Welcome to WebmasterWorld Guest from 188.8.131.52
Forum Moderators: open
Long time listener, first time caller, as they say.
I run a .com site that's been around since 2003 and is doing quite well in the US SERPs, particularly on Google. It gets crawled all the time and generally seems on Google's good side at the moment.
Now I'd like to translate the site into a few other languages - probably German, Swedish and perhaps Spanish to begin with - and am trying to determine the best way to proceed. I've read through many of the helpful posts in this forum and can't quite figure the right course of action, so I thought I'd post directly.
There seem to be 3 general options for how to proceed from a domain standpoint:
#1 & 2 have the clear benefit of harnessing the search engine mojo I've built over the last 3 years; Google and other engines may start to crawl and index these Swedish-language pages right away. Starting all over with a .se domain seems really daunting. And it seems that .com's do show up in international, non-English language searches.
What do you guys think? Should the decision be different based on the market? Might I lose some traffic by going the swedish.widget.com approach, but could it still be worth it due to the time, effort and money saved by not needing to start from scratch?
Any and all thoughts welcome. Thanks so much.
1) Google would treat this as a brand new domain and I can't see any advantages, unless you can host this sub-domain on another local server (in Sweden in this case), to really target the local market.
2) Has the advantage of being spidered immediately and will "piggy-back" on your existing site to immediately rank. Will not appear in local searches (see 3)
3) This is the best option if you really want to target the local market. Yes, .com's appear in searches in other languages, but not if they are not locally hosted and the user selects the "pages from Sweden/Spain/Germany etc." option that all local Googles allow. To be included in the SERPs for this option, you need a local ccTld or to be hosted in that country (see 1). Also applies to Yahoo & MSN.
Depending on your business however I'd translate the site in German, French, Spanish, Swedish and Italian too.
Be careful not to create duplicate content: do different copy on 2 and 3. This will not only benefit the popularity and "authority" of 2 but also help getting 3 be recognized as a domain of its own right.
I did 2 for many years and now start to spawn off lot's of 3's :-)
Having said that, I should mention that we recently put a 475-page dot-com travel planning site about Stockholm online in English. Amazingly, the majority of the business transacted is actually from Sweden, suggesting that neither English nor dot-se is really necessary. Furthermore, all of our dot-se sites are hosted in the UK and that hasn't hurt rankings at all. We are in the top 5 for all important keywords in a very competitive environment.
If you want to go for European business, you should start by translating for markets where people don't understand Engish as well as they do in Scandinavia: France, Spain and Italy, adding German for good measure - they will appreciate it.
Dot-coms work well everywhere. A nice touch is to translate widget.com into the respective language. If you check my profile site, you will see an example of how it can be done for 20 languages. Good luck!
[edited by: ThomasB at 5:38 pm (utc) on Mar. 10, 2006]
[edit reason] examplified [/edit]
Also make sure you consider locali(s)zation issues such as use of graphics, colours and other cultural issues.
I'm from the localization industry so if you need any stats or further advice feel free to stciky me.
.se/.fr/.de/etc is the only really good solution.
If you were to go with a subdomain/subfolder, use the ISO country code:
Do a search for 'Microsoft Office Online' and you will probably see the first result being your local language version of office.microsoft.com. From here I see office.microsoft.com/en-gb/default.aspx
The relevant contents of the <head> section are:
<meta name="MS.LOCALE" content="en-GB">
I have verified this for a few countries and it works well.
All versions I have tested have a UTF-8 header.
I might be wrong, but I don't know of any technology which allows a subdirectory to have a different IP (even if it has a different server) to the main domain. I did try running it through nslookup but that was unable to process something with a subdirectory.
From that I believe that having all the content grouped in subdirectories but in different languages is establishing that these are equivalent results, and correctly targetting them to the appropriate searcher.
Whether MS.LOCALE meta-tags are of value is an interesting question.
Thanks so much for all of the helpful replies. This has been great.
We are going to go with the widget.com/se approach. WebWalla, I agree that we will lose out on local searches, but even if that's 20% of the market I still think we'll be better off piggybacking on our existing domain vs starting from scratch with a .se domain. With the sandbox, it could be 12 months before we see any traffic there, so I'd definitely trade the 20% reduction for some immediate results.
Thanks also for all of your comments regarding which languages to pursue. We've decided to do our first experiment with a German version, so we'll use widget.com/de.
Paco, interesting question. I'm not sure on that.
Thanks again everyone. This has been really helpful.
>We are going to go with the widget.com/se approach.
Ok, but I wouldn't.
>I still think we'll be better off piggybacking on our existing domain
Short term yes, long term no. The engines will look more and more at local flavour to determine if a site belongs in a local language search. The more you can localise, the more you should.
>even if that's 20% of the market
You kidding, right?
Few reasons why:
- sandbox. Is it really that big a deal? You'd be surprised..
- 20% of your business could be a lot of money long term. A lot.
- ccTLD's builds trust. I'll choose the guy around the corner every time if his stuff is good enough.
>it could be 12 months before we see any traffic there
Really? Maybe a little ppc to test the waters?
>so I'd definitely trade the 20% reduction for some immediate results
I wouldn't - depending on your business of course.
>German version, so we'll use widget.com/de
Understandable, but remember in Germany you're swimming with the sharks, eg. you'll see the highest ammount of competition from the fact that it's the largest internet population after the UK.
I'm not saying you wont succeed with a com/country approach, Renke (Hi Jan!) is the proof that it cetainly can be done. However he's a seasoned pro and the founder of the euro forum, so he has a minor advantage imo :)
Registering and hosting locally can be a pain and you have to go with the do-able solution that's working for you.
Let us know how it goes! :)
Would redirecting from a local domain to a .com based in another country be considered as local by search engines?
I can give you a specific example I know of. It is a local widget.es Spanish domain which is redirected to widget.com/es with a 301 redirect. Widget.com isn't hosted in Spain. When searching for widget in google.es with the "Pages from Spain" option, widget.es doesn't appear at all. Searching for widget.es does give the widget.com/es URL, but this very specific search is the only time anything to do with this domain appears in the local SERPs.