Translation Guy Blog
Translations are like a hot stove to a savvy international manager. With any luck, they only get burnt the first time. The best solution for many is to sample. Get those potential vendors to prove their mettle by writing up some good translation on demand.
But that’s like going from the frying pan to pin the tail on the donkey. Because a sample translation does not necessarily reflect a sample translation workflow. It’s the quality control process that makes for good translation.
Quality control starts immediately, even for a paragraph or two. Especially if it’s the first time the translation team has worked on the product or service. At our shop, once you send that sample, that’s when we start asking questions. We’re asking about translation memories, glossaries, style guide, not only in the target language but in source languages as well. These are the bricks and mortar used to construct consistently well-styled translation across a product offering. That means that if your potential provider is not asking you those questions, then they are not using those best practices to translate your content. That doesn’t mean it will be wrong, but it does mean that it is less likely to be right.
If the client doesn’t have them, then the first order of business for the translation service is to construct those assets. If no glossaries exist, we will look to previous translation or industry standards combined with our best judgment to come up with the right terms. But the only way we can read the minds of our reviewers and get things exactly right is if we get their feedback. That can mean constructing the glossary beforehand, or doing the translation and getting terminology corrected by the in-country or client-side reviewer. So that means that the people evaluating our sample have to give us the support we need to do the job correctly. Even during the sample process, we work diligently with translation checkers to specifically understand and address the problems they’re having with our sample. This method is much closer to translation reality than luck of the draw by reviewers who were not properly supported in their own work. It’s as true for translation as anything else; you can’t manage if you can’t measure. Translation quality assurance does not have to be a subjective process, nor should it be.
So even if there’s some back-and-forth between reviewer and translation team on a sample, we are able to establish a reproducible method to get things right going forward. Otherwise, each time it’s like starting a client relationship from scratch. Normally I hate to post best practice doctrine on the web, since everybody comes and copies it. But I think evidence of this best practice extends far beyond boilerplate. The devil is in the details and a project management team that isn’t wrangling these issues from the get-go is going to get left in the quality assurance dust. It’s also important to find a vendor who will work with you to set up the necessary steps with minimum client-side headache. The long-term goal is to make the process as seamless as possible. That’s not always possible to accomplish on a shakedown sample of a paragraph or two.
The time required to prepare and review samples pays off in spades with reduced headaches, embarrassment and costly error further down the line. Vendors who ignore this process put their client’s success at risk.