International SEO - cannibalisation and duplicate content
-
Hello all,
I look after (in house) 3 domains for one niche travel business across three TLDs: .com .com.au and co.uk and a fourth domain on a co.nz TLD which was recently removed from Googles index.
Symptoms:
For the past 12 months we have been experiencing canibalisation in the SERPs (namely .com.au being rendered in .com) and Panda related ranking devaluations between our .com site and com.au site.
Around 12 months ago the .com TLD was hit hard (80% drop in target KWs) by Panda (probably) and we began to action the below changes. Around 6 weeks ago our .com TLD saw big overnight increases in rankings (to date a 70% averaged increase). However, almost to the same percentage we saw in the .com TLD we suffered significant drops in our .com.au rankings. Basically Google seemed to switch its attention from .com TLD to the .com.au TLD.
Note: Each TLD is over 6 years old, we've never proactively gone after links (Penguin) and have always aimed for quality in an often spammy industry.
**Have done: **
- Adding HREF LANG markup to all pages on all domain
- Each TLD uses local vernacular e.g for the .com site is American
- Each TLD has pricing in the regional currency
- Each TLD has details of the respective local offices, the copy references the lacation, we have significant press coverage in each country like The Guardian for our .co.uk site and Sydney Morning Herlad for our Australia site
- Targeting each site to its respective market in WMT
- Each TLDs core-pages (within 3 clicks of the primary nav) are 100% unique
- We're continuing to re-write and publish unique content to each TLD on a weekly basis
- As the .co.nz site drove such little traffic re-wrting we added no-idex and the TLD has almost compelte dissapread (16% of pages remain) from the SERPs.
- XML sitemaps
- Google + profile for each TLD
**Have not done: **
- Hosted each TLD on a local server
- Around 600 pages per TLD are duplicated across all TLDs (roughly 50% of all content). These are way down the IA but still duplicated.
- Images/video sources from local servers
- Added address and contact details using SCHEMA markup
Any help, advice or just validation on this subject would be appreciated!
Kian
-
Hey Simon,
The Australia is lang="en-au"
The UK is lang="en-gb"
The US is lang="en-us"
We've tried to keep these as tight per country as possible so opted not to use the straight 'en'.
In analytics, there has been some reduction is language referrals, mainly "en-gb" falling from the number one language type for the US site, which is a positive. Interstingly enough, once we removed the .co.nz fro mthe Index the .com site remove in to dominate the SERPs for brand and some core-KW searches in Google.co.nz.
Its a little unfortunate as Panda, from my understanding, is keen to spare ccTLDs from any harsh devaluations, but we'll hopefully be able to hit whatever threshold for % of unique content in the near future.
We have review functionality planned for each TLD which should help add value to existing duplicate content. Once this is up and I have some more robust data I'll pull a post together for YouMoz.
Thanks for the feedback!
Kian
-
Wow, that's a pretty comprehensive list of actions you've compiled there and you seem to have covered pretty much all the bases. I almost think your post should be promoted on Youmoz as a great step of actions for targeting regional websites.
My experience of hreflang is that it is not perfect in that you occasionally get the wrong versions of pages served in SERPs. I wonder do you specify the .com as 'en' in the hreflang mark up in order that it is the generic English language version as opposed to being country specific?
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 SEO and Indexing in Country Specific Search Engines
Hey everyone! My company has recently migrated to a new domain (www.napoleon.com) which includes migrating many TLD separate domains to the new. We have structured the website to have multi-language and regions, including regional content, product offerings etc. Our structure is as follows just to give an example. napoleon.com/en/ca/
Intermediate & Advanced SEO | | Napoleon.com
napoleon.com/fr/ca/
napoleon.com/en/us/
napoleon.com/de/de Currently, specifically the homepage version of the USA website is indexing into Canadian Search Engines, and I can't figure out why. It has been roughly 6 weeks since launch. Any thoughts on this? Thank you Dustin0 -
Same site serving multiple countries and duplicated content
Hello! Though I browse MoZ resources every day, I've decided to directly ask you a question despite the numerous questions (and answers!) about this topic as there are few specific variants each time: I've a site serving content (and products) to different countries built using subfolders (1 subfolder per country). Basically, it looks like this:
Intermediate & Advanced SEO | | GhillC
site.com/us/
site.com/gb/
site.com/fr/
site.com/it/
etc. The first problem was fairly easy to solve:
Avoid duplicated content issues across the board considering that both the ecommerce part of the site and the blog bit are being replicated for each subfolders in their own language. Correct me if I'm wrong but using our copywriters to translate the content and adding the right hreflang tags should do. But then comes the second problem: how to deal with duplicated content when it's written in the same language? E.g. /us/, /gb/, /au/ and so on.
Given the following requirements/constraints, I can't see any positive resolution to this issue:
1. Need for such structure to be maintained (it's not possible to consolidate same language within one single subfolders for example),
2. Articles from one subfolder to another can't be canonicalized as it would mess up with our internal tracking tools,
3. The amount of content being published prevents us to get bespoke content for each region of the world with the same spoken language. Given those constraints, I can't see a way to solve that out and it seems that I'm cursed to live with those duplicated content red flags right up my nose.
Am I right or can you think about anything to sort that out? Many thanks,
Ghill0 -
SEM Rush & Duplicate content
Hi SEMRush is flagging these pages as having duplicate content, but we have rel = next etc implemented: https://www.key.co.uk/en/key/brand/bott https://www.key.co.uk/en/key/brand/bott?page=2 Or is it being flagged as they're just really similar pages?
Intermediate & Advanced SEO | | BeckyKey0 -
What is considered duplicate content?
Hi, We are working on a product page for bespoke camper vans: http://www.broadlane.co.uk/campervans/vw-campers/bespoke-campers . At the moment there is only one page but we are planning add similar pages for other brands of camper vans. Each page will receive its specifically targeted content however the 'Model choice' cart at the bottom (giving you the choice to select the internal structure of the van) will remain the same across all pages. Will this be considered as duplicate content? And if this is a case, what would be the ideal solution to limit penalty risk: A rel canonical tag seems wrong for this, as there is no original item as such. Would an iFrame around the 'model choice' enable us to isolate the content from being indexed at the same time than the page? Thanks, Celine
Intermediate & Advanced SEO | | A_Q0 -
How to remove duplicate content, which is still indexed, but not linked to anymore?
Dear community A bug in the tool, which we use to create search-engine-friendly URLs (sh404sef) changed our whole URL-structure overnight, and we only noticed after Google already indexed the page. Now, we have a massive duplicate content issue, causing a harsh drop in rankings. Webmaster Tools shows over 1,000 duplicate title tags, so I don't think, Google understands what is going on. <code>Right URL: abc.com/price/sharp-ah-l13-12000-btu.html Wrong URL: abc.com/item/sharp-l-series-ahl13-12000-btu.html (created by mistake)</code> After that, we ... Changed back all URLs to the "Right URLs" Set up a 301-redirect for all "Wrong URLs" a few days later Now, still a massive amount of pages is in the index twice. As we do not link internally to the "Wrong URLs" anymore, I am not sure, if Google will re-crawl them very soon. What can we do to solve this issue and tell Google, that all the "Wrong URLs" now redirect to the "Right URLs"? Best, David
Intermediate & Advanced SEO | | rmvw0 -
Frequent FAQs vs duplicate content
It would be helpful for our visitors if we were to include an expandable list of FAQs on most pages. Each section would have its own list of FAQs specific to that section, but all the pages in that section would have the same text. It occurred to me that Google might view this as a duplicate content issue. Each page _does _have a lot of unique text, but underneath we would have lots of of text repeated throughout the site. Should I be concerned? I guess I could always load these by AJAX after page load if might penalize us.
Intermediate & Advanced SEO | | boxcarpress0 -
Duplicate Content Question
My client's website is for an organization that is part of a larger organization - which has it's own website. We were given permission to use content from the larger organization's site on my client's redesigned site. The SEs will deem this as duplicate content, right? I can "re-write" the content for the new site, but it will still be closely based on the original content from the larger organization's site, due to the scientific/medical nature of the subject material. Is there a way around this dilemma so I do not get penalized? Thanks!
Intermediate & Advanced SEO | | Mills1 -
Subdomains - duplicate content - robots.txt
Our corporate site provides MLS data to users, with the end goal of generating leads. Each registered lead is assigned to an agent, essentially in a round robin fashion. However we also give each agent a domain of their choosing that points to our corporate website. The domain can be whatever they want, but upon loading it is immediately directed to a subdomain. For example, www.agentsmith.com would be redirected to agentsmith.corporatedomain.com. Finally, any leads generated from agentsmith.easystreetrealty-indy.com are always assigned to Agent Smith instead of the agent pool (by parsing the current host name). In order to avoid being penalized for duplicate content, any page that is viewed on one of the agent subdomains always has a canonical link pointing to the corporate host name (www.corporatedomain.com). The only content difference between our corporate site and an agent subdomain is the phone number and contact email address where applicable. Two questions: Can/should we use robots.txt or robot meta tags to tell crawlers to ignore these subdomains, but obviously not the corporate domain? If question 1 is yes, would it be better for SEO to do that, or leave it how it is?
Intermediate & Advanced SEO | | EasyStreet0