Hreflang and canonicalization
-
When using hreflang in order to deliver the relevant version in SERs, should we also make use of a reference to a canonical version to avoid duplication?
Currently, we provide different regional versions of our content where the content is largely the same aside from minor changes due to spelling, units of measurement although occasionally larger amends are required.
We have implemented hreflang referencing all the alternative country Urls, e.g en-us, en-gb, en-aus etc but also specificied the canonical as the en-gb version since we are a UK based website and the majority of the content originated from the UK version of our site.
Recently, our rankings across all countries have been falling markedly and I'm wondering whether the canonical element may be at fault. We have not been engaging in any black hat activities that might have been responsible for any sort of fall.
When we implemented the hreflang and canonical in July 2012 our traffic has actually been increasing significantly until literally 21 Nov when the search traffic is plummeting considerably across all countries. It would be useful to know if you need to specify a canonical version when using hreflang or could there be another reason for our ranking falls.
Many thanks in advance of your assistance.
-
you can test it out and remove the canonical for the not fully equivalent pages ... and unfortunately there is no other solution than a canonical to fix the pages that have a fully equivalent content.
just test it out and keep a close eye on it and please do update this thread
thank u
-
Thanks Wissam. I posted the same question in a Google forum and was told that I should remove the canonical reference (but retain the hrelang elements) as some of the content was not entirely identical and had regional differences.
I've asked whether I should do the same (i.e not specify a canonical) when the content is entirely identical but equally relevant to different countries. Would the hreflang be enough to prevent them being considered duplicate?
-
Hi Simon
I think the implementation you did on the site is confusing and wrong.
you consolidated ur au to the .com domain without specifying which folder or subdomainis the au section is.
previously because you have the .com.au in the domain Google understood that signal that this website is relevant to au visitors. but when you consolidated to the .com you need now to TELL or HINT to Google (through Google Webmaster Tools) where the whole domain that was targeting this country went.
and HREFLANG is not about Geotargeting but about the Language.
-
Hi Wissam, yes indeed all the pages are informative article pages. I want each country specific version to rank highly in it's own country i.e en-us article to rank in US, en-au in Australia etc. Does specifying a canonical strangle your ranking in all the other countries?
-
Google has actually updated their Google webmaster help section of the hreflang and remove the reference of rel canonical because people tend to get confused and implemented incorrectly.
so my question to you are these pages informational pages? are they fully equivalent to the others ? in aus and us ?
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
International Blog Structure & Hreflang Tags
Hi all, I'm running an international website across 5 regions using a correct hreflang setup. A problem I think I have is that my blog structure is not standardized and also uses hreflang tags for each blog article. This has naturally caused Google to index each of the pages across each region, meaning a massive amount of pages are being crawled. I know hreflang solves and issues with duplication penalties, but I have another question. If I have legacy blog articles that are considered low quality by Google, is that counting against my site once or multiple times for each time the blog is replicated across each region? I'm not sure if hreflang is something that would tell Google this. For example, if I have low quality blog posts: blog/en-us/low-quality-article-1
Intermediate & Advanced SEO | | MattBassos
blog/en-gb/low-quality-article-1
blog/en-ca/low-quality-article-1 Do you think Google is counting this as 3 low quality articles or just 1 if hreflang is correctly implemented? Any insights would be great because I'm considering to cull the international setup of the blog articles and use just /blog across each region.0 -
x-default hreflang attribute value
Hi When is it best to use the x-default hreflang attribute value tags? Screaming frog is flagging lots as missing, but is it necessary? Thanks
Intermediate & Advanced SEO | | BeckyKey0 -
Hreflang : mixing with/without country code for same language
Hello, I would like to display 3 different english versions of my website : 1 for UK, 1 for CA and 1 for other english users. It would look like this for a page: . (english content with £ prices) <link rel="alternate" href="https: xxx.com="" en-ca" hreflang="en-CA">(english content with $CA prices)</link rel="alternate" href="https:> <link rel="alternate" href="https: xxx.com="" en="" " hreflang="en">(english content without currency)</link rel="alternate" href="https:> I wonder if I can mix this hreflang without country code with hreflangs with country code for the 2 other specific versions... or if the hreflang without country code version will appear whatever the country, even if i specified it . In other terms, is hreflang="en" > hreflang="en-CA" + hreflang="en-GB" if tagged together on a same page? Thank you
Intermediate & Advanced SEO | | AlexisH0 -
Hreflang Tags & Canonicals Being Used
We have a site on which both hreflang tags and canonicals are being used. There are multiple languages, but for this I'll explain our problem using two. There are a ton of dupe page titles coming up in GSC, and we're not sure if we have an issue or not. First, the hreflang tags are implement properly. UK page pointing there, US page pointing there. Further down the page, there are canonical tags - except the UK canonical tag points to the UK page, and the US version points to the US page. I'm not sure if this will cause an issue in terms of SEO or indexing. Has anyone experienced this before or does anything have any insight into this? Thanks much! Matt
Intermediate & Advanced SEO | | Snaptech_Marketing0 -
How to take out international URL from google US index/hreflang help
Hi Moz Community, Weird/confusing question so I'll try my best. The company I work for also has an Australian retail website. When you do a site:ourbrand.com search the second result that pops up is au.brand.com, which redirects to the actual brand.com.au website. The Australian site owner removed this redirect per my bosses request and now it leads to a an unavailable webpage. I'm confused as to best approach, is there a way to noindex the au.brand.com URL from US based searches? My only problem is that the au.brand.com URL is ranking higher than all of the actual US based sub-cat pages when using a site search. Is this an appropriate place for an hreflang tag? Let me know how I can help clarify the issue. Thanks,
Intermediate & Advanced SEO | | IceIcebaby
-Reed0 -
Hreflang many to one relationship (And canonical)
I have a question about the use of hreflang. In country one i have for instance web.one/BlueWidget web.one/YellowWidgetIn country (and language) two I have web.two/Widget In other words, the distinction between yellow and blue isn't there in country two. My best try would beweb.one/BlueWidget two" href="<a>http://</a>web.two/Widget" />web.one/YellowWidget <link rel="<a>alternate" hreflang="two" href="<a>http://</a>web.two/Widget" /></link rel="<a> <link rel="<a>canonical" href="web.one/BlueWidget"/></link rel="<a>web.two/Widget <link rel="<a>alternate" hreflang="one" href="web.one/BlueWidget" /> </link rel="<a> <link rel="<a>canonical" href="web.two/Widget "/></link rel="<a>So pointing both blue and yellow to the Widget in country two, and pointing widget in country two to Blue widget in country one, as BlueWidget is what I would prefer to promote if I had to choose.What do people think? Is this the way to go? If so, why? If not, why?And is it actually worth bothering with? Will country two get a boost because I get links to country one? Regards
Intermediate & Advanced SEO | | Peekabo0 -
Will this internal linking feature cause canonicalization issues?
This is a canonicalization type question, so I believe it should be a pretty straightforward answer. I just haven't had much experience with using the canonical tag so I felt I should ask so I don't blow up my site 🙂 Ok, let's say I have a product page that is at: - www.exampledomain.com/products/nameofproduct Now on that page I have an option to see all of the specs of the product in a collapsible tab which I want to link to from other pages - So the URL to this tab ends from other pages ends up being: - www.exampledomain.com/products/nameofproduct?=productspecs This will link to the tab and default it to open when someone clicks that link on another page. Correct me if I'm wrong, but if I understand canonicalization correctly I believe creating this link is going to cause a duplicate page that has the opportunity to be indexed and detract from our SEO to the main product page. My question is... where do I put the "rel=canonical" tag to point the SEO value back to the main page since the page is dynamically generated and doesn't have its own file on the server? - or do even need to be concerned with this? Feel free to correct me if I'm wrong on any of the above. Like I said - this is something I am fairly familiar with how it works, but I haven't had much experience with using. Thanks!
Intermediate & Advanced SEO | | CodyWheeler0 -
Canonicalization issue? - URLs with and without trailing slashes showing up as unique
For some reason, our urls are set to change from “www.apprenda.com/ANYTHING" to “apprenda.com/ANYTHING” These register as different pages though? We have rankings in SEOMoz Pro for terms where our homepage shows up 6th on google, but SEOMoz says it's not on the first page because it's checking against apprenda.com and not www.apprenda.com Also, it seems like for some reason pages with trailing slashes also register differently than those without. Should we be doing something for that? Something to make sure all pages get rewritten to having the trailing slash or not? For instance, this url: http://apprenda.com/saasgrid/features/multi-tenancy/ and this url” http://apprenda.com/saasgrid/features/multi-tenancy are really the same page. Yet in our analytics, they register as different pages with their own stats, etc. What should we do in our particular case, and how can we get this fixed? I really appreciate the help, and thanks in advance! Jesse
Intermediate & Advanced SEO | | ApprendaPlatform0