Href lang and multilingual question
-
Greetings Moz-Hive mind! I'm hoping you can help me on the internationalisation conundrum below;
We currently have a website with three distinct 'locales' US, SEA and UK we automatically redirect customers using IP recognition to a locale which matches, we also determine their currency based on IP. The issue we currently have is a lot of duplicate content and no use of href lang or rel=canonical tags etc...
My proposed structure would be to create a locale based directory for the three locales we offer.
- / - being US and most other Worldwide
- /uk - being UK
- /as - being Hong Kong and other Asian territories.
How would you suggest we set up the href lang tags for these? Because technically there are going to be multiple language possibilities within. Our main customers are English only if this helps.
Also as a secondary question, how should I set up the Google Search Console settings for each of these directories?
Many thanks in advance.
-
Thank you so very much!
-
Correct.
About using the hreflang="x-default" or the hreflang="en", that depends if you are interested in targeting also users not searching in English.
If not, you can go for the hreflang="en" option.
-
OK that makes sense. So let me get this straight;
1. USA/Worldwide version - /
- hreflang="en" (Can I use "x-default" here instead?)
- no targeting within Search Console
2. SEA version - /as/
- hreflang="en-HK" (repeat for all SEA territories)
- no targeting within Search Console
3. UK version - /uk/
- hreflang="en-GB"
- targeting United Kingdom on Search Console- use a selector pop up using IP sniffing to detect if the customer is viewing the wrong version of the page/site.
-
A solution could be implementing an alert system (Amazon style), so that people with those characteristics visiting the website and having eventually landed on it (but on / and /uk/ URLs) albeit they are located in HK or other EAS country, they are alerted that the best version of the site for them is the /as/ one.
You can do it via IP detection. Don't do automatic redirection, though.
-
Just to let you know, none of the above has been implemented yet, I am still planning the best course of action.
The majority of our clients browsing from Hong Kong have their language set to "en-US" and "en-GB". Is there a way to target these users without showing them the USA version of the site?
The USA version of the site needs to be the global one because the majority of our clients are based there.
Would you be willing to take this conversation offline into email? I would very much appreciate the help.
-
Mmm... question: do those users visiting the site from Hong Kong or from USA and/or GB?
Because if that is the situation, then you should check out if you have:
- Well implemented the hreflang;
- If you have geo-targeted the /uk/ subfolder to Great Britain/United Kingdom in Search Console.
Said that, following the logic I presented in my first answer, the hreflang="en" should suggest only "/" "(USA) URLs, because you told me that you want the USA version of your site to be also the "Global" one.
If it so, then the "/as/" cannot be paired to the hreflang="en" annotation, because you will have two identical hreflang (hreflang="en") suggesting two different URLs, and that would be wrong and confusing Google so much that it can ignore all the hreflang you may have implemented.
However, in the "home page" of the /as/ subfolder, you should have these hreflang annotations:
- <rel="alternate" hreflang="en-HK" href="http://www.domain.com/as/"></rel="alternate">
- <rel="alternate" hreflang="en-AU" href="http://www.domain.com/as/"></rel="alternate">
- REPEAT IT SO TO COVER ALL EAS COUNTRIES
- <rel="alternate" hreflang="en-US" href="http://www.domain.com/"></rel="alternate">
- <rel="alternate" hreflang="en-GB" href="http://www.domain.com/uk/"></rel="alternate">
- <rel="alternate" hreflang="en" href="http://www.domain.com/"></rel="alternate">
-
Thanks for your detailed response Gianluca! Very helpful.
One question, am I also able to include hreflang="en" under the /as/ subfolder? Looking at Analytics and 80% of our Hong Kong clients are browsing in either en-US or en-GB (complicated).
-
Your situation is a tricky one, because the SEA area is not a country, and therefore cannot be geo-targeted via Search Console.
To use the hreflang in order to suggest Google to show the /as/ subfolder is possible, though. Simply, you have to implement as many hreflang="en-[Code of the Country] as many are the countries included in the SEA area.
Then, in order to have the USA version being the global one:
-
You should not geo-target it toward USA in Search Console;
-
You should use the hreflang="en" and not the hreflang="en-US" (that will target only users from USA).
-
As an option, you can
-
set up the hreflang="en-US" to suggest google to show / to people using English in the USA;
-
set up the hreflang="x-default" to suggest google to show / to everybody (using English or not), who are not geo-targeted by other hreflang annotations.
Obviously, for UK, you must use the hreflang="en-GB".
-
-
Hey Dmitrii, sorry I should have mentioned in the question I understand how href lang works. I am more asking for the best practice implementation when the language might possibly vary? With US and Worldwide and the Hong Kong and Asian territories.
-
Hi there.
Well, you setup hreflang very normally - on each of those language based pages specify rel="alternate" hreflang="whatever" href="matching url".Refer to this video from Google, they explain pretty good what's what and when to use which version: https://support.google.com/webmasters/answer/189077?hl=en
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
-
Question regarding Site and URL structure + Faceted Navigation (Endeca)
We are currently implementing the SEO module for Endeca faceted navigation. Our development team has proposed URLs to be structured in this way: Main category example: https://www.pens.com/c/pens-and-writing/ As soon as a facet is selected, for example "blue ink" - The URL path would change to https://www.pens.com/m/pens-and-writing/blue-ink/_/Nvalue (the "N" value is a unique identifier generated by Endeca that determines what products from the catalog are served as a match for the selected facet and is the same every time that facet is selected, it is not unique per user). My gut instinct says that this change from "/c/" to "/m/" might be very problematic in terms of search engines understanding that /m/pens-and-writing/blue-ink/ as part of the /c/pens-and-writing/ category. Wouldn't this also potentially pose a problem for the flow of internal link equity? Has anyone ever seen a successful implementation using this methodology?
Intermediate & Advanced SEO | | danatanseo0 -
To subdomain or to subfolder, that is the question.
Hi All, So I have a client that has two restaurants that they are wanting two sites for. Right now they have one site for their two locations that ranks pretty well for some bigger keywords for their style of food. With them wanting two sites, i'm struggling on whether we should just build them all within one site and just use separate folders on that site restaurant.com/location1 & restaurant.com/location2 with a landing page sending you to each, or if we should split it into subdomains. The content will be roughly the same, the menus are identical, i think each branch is just owned by a different family member so they want their own site. I keep leaning towards building it all into one site but i'm not sure. Any ideas?
Intermediate & Advanced SEO | | insitemoz10 -
Canonical questions
Hi, We are working on a site that sells lots of variations of a certain type of product. (Car accessories) So lets say there are 5 products but each product will need a page for each car model so we will potentially have a lot of variations/pages. As there are a lot of car models, these pages will have pretty much the same content, apart from the heading and model details. So the structure will be something like this; Product 1 (landing page) Audi (model selection page)
Intermediate & Advanced SEO | | davidmaxwell
---Audi A1 (Model detail page)
---Audi A2 (Model detail page)
---Audi A3 (Model detail page) BMW (model selection page)
---BMW 1 Series (Model detail page)
---BMW 3 Series (Model detail page) Product 2 (landing page) Audi (model selection page)
---Audi A1 (Model detail page)
---Audi A2 (Model detail page)
---Audi A3 (Model detail page) BMW (model selection page)
etc
etc The structure is like this as we will be targeting each landing page for AdWords campaigns. As all of these pages could look very similar to search engines, will simply setting up each with a canonical be enough? Is there anything else we should do to ensure Google doesn't penalise for duplicate page content? Any thoughts or suggestions most welcome.
Thanks!0 -
Content Cannibalism Question with example
Hi, Since I love writing and I write a lot I always find myself worried about ruining for my self with Content Cannibalism. Yesterday, while looking to learn about diamonds I encountered a highly ranked website that has two pages ranking high on the first page simultaneously (4th and 5th) - I never noticed it before with Google. The term I googled was "vvs diamonds" and the two pages were: http://bit.ly/1N51HpQ and http://bit.ly/1JefWYS Two questions: 1. Does that happen often with Google (presenting two lines from the same site on first page)? 2. Would it be better practice for the writer to combine them? - creating a one more powerful page... Thanks
Intermediate & Advanced SEO | | BeytzNet1 -
Migration Challenge Question
I work for a company that recently acquired another company and we are in the process of merging the brands. Right now we have two website, lets call them: www.parentcompanyalpha.com www.acquiredcompanyalpha.com We are working with a web development company who is designing our brand new site, which will launch at the end of September, we can call that www.parentacquired.com. Normally it would be simple enough to just 301 redirect all content from www.parentcompanyalpha.com and www.acquiredcompanyalpha.com to the mapped migrated content on www.parentacquired.com. But that would be too simple. The reality is that only 30% of www.acquiredcompanyalpha.com will be migrating over, as part of that acquired business is remaining independent of the merged brands, and might be sold off. So someone over there mirrored the www.acquiredcompanyalpha.com site and created an exact duplicate of www.acquiredcompanybravo.com. So now we have duplicate content for that site out there (I was unaware they were doing this now, we thought they were waiting until our new site was launched). Eventually we will want some of the content from acquiredcompanyalpha.com to redirect to acquiredcompanybravo.com and the remainder to parentacquired.com. What is the best interim solution to maintain as much of the domain values as possible? The new site won't launch until end of September, and it could fall into October. I have two sites that are mirrors of each other, one with a domain value of 67 and the new one a lowly 17. I am concerned about the duplicate site dragging down that 67 score. I can ask them to use rel=canonical tags temporarily if both sites are going to remain until Sept/Oct timeframe, but which way should they go? I am inclined to think the best result would be to have acquiredcompanybravo.com rel=canonical back to acquiredcompanyalpha.com for now, and when the new site launches, remove those and redirect as appropriate. But will that have long term negative impact on acquiredcomapnybravo.com? Sorry, if this is convoluted, it is a little crazy with people in different companies doing different things that are not coordinated.
Intermediate & Advanced SEO | | Kenn_Gold0 -
To merge or not to merge? That is the question.
I am planning to do something I never did, and I am wondering if it's really a good idea or not. I have four websites, all of the same company, each one with a different domain and different content: one has been the main official site for 16 years, 200 unique per month, indexed for 134 keywords, Domain Authority 17, 13 linking root domains one has been used as the main site from 2003 to 2006, it's focused on a specific business they actually discontinued, still online, no update since 2006, 500 unique per month, indexed for 92 keywords, Domain Authority 13, 8 linking root domains another has been a built on 2010 and maintained for less than year, and it's focused on a business they never really started, still online, no update since 2010, 3000 unique per month, indexed for 557 keywords, Domain Authority 25, 84 linking root domains a fourth one has been also built on 2010 and focused on a business never really started, still online, no update since 2010, 100 unique per month, indexed for 4 keywords, Domain Authority 6, 3 linking root domains Each website has traffic and links, all links being natural, they never tried to gain links in any way, they never did on page optimization, they never ever thought about SEO. They are not event interlinked. So, my idea is to merge all of them, putting websites 2, 3 and 4 as subfolders of the main site and replicating the old content there. Because those sites have traffic, incredibly one of the abandoned sites has 3000 unique per month, while the main site just 200! My doubts are: does it make sense to merge everything from a SEO prospective? A part from doing 301 correctly, what else should I be careful to do or not to do? website number 4 it's really outdated, content and structure is not easy to merge with the rest, traffic is really small, is it worth spending the time to merge it? Finally I also have a problem; customer didn't want to merge them, they agreed to, but they don't want visitors of the main site to be able to navigate to the old ones, so once moved and redirected I would have to put them in the sitemap of the main site but avoid linking to them on the actual "main" site. As far as I know google crawler doesn't like to find pages in sitemaps which are not reachable through a linking path on the website, is that correct? Is that going to make all the merging work useless? Should I convince the client to at least put small links in the footer or on a page linked from the footer?
Intermediate & Advanced SEO | | max.favilli0 -
Anyone have an hour right now to cover some SEO questions
Hi folks, I need someone to Skype with me today, on some seo questions, for a multi wordpress set up I`m in middle of developing for franchise local sites. Will pay you $95 for the hour. Thanks Brent Sky pe me: cyberbrent (Brent H, Richmond, BC)
Intermediate & Advanced SEO | | MenInKilts1 -
Canonical Tag - Question
Hey, I will give a thumbs up and best answer to whoever answers my question correctly. The Canonical Tag is supposed to solve Duplication which is fine. My questions are: Does the Canonical Tag make the PR / Link Juice flow differently? If I have john.long.com/home and john.long.com but put a Canonical Tag on john.long.com/home reading john.long.com then what does this do? Does it flow the Link Equity back to john.long.com? Can you use the Canonical Tag to change PR flow in any means? If I had john.long.com/washing-machines and john.long.com/kids-toys... If I put a Canonical Tag on john.long.com/kids-toys reading john.long.com/washing-machines then would the PR from /kids-toys flow to /washing-machines or would Google just ignore this? (The pages are completely different in this example and content is completely different). Thank you.
Intermediate & Advanced SEO | | AdiRste0