Here are the FAQs for International SEO which were available on Gooogle earlier this year.
Q: What’s the best URL structure for a multilingual site?
A: We detect the language on a per-URL basis; you don’t need to use any special structure. However, it’s important that each language version is on a separate URL, and that Googlebot is able to crawl all versions.
Q: What can I do if a user reaches the “wrong language” version of my pages?
A: We recommend letting users access all language versions (without automatic redirection), regardless of their location or browser settings. For example, it’s possible that a user in France may wish to see English-language content, and may even have explicitly searched in English to find that content. Also, keep in mind that Googlebot must be able to crawl all versions in order to index them properly and remember to make it easy for users to switch to their preferred language using links or a drop-down menu for the different language versions of a page.
Q: Google shows the “translate” link for my page in the search results, even though my page is already in the language used when searching. How can I prevent this?
A: We recommend making sure that the language of the page is obvious and clearly recognizable. For example, if the main content is in one language, it may help to have other text on that page (menus, sidebars, footers, etc.) in that same language. In some situations, Google is able to recognize multiple languages on a page (for example, if a page in English includes individual Spanish translations or names). Displaying a “translate” link does not necessarily imply any problems. To suppress translations all the time, use the “notranslate” robots meta tag. To block translation of individual elements on a page, use the “notranslate” HTML class.
Q: Should I use character encodings or HTML entities for non-ASCII characters?
A: Either using a specific character encoding (like UTF-8, ISO-8859-1, GB 2312, etc.) or using HTML entities is fine. We generally recommend using UTF-8, but you can use other encodings if they are clearly specified.
Q: Can I use automated translations?
A: Yes, but they must be blocked from indexing with the “noindex” robots meta tag. We consider automated translations to be auto-generated content, so allowing them to be indexed would be a violation of our Webmaster Guidelines.
Q: Do you support multilingual content in rich media files like Flash?
A: Yes. Since rich media content is frequently indexed as a single page, it helps to separate the content into separate files per language. Also, see our guidelines for using Flash and rich media.
Q: Can I use meta tags to specify the language of a page?
A: We don’t use meta tags or HTML attributes in determining the language of a piece of content. Keep in mind that other products and services, such as screen-readers, may use meta tags or HTML attributes, so it’s not pointless for you to consider using them.
Q: What’s the best URL structure for a multi-national (geotargeted) site?
A: It helps to have the country-specific sections of the site clearly separated on a per-country basis. You can do that either by using ccTLDs or by correctly configuring gTLDs in Search Console. With gTLDs you can use subdomains (country.example.com/) or subdirectories (example.com/country/); either way is fine. Google also allows geotargeting for some ccTLDs..
Q: Is the server location important for geotargeting?
A: If you can use one of the other means to set geotargeting (ccTLD or Search Console’ geotargeting tool), you don’t need to worry about the server’s location. We do, however, recommend making sure that your website is hosted in a way that will give your users fast access to it (which is often done by choosing to host near your users).
Q: Apart from ccTLDs, Search Console, and server location, what can play a role in geotargeting?
A: Anything that makes it clear what your pages are trying to target: for example, physical addresses, a local phone number, a Google+ Local listing, the use of the appropriate currency, etc.
Q: What can I do if a user reaches the “wrong country” version of my pages?
A: We recommend letting users access all country versions, regardless of their location. Avoid automatically redirecting users based on their perceived location. To guide users to the appropriate version of your site, you may wish to show the user a banner or otherwise encourage them to visit a local version of your site.
Q: I want to target the whole world; what should I do?
A: Geotargeting helps when users want to find locally-relevant content. If you want to target the whole world, then you generally don’t have to do anything special with regard to geotargeting.
Q: Can I use meta tags to specify geotargeting?
A: Google does not use meta tags when determining geotargeting.
Q: Can I target different English-speaking countries separately with the same content?
A: We recommend using unique and locally-relevant content whenever possible, but we understand that this is not always possible. The “rel-alternate-hreflang”
Rel Alternate Hreflang
See also our Help Center article on rel-alternate-hreflang.
Q: Can I use “rel alternate hreflang” across domains / across TLDs?
A: Yes! Keep in mind that this markup is on a per-URL basis, so if there are alternate URLs across those domains, you would need to add the markup for each individual URL.
Q: Does “rel alternate hreflang” replace geotargeting?
A: No. This link-element provides a connection between individual URLs, and only allows Google to “swap out” the URLs from your site currently shown in the search results with ones that are more relevant to the user. It does not affect ranking, as geotargeting would.
Q: Since this markup can be specified in an XML Sitemap file, should I use separate Sitemap files for each language and location?
A: The location of a URL within your Sitemap files does not matter. You can structure the Sitemap files in any way that makes sense for you, and that helps you to maintain them.
Q: I use “rel-alternate-hreflang” on my pages, why does it sometimes not work?
A: It can sometimes take a bit of time for this markup to be discovered, extracted, and processed on all of the URLs that are involved. Additionally, keep in mind that this markup is seen as one of the signals, not as a directive. See the previous questions for information on how to handle users who arrive on the “wrong” version of the page.
Q: Can I use “rel=canonical” together with “rel-alternate-hreflang”?
A: We recommend not using rel=canonical across different language or country versions. Using it in the same language/country version is fine and one of the recommended ways of handling canonicalization.
Q: How can I check the indexing of my multilingual or multiregional content?
A: A simple way to get a very rough check is to use the advanced search options on http://www.google.
1. Enter your site or domain in the appropriate field
2. Select the desired language or region
3. Click “Advanced search”
Keep in mind that the numbers shown here are very rough approximations and not meant to be used as actionable metrics. Remember that some URLs may be recognized as having content in multiple languages. For a more actionable metric, consider using separate Sitemaps files for each language or country and referring to the Sitemap indexing numbers shown in Search Console, or depending on the site structure, the data shown under Index Status in Search Console.
Q: Can I use non-English / non-ASCII URLs?
A: Yes. Past the domain name, which can be an IDN, we recommend using UTF-8 for path, filename, and query parameters. Google supports UTF-8 path, filename, and query parameters in the robots.txt file. Tip: Make it easy for users to link to your content by providing copyable links with escaped URLs so that these URLs remain valid regardless of where they are linked from.
Q: Can I block individual countries from accessing my content?
A: Yes, however, you must treat Googlebot the same as a normal user from that region. If you are blocking users in the country where Googlebot is crawling from, you must block Googlebot as well; treating Googlebot differently would be considered cloaking, a violation of our Webmaster Guidelines. If you need to block Googlebot but would still like to have some content indexed, it may be possible to create separate pages with content that’s globally accessible. It’s not possible to selectively block the cached page or Instant Preview on a per-country basis, but you can use the “nosnippet” and “noarchive” robots meta tags to block these features in general.
Q: Can I use interstitials in some countries?