The following was recently posted on a localization forum: “We sometimes get a letter from a user in a country whose language our software hasn’t been translated into. The user would offer their help in translating our product into their language promising to do a good job.
Even when a company uses a professional translation service that they’ve had a relationship with for years, there may be one or two languages that this service doesn’t handle very well and you start getting letters from users saying how there are so many errors in their language. The main problem I see with translations is that translators do not know the actual product for which they are translating, and it’s hard to translate words and phrases taken out of context. So theoretically, accepting services from a user, who actually knows your software very well and knows where most of the phrases are used in the program interface, wouldn’t be such a bad idea. I can think of a couple of reasons why it may still be a bad idea, but would be curious to hear other people’s opinions on this.
Do any software companies ever accept these offers? If you’ve done that, what was the result?”
In short, unless your company has a very strong localization staff and technology infrastructure to handle the localization process and QA it, complemented with many ready-to-assist pre-screened user volunteers, be aware of what you are about to do, for you are destined to run into numerous potentially horrible consequences.
These are some limitations that this scenario will ultimately impose:
1- Users are more knowledgeable of the product than translators inexperienced with your industry or product. This is why you can’t just hire any translator and instead have to look for translators that are subject-matter experts (SMEs). During the initial translation effort, it is best to involve both, professional SME translators and experienced users. The translators translate while in-country users perform testing and quality assurance.
2- Translating out of context is very difficult to do and error prone. Providing context around the source text is extremely important to achieve accurate translation. But this is not always possible to do. Training a translator on the product and giving them access to the author of the source text (developers and tech writers) and an in-country product expert, empowered by an online query portal that permits effective collaboration, is the best recipe for success.
3- Users don’t understand as well as translators the language requirements and more importantly the importance of a correct and efficient translation process. For instance, does this user know how to correctly translate resource files, or resize dialog boxes or apply translation memory technology to store all translations in a database for efficient and consistent reuse, or correctly and consistently use industry terminology, or apply correct style guides? Translating by mistake a source-code string that should not be translated, can result in many hours of debugging and recompiling before your software can be built, or worst yet, can cause operation crashes in the field requiring costly product recalls or updates.
4- Many users ad lib or modify the intent of the text. They alter the integrity of your source in the target language. They think they know better than your technical communicators and take it upon themselves to unilaterally correct a wrong. Can your company assume that liability? Why take the risk and possibly ruin your international image?!
5- When just one person is involved, an edit/review cycle will be missing, leading to inferior translation quality and makes possible deviations from the source. It is always helpful to have a second pair of eyes review the translation, source against target, and not just do a cursory proof of the target language. Waiting for a post-compilation linguistic quality step that is much further down the localization process makes applying corrections to the translations costly and time consuming.
6- Depending on just one person will limit capacity and constrain schedules. If you are seeking simultaneous releases, you do need a professional team supporting you. Time-to-market is essential, particularly when you have a short life cycle as software products typically do.
7- Software products undergo frequent last minute changes. If this person is not available when urgently needed, I’ve seen engineers resort to machine translation to convert new or additional strings, corrupting the integrity of the entire localized product! Engineers don’t realize how bad machine translation output is and when in a bind, turn a blind eye to the correct translation process and release what looks like a localized product to them. Will you blame them? Software release deadlines can be brutal!
I can give many more reasons why such a strategy can and will ultimately lead to disaster. Many clients have come to us with existing “translations” performed by their field personnel asking us to take over their updates. They did that usually after reaching capacity limitations or getting frequent complaints from their end-users. Unfortunately, correct localization had to be often done from scratch on the entire product causing havoc with the installed user base while they get re-oriented to using the corrected product.
An Arabic adage says to give your dough to the baker even if he eats half of it. Software localization is a team effort and there are many techniques to save on costs; accepting a volunteer’s offer to handle the bulk of the localization process and its numerous important tasks is decisively not one of them!
|Whitepaper – 10 Ways to Reduce Localization and Translation Costs|
This white paper will give you 10 tangible methods to reduce your localization costs without sacrificing quality. Guaranteed to save you money and a must read for every translation localization manager.
Download Whitepaper Now!