Facebook Taps Users To Create Translated Versions Of Site. Spanish, French and German Available Now
Michael Arrington
46 comments »
MySpace continues to roll out local versions of its social network. As I wrote last week, they tend to put a team on the ground locally (they are now hiring in Turkey) and then build the site not only in the local language, but promote local artists and other popular culture as well. MySpace now has offices in London, Paris, Berlin, Madrid, Milan, Stockholm, Helsinki, Oslo, Copenhagen, Sydney, Mexico City, Sao Palo, Buenos Aires, Toronto, Tokyo, and Beijing. Offices will be opening up soon in Mumbai, Moscow, and Istanbul.
Facebook is taking a radically different approach - tapping users to do all the hard work for them. They are picking and choosing markets (Spanish was opened first, two weeks ago; today German and French were launched) and asking just a few users to test out their collaborative translation tool. Once the tool is perfected and enough content has been translated, Facebook will offer users the ability to quickly switch the language on the site, per their preference.
Like Google has done with image search and Wikipedia has done all along, sometimes it just makes sense to ask you users to do as much work as possible. As long as users go along with it, everyone is happy.
The translation application can be found here, but for now only invited users can add and use it.
Once users add the application and enable “translation mode,” they’ll be given phrases from pages they browse that need to be translated. Each phrase is an easy, bite-sized chunk. Once submitted, other users vote on the submissions and/or submit their own version. More details are here.
More than half of Facebook users are now outside of the U.S. Given that all of them need to browse the site in English today, that is an amazing statistic. Today there are just 600,000 active Facebook users in Germany, Facebook VP Matt Cohler told me this morning (actually, he told that to a Business Week reporter, but I eavesdropped and generally made a nuisance of myself). When a German version of the site launches, look for that number to increase dramatically.
Wow… a 15 billion$ company can not afford a few translators?
Into Chinese presumably soon..
Nothing like using crowdsourcing to do what you should be paying for. Good on you facebook for finding yet another way to exploit your users.
If that’s the case (of exploiting users), what about the volunteer coders contributing to those open source projects (like, Mozilla’s, MySQL, Zimbra, etc?)
113 , you aren’t seriously comparing Mozilla and MySQL to Facebook. RIGHT?!
Crowdsourcing translations is a hot idea.
We made a web app that any site can use to do this too: Languify. Just released recently. Very open to feedback.
I think on the long term MySpace’s strategy will pay off more. It may be more costly, but creating local versions of a network is not only about translation, it’s also about understanding the local culture and adapting the content.
You should also know that Meebo (http://wiki.meebo.com/doku.php?id=translations) and Netvibes (http://translate.netvibes.com/) have done the same thing, and it’s worked great. I don’t see anything wrong with it.
@Jason - well, to be or be not.. to have or have not.. to care but for what..
Don’t forget that Google has been doing this for years:
http://www.google.com/transconsole
(Although they also professionally translate a lot of their stuff too.)
Facebook is also looking around in Europe, considering at the somewhat mysterious answer on a question at the DLD conference about acquisitions by Facebook in Europe: http://thenextweb.org/2008/01/.....in-europe/
Right.. and Google had explicitly stated unequivocally to all contributing translators that their legal rights to their work were limited if any..
The comparison should be with Wikipedia, not with Google’s image labelling. Translation is a very difficult, demanding job, requiring not only two huge joined-up vocabularies but a significant amount of creativity. The amount of creativity required increases the more different the two languages are, but even translating an English sentence into German, or vice versa, requires considerable skill with both languages. You have to know how to make it sound ‘right’ in English while retaining the original meaning. In that respect it’s closer to researching and writing articles for Wikipedia than looking at a picture and writing ‘cat’.
One of the things about contributing to Wikipedia in a non-casual way is that you do get recognition for it. Anyone who spends more than a few minutes doing valuable work will have it recognised, whether through compliments, name recognition, or ‘Barnstars’, pseudo-awards that anyone can give to anyone for anyone. Wikipedia critics have, with some justification, called this cultish, weird and valueless, and as an ex-admin I’m not going to bother defending it.
My point is that working for Wikipedia is not without reward. There may be no monetary payment, but there is recognition. That recognition is there because there is a community of editors who read, check and edit each other’s work, and it’s helped Wikipedia to grow as big as it has. What such community will recognise the work of volunteer translators for Facebook? In short, why would I spend time translating Facebook into (say) German? Some guy I don’t know saying “hey, good job” isn’t enough, to feel that your work is valued you have to feel like part of the community that values it.
And that isn’t even the most obvious point, which is that Wikipedia is a non-profit, while Facebook is a for-profit company. There’s something intrinsically wrong about asking people to make you money for nothing.
I don’t think this is really “radically different” by any stretch. WordPress.com has been doing the same thing for ages:
http://translate.wordpress.com/
Yet didn’t get any press.
Re-reading the post, I’ve clearly neglected that the phrases will be split into bite-sized chunks, so a lot of what I’ve said above is pretty much invalid as a comparison.
Now that I know what’s going on, I think Facebook is taking the wrong approach to translation - translation shouldn’t be done in bits, there needs to be a consistent approach. There’s a reason translated books generally have a single translator on the cover and not the name of a company employing 20 temporary staff who all translated a chapter each. Even if this is just menus we’re talking about translating, you could easily end up with the same word in the translation being used for two entirely different things on the site. However, I’m not going to develop this further because I have no business talking if my reading comprehension isn’t turned on
Meebo has been doing this for awhile, and I think it’s a great way to extend the reach of your app faster and more effectively at a low, low cost.
I have to agree with Tim (not that I agree on the 15B valuation, but that’s another story :-).
Cheap..ahem..labor is one thing. Time to marker and strategy another. What they may save in localization/translation they may have to spend on acquiring local social networks at a much higher premium price tag for not having entered those markets soon enough. Then again, they might already have to to that regardless of who/how long it takes them to localize their site.
I suppose this is cheaper than buying studi VZ.
The issue Facebook needs to deal with is localization and not just translation. Localization is a completely different cup of tea and I have my share of doubts that Facebook has the necessary know-how.
How about Malay/indonesia language?
Definitively that is going to be of much utility. To cover the market and to obtain the greater amount of users is the goal of Facebook and that better way to do, that implementing new tools of language…
Apples to oranges. FB makes a smart crowd-sourcing move on translation. MySpace makes a smarter move by getting into the society (that’s the same root as “social”, right?…duh). A Spanish language version of FB is bound to alienate most Spanish speakers, becuause not all Spanish is the same - every Spanish society has it’s own variations on the language. So if users vote on acceptable definitions, it’ll be like the Brits being forced to read “pants” when the meaning is really “slacks”. To the Brits, “pants” = “underpants”. Ergo, local is not just a translation exercise, crowd-sourced or otherwise.
Does anyone have an opinion on whether the actual facebook.com domain will be translated into other languages to represent different languages? For instance, would livredevisage.com be used for the French facebook?
How about translating facebook to ascii in decimal - 102097099101098111111107
henceforth, 102097099101098111111107.com :p
Ben, for Germany http://www.facebook.de is already prepared.
Some may call this exploiting users, but it’s not. It’s actually to the benefit of the users, and they get to be part of the creation.
p.s. Flickr started on this same ‘exploitation’ model before it was turned into a full blown photo community. It was a game involving tagging the random images that showed up. free labour!
Fantastic strategy by FB.
You can also preview the translated version and translate while you surf.
And you can vote other people translations in order to make fb know if those are OK
That is very cool. I think is harnessing the power of the user and allowing them to have a voice in development.
This is a lazy idea on their part, but sounds like fun for users. This could be another way for users to impress friends.
http://www.webepags.com
Contrary to popular (and Michael’s) belief, translating a site doesn’t do much at all for its popularity. Foreign early adopters tend to not be scared away by English-language sites, and once a base is built, users find ways around not understanding precisely what’s going on. Prime examples: orkut in Brazil (wasn’t available in Portuguese until much, much later) completely taking over social networking in Brazil (if not the country itself), Hi5 in Latin America in general (biggest player down there although only in English for a long time), and a lot of other examples. Language doesn’t matter at all in a globalized (online) world; localization, local user base and local momentum do. So Michael’s wrong if he thinks that this will do anything to FB’s traction with users in Germany.
I blogged about this on December 27th.
http://facereviews.com/2007/12.....wdsourcing
Just to pick up on the point in the last paragraph - the US isn’t the only English-speaking country. Therefore, it would be more interesting to know how many users are using Facebook outside English-speaking countries, than purely the US.
With the amount of money facebook has they should pay professional translators to do the job and make sure that the translations are quality instead of relying on someone who may do a poor job.
Remember the Milk has used this successfully for a long time…
http://www.rememberthemilk.com/services/translate/
click More Languages to see all 19 they have completed.
I think FB’s approach is better, more scalable (and cheaper) than MS.
I wrote about this a while back and I think the timing off this is bad. Only a few weeks after pissing off pretty much everyone and everything they have decided to ask their users to help them do THEIR work for them, for FREE!
I for one will not be helping them! Also, I believe their reach outside the English speaking world is rather small, so it will be interesting to see if they can find enough people to pull this off.
http://www.taxidrivermarketing.com/?p=19
I think they don’t have much chance in Poland, even if they translate everything into Polish. Most of internet users in Poland are using http://www.nasza-klasa.pl (something like classmates.com) and their growth is amazing in last few months. They have over 7 mln users already.
Little bit of history: started by 4 students, investment around 1 mln$ from Samwer Brothers. Now they overtook almost all largest Polish portals.
http://www.alexa.com/data/deta.....ize=Medium