It’s day two of Translation Review Week, where we plan to devote the next few days to talking about translation review and its documented correlation with migraines and road rage in the localization industry. Get ready to talk geek, folks, ‘cause today’s entry is a good one.
Now, for those of you who just stopped by, the first entry in Translation Review Week dealt with knowing what you want. That is, it’s important to understand the types of documents you’re dealing with before translation and—more importantly—the purpose behind translating them. This information should in turn be used to inform your own requirements for translation review and overall linguistic QA. Based on what you understand of your own business and its translation objectives, you then need to interface with your language service provider and make sure that their review practices can assimilate your needs.
So yesterday, the emphasis was on identifying what you want. Today we’re going to talk about getting what you want by defining it for others. To wit, we’re going to talk about the kind of guidelines you should consider when defining the scope of, preparing for, and implementing your translation review process.
In keeping with our “tell your LSP to shut up and listen” philosophy (in business-ese, that’s “fostering a listening culture with your localization vendor”—but a rose is a rose is a rose is a rose), the following suggestions are best practices that we’ve gleaned from our experience as a leading language service provider, and by no means should they be taken as truisms that trump your internal processes and needs. They can, however, be used to complement the translation review requirements that you’ve already identified for yourself, regardless of whether you employ in-house, independent, or vendor-driven methods of review.
Above all, show the review process a little respect.
It surprises us whenever a client mentions how localization isn’t recognized nor really respected throughout their organization. Given that positive localization ROI is a well-documented actuality, you’d think that more emphasis, time, and resources would be applied to, well, making localicious products. While they appreciate the extra revenue (one of our clients in the Life Sciences actually makes 60% of their total revenue from foreign markets—sixty percent!), there’s still a mind-boggling tendency among translation buyers to tack localization on at the very end of the product development cycle. Almost like an afterthought.
What does this have to do with the translation review process? Well, I know none of you have taken the SATs in a while, but check this out:
localization: product development cycle :: translation review : localization cycle
Now, analogies with colons usually make my brain hurt, but this one’s pretty simple: Localization is to the Product Development Cycle as Translation Review is to the Localization Cycle. The quality of localization defines the success of your product abroad, yet it’s tacked on to the end of the product development cycle.
Similarly, the quality of translation review in a way defines the success of your localization cycle, and yet the review process is tacked on at the end too. Translation reviews are often treated like that weird guy in the office whom you didn’t want to invite to the movies, but whom you had to invite because he overheard you inviting someone else. Seeing as it’s sometimes the last quality barrier between product development and international deployment, translation review deserves a little respect.
(It’s a little known fact that Aretha Franklin’s 1967 hit, “Respect,” was originally written about the translation review process. Digital props to Ryan Arrowsmith for the picture.)
How to show a little respect:
For one, you should plan better for it. Preparation for translation review should begin before the first translator gets a frantic e-mail from your LSP’s project management team.
Get reviewers on board from the start.
As a best practice, we suggest that you first define who will perform linguistic review on your documents (“the who” in translation review will be discussed on Thursday) and then arrange for your LSP to give the project guidelines to their translators and reviewers at the same time. Get them in the same (digital equivalent of a) room; give them glossaries, style guides, and explain the scope of the project at the same time; give translators and reviewers a chance to build rapport and understand the project together in order to align them with a common goal.
This gets everyone on the same page from day one, and keeps translators and reviewers from biting out each others’ jugulars.
Give reviewers ample time.
As a secondary best practice, not only should reviewers be brought on board before project kick-off, but when it comes time to perform reviews (whether they’re pre-translation terminology reviews, mid-translation sample reviews, post-translation but pre-production reviews, final proofreading or translation approval, etc.), reviewers should be given enough time to do just that: sit back with their red pens, don their monocles and review.
Just as localization shouldn’t be tacked on as a last-minute step in the production cycle, the review cycle should also be threaded throughout—and clearly scheduled into—the localization process. We understand that time is of the essence and that, the faster your product hits the market, the more of a competitive advantage you’ll have. But there’s always, always a way—and that’s when good time management comes in. Demand it of your language service provider.
This makes sure that reviews are performed well and, incidentally, that they’ll be delivered on time.
Give your reviewers specific instructions on what should be reviewed.
When a project is handed off to a translator, he or she will invariably receive a clear set of written instructions that outline the customer’s expectations for the project. Likewise, reviewers should also receive a clear set of parameters (including checklists) regarding what’s expected of them. Before translation begins, sit down with your LSP and define what constitutes an error, then weight each error accordingly. The types of errors reviewers should look for depends on the type of document—which is why it’s important to know what you want.
There are a lot of metrics out there for defining what constitutes a translation error. One you’ll often hear about is the SAE J2450, which is a metric developed by the Society of Automotive Engineers for objectively and consistently measuring the quality of translated material. A lot of language service providers, including CSOFT, have repurposed this translation quality metric for other types of projects.
Although it addresses seven different error categories focused on quantifiable (read: not subjective) translation mistakes, the J2450 is not enough as a standalone filter. Other metrics, such as those based on the LISA QA Model, the EN 15038, the ATA framework for error marking, etc., should all be considered and—more importantly—weighted to reflect the seriousness or negligibility of a given type of error.
Take your self-defined list of requirements (based on your understanding of your own business and products) to your LSP and work with them to find which error-marking systems should be applied to which types of projects. Once these parameters are set, stick to them, and do not let your reviewers venture into other waters.
This helps mitigate the widely detested subjectivity of linguistic reviews.
Finally, establish a clear pecking order.
Once you’ve minimized the likelihood of your translators and reviewers committing manslaughter, and after you’ve liberally scheduled for a thorough review process and have defined how reviews should be performed… you’re still going to have problems. It’s unavoidable. But disputes can be minimized with proper preparation. So it’s important that you outline your trump cards on paper before the, erm… proverbial excrement hits the wind-circulation device.
When an issue or dispute arises, how are you going to deal with it? How should people argue, e.g., via phone, e-mail, in an Excel sheet or in a review management system? And, most importantly, who makes the final call, or who has the power to stop the madness and implement changes? All roles should be defined at the get-go, regardless of what type of review is being performed.
This prevents the translation review cycle from becoming a downward spiral of pedantry.
Getting what you want
Essentially, the above are all requirements that you should put down in writing in order to progress from knowing what you want to getting what you want. As they are, these suggestions are not complete, nor will they solve all of your review-related problems. But from our experience, they will go a long way in helping to ensure that you’re getting what you want.
This is the second step toward achieving Review Nirvana.
Stop on by this Thursday (because tomorrow is Wacky Word Wednesday) for the third and penultimate step in Translation Review Week, where we’ll discuss Getting the Right People on Board.
Are you a translator, LSP, or translation buyer?
Great! Thanks for reading. At CSOFT, we’re always interested in hearing what our peers have to say. Translation Review Week is all about improving the review process for all parties involved in localization. So if you have any thoughts, suggestions, or disagreements about what you just read, please feel free to leave a comment.
If you want to learn more about translation review, or if you just want to stay in the loop, make sure to visit csoftintl.com!