If you have researched translation for multilingual websites and mobile apps, you’ve seen a wide range of solutions for translating and updating multilingual digital media. These are packaged as translation "platforms" or "proxies" or, more recently, "localization management solutions." Some are developed by startups focusing specifically on websites and mobile apps, while others are proprietary tools that have been developed by language service providers. In this article, we will refer to these as "automated solutions."
In general, automated solutions establish a cloud-based framework to translate a website or app and, in some cases, for continuing translation updates on an ongoing basis. After the initial translation project, the client can use the interface to make updates, which trigger micro-translation projects for translators to fulfill almost in real time. Translators log in directly to the tool to translate the client’s changes. The leading platforms can handle an impressive array of file formats, even converting them on the fly. Many also provide the translators with an interface that "mirrors" the look of the source application or website. This helps the translators see the content in context and compensate for text expansion or contraction during translation (some languages use more characters than others, which impacts the "look" of the project).
When an automated solution claims that it provides a seamless alternative to what they characterize as tedious "manual" processes, be skeptical. Translation project managers are already skilled in managing digital content, and industry-standard localization project management is NOT limited to cutting and pasting content to and from Excel worksheets. The translation industry has embraced content management technologies on pace with the field of technical communication in general. In fact, automated platforms offer many of the functionalities of translation tools that have been in use since the 1990s. Although we use the term "hands-on" processes to distinguish them from automated solutions, they are still a far cry from "manual."
What tools are already in use?
Like professional technical writers, professional translators use tools for structuring content, ensuring consistency of style and terminology, and performing quality assurance. Market leaders like MemoQ and SDL Studio (and a host of competitors) have been developing and refining their products for decades. Increasingly sophisticated CAT (computer aided translation) tools are available as both desktop applications and cloud-based collaborative platforms. Competition between brands is intense, but a certain amount of compatibility has developed between them. The industry standard XLIFF file (XML Localization Interchange File Format) allows for translation files to be shared between different tools, with only occasional issues caused by variations in implementation. Nowadays, being able to use these tools has become necessary for a successful translation career. Practically every reputable freelance technical translator and every professional language service partner or agency has expertise with one or more CAT tools.
CAT tools break content down into segments and present them in a two-column source-target interface. The tools provide termbase management, controlled authoring, style guidance, and QA functionality. Translation memories (TMs) make pre-translated segments available for re-use across media platforms and over a lifetime of updates. Automated solutions have adopted these capabilities as well, albeit with varying levels of quality and degrees of success.
From the translator’s point of view, years of subject matter expertise is codified in their own personal termbases and CAT toolcustomizations, in addition to investments in training time and licensing fees. Translators prefer to use their own tools (or combination of tools), and in-demand translators can be choosy in declining projects that require them to learn new tools. When quality is paramount (and it usually is), businesses should avoid automated solutions that very few translators are willing or able to use.
Why is localization engineering necessary?
In a sense, the purpose of CAT tools has always been "automated localization engineering" for digital content. Most file formats have become so commonplace that we hardly even consider them "digital" anymore. Twenty years ago, translation looked very different than it does today. Today, a translator can import a Word document or an InDesign file into a CAT tool, view the text without the distraction of markup/tags, translate it, and export a target document with the formatting intact. SDL Studio, for example, currently supports 70 different file types.
For mobile apps, websites, technical drawings, and elearning modules, separating content from code is rarely as simple. Putting the application or module back together again after translation can also pose challenges, especially when multiple languages are involved. When code is mistaken for content and vice versa, problems occur. Recoding can be required to create a usable deliverable.
As digital technologies multiply, technical translators face these problems:
CAT tools are compatible with many file formats. However, they are not compatible with ALL file formats.
Even if the file formats are compatible, the export and import functions for the client’s authoring platforms will vary in quality.
Authoring standards for the source content itself can also vary in quality. Custom coding, shortcuts, and workarounds can all interfere with the CAT tool’s ability to read and manage the content.
When choosing a localization strategy, a client should not assume that "automating" the process is going to be more cost-effective. Instead, they should ask these questions:
- How localization-friendly is the website or app?
- What is the expected frequency of changes and updates?
- What in-house resources can I devote to localization engineering?
Regardless of whether you will use automated or hands-on solutions, the best way to reduce the costs of localization engineering is to follow best practices for internationalization from the start.
Internationalization best practices
What is the first step toward localization-friendly content for apps and websites?
- Protect the code. Keeping the content separate from the code makes it easier to isolate the content for translation.
- Don’t hard-code dates, times, measurements, or currencies.
- Don’t concatenate strings to form sentences. Remember that grammar and word order vary across languages.
- Don’t embed text in graphics.
- Support different character sets by using Unicode.
- If certain features won’t be used internationally, make them easily disabled options.
- Store strings in resource files.If an app or website is not already internationalized, recoding will be necessary to accommodate the needs of a global audience. Some automated solutions promise to take the internationalization step out of the equation by creating a proxy site or replica to serve as the source for the localized sites.