A recent question posed at the Google Webmaster Central forum shows exactly how complex multi-language and multi-region sites can be to work with, in order to try and get the best localized results in Google. The poster has a somewhat unique situation, as he (she?) works with a Club finder Web site that has interests in multiple countries. In an example scenario, they want someone from Spain who is traveling in Germany to be able to find out about events at German clubs from a sub domain dedicated to Germany, but content written in Spanish.
Does using different subdomains (specifying in webmaster tools the country for each one of them) and folders for specifying languages is already ok ? We want google to index both, not treating them as duplicate, but knowing that each one, appart from being in the same language, are prepared for different countries.
Googler John Mu, who has provided sound advice this year to Webmasters seeking to translate content automatically and advising against using more than one language per page, has recommended that the poster consider using a canonical tag to help improve his rankings.
John comments that as long as he is using the proper geotargeting settings, the current use of similar content in multiple languages on different sub domains shouldn't be an issue. He feels that in some instances that the content should be tagged with canonical though.
The poster describes a scenario where a particular DJ might have a same page on every sub-domain. This likely would be close to duplicate content, and makes sense to canonicalize, especially if trying to rank for the DJs name. The drawback would be that the canonical version would have the best shot at ranking for the DJ, which leaves those sub domains/regions that are passing the value over unlikely to be able to rank as easily with the sub domain. Of course, the .com could rank in the specific ccTLD version of the search engine, but may have a handicap. And there is also the question of just how competitive the term is (is it a super well-known DJ or one of those guys with money that thinks he can spin?).
The rest of the content would be likely to be unique, since it describes events occurring in the specific country, although in different languages. I am curious to see how it would compete if someone was to do a search in their native language but in the search engine ccTLD specific to the country they are visiting. Does anyone have that experience?
Please share your thoughts and experience in the comments below or at the Google Webmaster Help Forums.