The ABC of Cost-Effective Translation Workflows: Part I
This time in our blog series on translation and cost efficiency, we’re taking a closer look at different technologies and workflows that can help companies to save money.
The modern translator is no longer seen as a humble wordsmith who sits alone in their solitary chamber, poring over piles of dictionaries. These days, they’re considered “linguistic engineers” – and they play a key role in increasingly complex translation projects in which they rely on cutting-edge technologies. But which tools and work methods are considered state of the art in this rapidly developing, high tech sector? And how exactly do they help companies to cut down on excessively high translation costs? For the next two weeks we’ll be publishing four best practices for cost-effective translation workflows. Here are the first two.
The sliding scale that is translation quality
SCENARIO
In the same way that you would spend much more time composing an important press release than writing an e-mail to a fellow team member, the amount of work that goes into a translation should always be proportional to its purpose.
Your software’s user interface should gain the user’s trust and be highly accurate. In order to craft the best possible user experience, you’re best off having this content translated by an experienced translator and then having it revised. On the other hand, if you needed the text quality to be good – but not highly polished – then a translation and review (such as for a blog post on your website) may well be the right choice.
However, if you really wanted to free up your budget, then machine translation (MT) would be right up your alley, as this can be up to 60 percent cheaper than conventional translation. This workflow is used when it’s important for a text to be comprehensible and its terminology accurate. There are a few areas of software in which it makes sense to employ MT, such as for extensive user documentation.
Milengo’s Managed MT service offers a highly automated translation workflow including a human quality assurance step that can be used, for example, if you would like to have your online help translated into another language – but would like to save on costs due to a low number of users in the target country and a minimal return on investment.
If all else fails, you also have the option of temporarily doing without translations that are not essential. We’ve illustrated what you need to bear in mind when it comes cutting costs for specific content types and target markets in our previous blog posts in this series – check them out to maximize your software localization budget!
STRATEGY
Define flexible quality levels for your translations – these will help you to avert additional costs due to superfluous workflows.
POTENTIAL SAVINGS
Automate and integrate
SCENARIO
The more seamless your software and documentation localization processes are, the less time and mental space you’ll have to dedicate to administrative tasks.
That’s why companies are moving towards middleware applications that are able to connect a range of otherwise incompatible content authoring, content management, and translation management systems involved in the translation process with one another.
The advantages become crystal clear when you compare this approach with outdated software localization workflows in which exporting and re-importing software strings had to be done manually and using exchange formats like CSV or XLIFF. In some cases, companies would even compile the text components to be translated in Excel tables before laboriously copy and pasting the resulting translation back into their software repositories. These type of arduous workflows lead straight to bottlenecks – especially when it comes to agile software development with its tightly scheduled release cycles.
STRATEGY
Get rid of manual workflows to reduce overhead costs.
POTENTIAL SAVINGS
In the second part we’ll cover how using translation memory and turning your gaze towards outsourcing can help you free up your localization budget.
