Can hreflang replace canonicalisation ?
-
Hi
Im working with a site that has ALOT of duplicate content and have recommended developer fix via correct use of Canonicalisation i.e the canonical tag.
However a US version (of this UK site) is about to be developed on a subfolder (domain.com/uk/ & domain.com/US/ etc so also looking into adopting the hreflang attribute on these.
Upon reading up about the hreflang attribute i see that it performs a degree of canonicalisation too.
Does that mean that developing the international versions with hreflang means there's no need to apply canonicalistion tags to deal with the dupe content, since will deal with the original dupe content problems as well as the new country related dupe content, via the hreflang ?
I also understand that hreflang and canonicalisation can conflict/clash on different language versions of international subfolders etc as per: http://www.youtube.com/watch?v=Igbrm1z_7Hk
In this instance we are only looking at US/UK versions but very likely will want to expand into non english countries too in the future like France for example.
So given both the above points if you are using hreflang is it advisable (or even best) to totally avoid the canonical tag ?
I would be surprised if the answers yes, since whilst makes logical sense given the above (if the above statements are correct), that seems strange given how important and standard best practice canonical usage seems to be these days.
What best ? Use the Hreflang alone, or the Canonical tag alone or both ?
What does everyone else do in similar situation ?
All Best
Dan
-
Thanks Tommy
Its just a .com at the moment but language versions will be on subfolders .com/us and .com/uk with hreflang handling the rest of the internationalisation in conjunction with webmaster tools etc
All the dupe content is on the .com
cheers
dan
-
That great info thanks for clarifying that Lynn !
Im sure ill have some further questions though when going through the process with the developers
Many Thanks
Dan
-
Hi Dan,
Which subdomain do you have a lot of duplicate content? If you have duplicate content for the EN-US, then I would use canonical tags for the US content. However, when you create a new sub-domain for UK, I would use hreflang on those contents to avoid causing duplicate content issues with the US contents that are the same.
-
Hi Dan
The hreflang will tell google that different pages are targeted to different languages/regions, but it will not help solve your existing dupe content problems as you describe them. So for these you should use the canonical tag.
Hreflang and canonical tags can be used together but can cause issues if you are not careful. Main thing to be sure of is that you keep any canonical tags within the same lang/region pages. You do not want to be having a US page with a canonical tag pointing to a UK page in other words.
Having multiple UK pages with a canonical pointing to the main UK page and the main UK page having an hreflang tag to its US counterpart is ok (as is having a self referencing canonical tag on the main UK and US pages if needed). Google mentions here that this is an acceptable way to handle canonicalisation issues in an international setup.
What you would also want to avoid is having hreflang tags on your dupe content pages in the UK version. Where would they go? to the US version of the duplicate content? To the main canonical version of the US content? That could get messy pretty quickly. So the trick is to consolidate your UK pages with rel canonical and then properly markup your final regional variations with hreflang.
Hope that makes sense!
-
Thanks for commenting Tommy !
The problem is though that the site already has lots & lots of duplicate pages, so want the developers to fix that, and i have been advising they do this via canonical tag since standard practice.
So are you saying that if we use hreflang then no need to use canonical for this existing duplicate content since using hreflang on the new international pages will deal with the old dupe content as well as the new uk/us dupe content too ?
cheers
dan
-
Hi Dan,
To my understanding on hreflang, since the UK site is only targeting searchers in UK, i don't think you will need the canonical tag and there is no reason to use it. As the canonical tag is to search engines find the original piece of content and avoid duplicate content. However, since the hreflang already tells the search engine that the UK page is targeting UK, it wouldn't appear for people that searches in US so duplicate content wouldn't be an issue.
In the end, if you are targeting different countries with the hreflang, i think that is enough. I would only use canonical if the same content is targeting searchers in the same geo location.
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
-
Can I safely delete unused tags??
Can I safely delete unused tags, ones with 0 posts connected to them? Thanks, Mike
Technical SEO | | naturalsociety0 -
How can I get Google to forget an https version of one page on my site?
Google mysteriously decided to index the broken, https version of one page on my company's site (we have a cert for the site, but this page is not designed to be served over https and the CSS doesn't load). The page already has many incoming links to the http version, and it has a canonical URL with http. I resubmitted it on http with webmaster tools. Is there anything else I could do?
Technical SEO | | BostonWright0 -
Blocked URL parameters can still be crawled and indexed by google?
Hy guys, I have two questions and one might be a dumb question but there it goes. I just want to be sure that I understand: IF I tell webmaster tools to ignore an URL Parameter, will google still index and rank my url? IS it ok if I don't append in the url structure the brand filter?, will I still rank for that brand? Thanks, PS: ok 3 questions :)...
Technical SEO | | catalinmoraru0 -
Can too many pages hurt crawling and ranking?
Hi, I work for local yellow pages in Belgium, over the last months we introduced a succesfull technique to boost SEO traffic: we have created over 150k of new pages, all targeting specific keywords and all containing unique content, a site architecture to enable google to find these pages through crawling, xml sitemaps, .... All signs (traffic, indexation of xml sitemaps, rankings, ...) are positive. So far so good. We are able to quickly build more unique pages, and I wonder how google will react to this type of "large scale operation": can it hurt crawling and ranking if google notices big volumes of content (unique content)? Please advice
Technical SEO | | TruvoDirectories0 -
While SEOMoz currently can tell us the number of linking c-blocks, can SEOMoz tell us what the specific c-blocks are?
I know it is important to have a diverse set of c-blocks, but I don't know how it is possible to have a diverse set if I can't find out what the c-blocks are in the first place. Also, is there a standard for domain linking c-blocks? For instance, I'm not sure if a certain amount is considered "average" or "above-average."
Technical SEO | | Todd_Kendrick0 -
600+ Visitors a day after 6 months, can you do it?
So since the Penguin update the clients of the company I work for have gradually been losing traffic and money. Noone (except for me) has noticed this yet and connected the dots. Yesterday we all get called in to have a bollocking and the manager asks the head of our department if he would be confident of being able to get 600+ visitors a day to an 'average website' that has just started up, to which he replied 'yes'. Since I started here back in February there has not been a single new client that has been able to gain that many visitors (many have not gained even 25% of that figure), which in the post-panda and post-pegnuin world, I find completely understandable. When I first started here they were using SEO 'tactics' which people used to employ 5+ years ago and didn't even use exact match keyword data.I have had a few talks with them about how SEO has changed over the last few years and they still don't seem to understand that it is now significantly more difficult to gain traffic using SEO than it once was. If you were asked about the same question, thinking about the 'average' client you might get, would you be confident enough to guarantee that at the 6 month mark they would be getting 600+ visitors a day?
Technical SEO | | Kinsel0 -
Can someone help me get this site ranked? www.2sponsors.com
Hi, I am have been trying for months to get a site ranked for one of my customers and I am not doing very well. I have been doing SEO for years and have gotten lots of sites ranked but this one has been the most difficult. Does anyone have time to look at it for me? Thanks The sites PR=4. I am trying to get it ranked in www.google.com.ar Thanks Carla skype: carla.dawson78
Technical SEO | | Carla_Dawson0 -
Canonicalisation - different languages and channels
Hi If the same content is placed on different URL's for the purposes of providing information on different channels (i.e mobiles), or has been translated into a different language (but is still the same content), do the serach engines still count this as duplicate content and will a canonical URL have to be tagged in these instances? Thanks in advance for your assistance.
Technical SEO | | jimmyseo1