Should I use greek characters in my urls?

Question

I'm setting up a website for a client who's Greece based and I'm not quite sure how to go about the urls.

There are two options I'm considering at the moment:

1. Use greek characters and have the urls look like that: www.mydomain.gr/γυμναστικη

2. Use latin characters and have the urls look like that: www.mydomain.gr/gimnastiki

I could always use the english version of the word which in that case is 'fitness' but it won't be very helpful from a SEO point of view.

I would really appreciate any opinion in the issue.

Answer 1

Aside from technical difficulties mentioned in @mookamafoob's answer, there potentially is this to consider: Many users in countries that have non-ascii characters in their alphabets have learned that URL's can not be made up of anything but ascii characters. Using identifiers, even if technically possible, might cause users to wonder, if an address they enter really works, or if they should themselves somehow "attempt to convert" the identifier to ascii.

I don't have any official research to back this up, but can distinctly remember a very convinced client telling me that we can not use "äöå" in the folder structure of a website we were developing. The technical reality simply does not always match users' expectations.

Thus my advice would be to stay clear of localized URIs for the time being. At the very least, map ascii version of your URIs to the localized URIs, if you decide use them (i.e. www.foobar.com/aoa redirects to www.foobar.com/äöå, just to be sure that users trying to "guess" an ascii URI find the right page).

Answer 2

Oddly enough there are a couple similar questions on SO that address this particular issue. I would begin by look at https://stackoverflow.com/questions/2015546/allowing-non-english-ascii-characters-in-the-url-for-seo.

Then look at the linked questions here and here.

In essence, the URL won't have any effect on SEO. Most languages will change the characters to be Latin based as pages tend to and should be served as UTF-8. IE will also often have problems with non-Latin characters. So option number 2 would probably be most ideal.

Hope this helps!

Answer 3

I'd say it depends on your audience: if they are almost only visitors who will be able to type these chars on their keyboard, you're fine. If not, normalize.

amazon uses IRI's on their local pages (i.e. http://www.amazon.de/Haushaltsgeräte/b/ref=sa_menu_es8/276-7584535-8896931?node=3169211). So, if a global player can do it, why shouldn't we?

If my projects target an international audience, I always normalize filenames and links to ASCII, for local projects I use IRI's (but still normalize filenames of uploaded media).