Welcome to the Q&A Forum

Hi Gianluca,

Thanks for responding. I took a look at your guide, and I definitely understand the gold standard would be to get everything translated professionally, and provide a completely native experience.

Unfortunately due to our catalog size that would be prohibitively expensive, so I need to think of another solution. It sounds like from your guide that we are doing more harm than good, even with alternate language syntax in place.

Based on your answer, my thought would be to meta noindex any product page where the site language is not the same as the product. That way every page in the index will be 100% localized for potential visitors.

So if its a Spanish product index: site.com/es/product, but meta noindex site.com/de/product, site.com/product, etc.

If we follow that path, does it make sense to remove the alternate language syntax, since all the linked URLs will be no index?

Thanks again for your help.

Hi Mozzers,

We have a marketplace with 20k+ products, most of which are written in English. At the same time we support several different languages. This changes the chrome of the site (nav, footer, help text, buttons, everything we control) but leaves all the products in their original language.

This resulted in all kinds of duplicate content (pages, titles, descriptions) being detected by SEOMoz and GWT. After doing some research we implemented the on page  rel="alternate" hreflang="x", seeing as our situation almost perfectly matched the first use case listed by Google on this page http://support.google.com/webmasters/bin/answer.py?hl=en&answer=189077.

This ended up not helping at all. Google still reports duplicate titles and descriptions for thousands of products, months after setting this up. We are thinking about changing to the sitemap implementation rel="alternate" hreflang="X", but are not sure if this will work either. Other options we have considered include noindex or blocks with robots.txt when the product language is not the same as the site language. That way the feature is still open to users while removing the duplicate pages for Google.

So I'm asking for input on best practice for getting Google to correctly recognize one product, with 6 different language views of that same product. Can anyone help?

Examples:

(Site in English, Product in English) http://website.com/products/product-72

(Site in Spanish, Product in English) http://website.com/es/products/product-72

(Site in German, Product in English) http://website.com/de/products/product-72

etc...

Rogerbot just completed its first crawl and is reporting all kinds of duplicate content - both page content and meta title/description.

The pages it is calling duplicate are used with  rel="alternate" hreflang="x", but are still being labeled as dupes.

The title and descriptions are usually exactly the same, so I am working on getting at least those translated into different languages.

I think its getting tripped up because the product page its crawling are only in English, but the chrome of the site is in the translated languages. The URLs look like so:

Original: site.com/product

Detected duplicates: site.com/fr/product, site.com/de/product, site.com/zh-hans/product