Did you recognize conversations you’ve overheard in your company about translation in part one of our series? Below are five more myths we encounter on a regular basis, and the odds are you might, too. Prepare yourself for your next team meeting with this expanded list of translation myths.
Myth #6: “It’s cheaper to have internal staff handle translation for us.”
The call to “do more with less” is fairly common among executives and project leaders, and when it comes to recruiting outside agencies to assist with translation projects, there’s a often a reflex reaction to round up an ad-hoc team familiar with the target language. If this is the case in your company, you should apply the other management adage: “Inspect what you expect.”
The results of an in-house translation project can often fall far short of your business objectives. Assigning translation duties to employees with existing full-time obligations typically shorts the project of the focus and attention it requires. A lack of attention to detail means your cost savings are obliterated when the project drags on, launches with errors, and enlists more resources for crisis management. Only then does it become clear: Investing in a professional translation partner isn’t as expensive as you think.
Myth #7: “There’s no need for technical QA after translation.”
If a translated document says what it needs to say clearly and accurately, there’s not a lot left to do. When a translated app or software release says what it needs to say, you’re only halfway to the finish line. A good quality assurance (QA) process goes beyond the message and digs into the interactive experience in translation. Your domestic check-out process might make perfect sense to a user in the U.S., but perhaps there are steps in the Japanese version which cause users to abandon the process. Is it about trust? Design? Usability? Is an icon or direction unclear? Are the forms working? All of these details play into the need for technical QA.
Myth #8: “Translation bugs can be deferred until someone’s free to work on them.”
When it comes to building trust with your customers, 90% right isn’t 10% short. It’s 100% short. Mistakes, however small, can erode your brand equity or launch your brand with a bad reputation in a new market. Poor translation sends the message that you don’t take the market seriously. An image that didn’t get vetted by localization experts can be downright offensive. Technical errors can lead to larger problems with snowball effects. Consider, for example, if your users are entering data in the wrong form fields because of a misunderstanding in cultural expectations, currencies, or other measures? Unchecked, you can be creating a tangle that will only be more expensive to fix later. Don’t let the bugs ride… squash them ASAP.
Myth #9: “We don’t need a glossary because we’re using translation memory.”
There are a lot of terms in the translation and localization industry that sound the same, but mean completely different things. (After all, we’re in the business of using the right word in the right place!) translation memory is a database tool leveraged by software to help human translators accelerate their translation speed and accuracy. It contains segments of text in the form of phrases, sentences, or paragraphs that are paired with accepted translations in one or more target languages. Translation memory is frequently used in projects where content is repetitive, such as technical documentation.
Glossaries are collections of terms or phrases and their approved translations. Glossaries often contain the accepted translations of marketing terms and corporate language, and help translations maintain consistency, even when projects are managed by multiple translators. Reducing this uncertainty through a sanctioned and widely available resource can greatly reduce the cost of a project.
Myth #10: “Translation memory and machine translation is basically the same thing.”
Most confusion about translation memory (also referred to as computer-assisted translation) and machine translation (MT) centers on the very squishy term “computer translation.” There’s a natural tendency to divide translation into “human” and “computer.” The truth is not so binary.
As we discussed in Myth#9, translation memory/CAT is used in software designed to help translators improve speed and consistency by automatically matching and proposing translated phrases. Machine translation (MT) is used in situations where there is an extreme degree of standardization and zero room for interpretational ambiguity.
Typically the investment in MT is only justified when there is a tremendous volume of content to be translated, as the initial setup for MT can be costly. Even with the best MT, results are below the high quality you’d expect from human linguists (and may require post-translation editing).
Congratulations! You’ve just completed the final five of our top ten misconceptions about translation. You’re now well armed for your next internal roundtable on taking your company global. Feel free to share these articles in your organization. There’s tremendous opportunity waiting for you worldwide.
