that one link.
Great article, I really enjoyed reading this.
Quick question, does anyone know if the SEO MOz pro campaign web app takes canonicals into consideration when looking at duplicate content ?
This is great information, thanks for posting this. This validates some of the tests I've been trying and some other stuff I thought about. Good stuff.
Great Post! This actually answered quite a few questions I already had, since I rarely have an instance to really use the canonical tag.
Wow, i would never have thought of pointing the "update" blog posts at another "main blog post". That seems quite bizarre to me - arent you negating all the positive effects of blogging all those updates? Instead of google seeing you adding content to your site daily you're telling google that you just have one page/week.
Outside of an experiment, you wouldnt do this would you?
Edit: Sorry #1 and #6 were both intentional. The comments below apply to #6 (not #1):
Actually, I did - #1 was the only one of the 6 that wasn't an experiment, in the sense that I deliberately made the change with long-term intent. New content is good to a point, but I don't think that the value of adding a post/day vs. a post/week, from an SEO standpoint is all that much. On the flip side, the downside of pouring out tons of low-quality content is clear, especially post-Panda. Controlling your index can have a big positive impact, in my experience.
For now, on my blog, I don't think the SEO repurcussions are that strong either way. Long-term, though, I thing focusing on the best post in the series and culling the mini-updates will strengthen my SEO considerably.
There are exception, of course, such as a news site (and News search), but for most sites, just pushing out new content every day isn't the advantage it once was - at least, in the SEO sense.
Great post Dr. Pete! I'm most interesting the facebook and twitter buttons. Extremely facinating.
Great information! I love how Facebook and Twitter respect the canonical. I noticed this about one of my own domains. I tried using Addthis to share a url and when it posted to Twitter it shared the canonical. So, if someone lands on a non-canonical url and shares it, the proper url gets the credit. Very cool!
Awesome read. Thanks Dr. Pete. Once again I'm ultra impressed by your ability to have fun running experiments like this. These are nearly all things I've been wondering, but never had the time to test on my own.
We'll be doing a lot of canonical tags on our new site. I'll make sure to point my developers to this post so they don't send any mixed signals :)
Im interested to know the effects of using rel=canonical on detail pages that have effectively been scraped from other domains not related to your own. This could be used if you want to provide information to your users for user experience but not claim it as your own. Would this only impact that particular page or with many rel=canonicals could this have further effects on your domain?
Dr. Pete great tests and great job!
Love it ! take a quick peek at Rand's post as well back in 2009 and now we are gonna apply it.
-- Russell
Great experiments. The first one is very timely for me. I've a had a personal blog for a while that has some good history, but just aunched a new SEO blog this month. One of my most trafficked pages on my personal blog was about social media marketing but I thought it was a better fit on my NEW SEO blog. Just yesterday I moved that post from my personal blog to the SEO blog, added a "This post has moved" link on the personal blog (did not want a 301 redirect) and was wondering just this morning if a cross domain canonical tag on the personal blog to the SEO blog would work...I guess it will!
Really useful post thanks :)
I knew already about the cross domain canonical, but I was surprised to see that Facebook and twitter got a hold of the canonical tag. There is more beneath the eye than it seems...