Tuesday, February 28, 2012

Localization Struggle: Quality Standards Vs. Cost Reductions


Some time ago, for the first time since our company was founded, a Japanese video game publisher requested us to use Trados 2007 for the localization of their brand new Playstation 3 video game. This request has come to us for the first time after more than 60 million characters translated -often into more than 6 or 7 languages-. 

Why is this possible? How come can the Japanese localized such voluminous video games without using a trendy TM such as Trados, Dejavu or Wordfast? My personal opinion is that Japan has traditionally been a software developer country with little or no idea of video game localization. All they had to do so far was wait for European companies to come to Japan, buy their licenses and localize the product, while the Japanese smiled and looked forward the royalties (usually a 15%). Changing the dollars for yen, was the only localization they used to do.

However, the fall of the sales in the Japanese market (together with the piracy), forced the Japanese developers and publishers to push to license out their products in Europe and America. Need less to say, the big ones still are lucky and have no trouble to sell abroad, but tiny developers have to be aggressive if they want to sell their product. And, also, they must be price competitive. The developer usually wants to save as much money as possible, and one of the usual phrases you can hear in that kind of negotiations is: "Oh, we will also localize the game to save some costs"...

And this is how the localization nightmare begins. Their cost saving strategy is basically the below:

Find the CHEAPEST translation agency.
Ask the translation agency not to use TM tools, as rates are higher when you use them.
Ask the translation agency to translate into Spanish, French, German or Italian from English rather than from Japanese, which is more expensive.

The consequences are usually as follows:
A. Not using TMs for a 500,000 character translation means inconsistencies will appear in the text (some video games have hundreds of items that are repeated through all the text).
B. Not translating directly from the source language -Japanese-, means that the 5% of the meaning is already lost in the first Jp-En translation. Moreover, if there was a single error in the first translation, that error will be repeated in the French, Spanish, English and Italian translation.
C. The game is usually rejected by the hardware maker (Nintendo, Sony or Microsoft), and consequently the developer has to spend more time and money testing the game in order to release it.
So, in this case also cheap is expensive. Even if the translation is only a little part of the localization process (being the familiarization, text debug and system debug the resting parts), we consider the translation as the HEART of the localization. Usually, I like to explain our clients that investing a 10% more in a proper translation, they will save more than the 50% of the total localization cost. Furthermore, they will also receive a translation memory of their game text, which will allow them to be more cost effective in future projects -as we all know, it is very common to have sequels in the video game translation world. However, most of them prefer us just to drop the prices and struggle to meet impossible deadlines. 

In Active Gaming Media, we believe it is crucial to focus on quality once more in this era of global crisis. Users not only look for the cheap, they also sharpen their targets and buy the best localized ones.  When this client asked us if we could use Trados 2007 for their localization, I felt happy to know we are not alone in this crusade for quality. I can guarantee, our clients's clients will also be happy with the game on their hands.