The Only Localization Glossary You’ll Ever Need
The Only Localization Glossary You’ll Ever Need
We aim to do more than just define terms with this localization glossary. We’re here to offer you a comprehensive guide to the localization industry so you’ll understand how to actually apply these terms and concepts in your dealings with LSPs and other localization companies.
With this authoritative localization glossary in your back pocket, you’ll be prepared to:
- Evaluate a localization company’s offerings;
- Ask detailed questions about their “advanced technology”;
- Identify any potential drawbacks in their processes;
- Determine whether a localization partner is a good fit for you.
When you understand and use the jargon of the localization industry in these conversations, you’ll be better equipped to identify transparent localization companies among all the other “black box” providers. That could result in your company solidifying a localization partnership that lasts for years and results in thousands of dollars saved.
It all starts with knowing the language.
The Ultimate Localization Glossary
Taking written words from a source language and conveying them in a target language. Translation is usually performed by skilled linguists but can also be completed by machine.
Machine Translation (MT)
Translating content into other languages using automated translation tools such as Google Translate or Lilt. Machine translation can be used alone or as a first step for human linguists (a practice which can be referred to as “post-edited machine translation” or PEMT). There are a few classes of machine translation software currently available on the market, and those are listed below in order of increasing sophistication:
- Rules Based (RBMT)
- Statistical (SMT)
- Neural (NMT)
A multifaceted approach to translation, where linguists combine cultural relevance with technological function to produce translations that are more effective in a digital context. A localization project always focuses on a specific locale—a location on the globe, characterized by the language and culture of the target audience there. As such, localization allows companies to target far more specific markets than does translation alone.
The ultimate goal of the localization industry: fully continuous, automated translation that runs simultaneously alongside agile development timelines. The automation present in an API-integrated platform such as Bureau Works gets close real-time localization, but the industry as a whole still has a long way to go. With each new technological advancement, we get closer to the reality and the opportunities of continuous localization.
Altering content to create a better translation result, instead of translating “word for word.” In truth, this is an overused and outdated term that’s mainly intended to manipulate LSP clients into purchasing higher-value services. When most people ask for transcreation, what they really want is high-quality translation. A good linguist does the very same thing. Creative target language copy-editing is a good alternative to transcreation.
Designing or retrofitting a product or website to be compatible in a multi-language format. Internationalization is a crucial step in the localization process, and should be completed far ahead of any translation work. Internationalization often involves using pseudo-translation to check for proper text expansion, optimize LTR and RTL user interfaces, and test UI/UX in other languages.
Not to be confused with globalization—glocalization is a term that was originally coined in the 1980s that refers to the combination of global and local considerations—for specific locales around the world. The term was later adopted by social scientists and business professionals around the world to describe the practice of having a simultaneous global and local mindset.
To apply these buzzwords in your own localization efforts, ask any potential localization services how they integrate machine translation into their process. Don’t get hooked by the transcreation bait when high-quality translation is the answer. Regardless of which service you need, ask to know your translators by name and test their abilities before you sign a contract. Adopt a glocalization mindset now, and apply it to everything from product development to social media marketing. That way, you’ll be fully prepared when it comes time to localize your content.
Types of Language Vendors
A “single language vendor” is a translation company that offers content services in only one language pair—for instance, English to Korean. Some enterprises assign project managers or language managers to wrangle a whole host of SLVs to translate their content.
A “multi language vendor” consolidates work relative to an SLV. MLVs offer translation for many language pairs and often operate by contracting the work out to various SLVs abroad. MLVs take care of the project management work. It can save time, but they may also increase the distance between you and your translators.
A “language service provider” is technical any company that offers services related to language, but it is often used as a synonym for MLV. They offer a combination of translation, localization, or other language-related services in many different languages. Most LSPs (whether offering one or multiple languages) tend to operate with a “black box” mentality—never letting their clients see behind the curtain and understand how they do business. This old paradigm often involves underpaid translators and low-quality results.
A “localization management platform” is a software and service combination that can centralize all of your language-related efforts into one convenient and highly efficient platform. BWX is one such platform; it offers API and CLI integration, high-quality translation, and back-end services, such as multimedia localization tools, to help localization run even smoother.
Before deciding which type of vendor might be the best fit for your company’s content needs, interview a few of each to consider how they might support your objectives and your workflow. We recommend providing every vendor on your shortlist with a very specific sample project (say, a set of multimedia files) and evaluating their process along the way. That way, you can get an accurate feel for their pricing and their operating style before signing a contract.
Tools and Technology
Software that facilitates translation by human linguists. At one point in recent history, these tools were referred to as TEnT (translation environment) tools. In the software world, managing file formats and isolating translatable content can get very complex. Computer-assisted translation or CAT tools automatically intake or parse files, lock anything that is not translatable (including code, mark-up, and variable placeholders), and give translators access to only the translatable content and protected placeholders that the translator may need to move within a sentence. CAT tools allow linguists to work with source and target languages side-by-side. They often incorporate additional tools like translation memories (see below), term bases (see below), and machine translation tools which make translation faster and easier.
The term “strings” is most often used in the context of software, and refers to localizable content. A string is a series of characters that represent a data type in software development. These are usually phrases that fall somewhere between individual words and entire sentences. During localization, you will export strings from your code repository or content management system to be translated by machine or by a linguist.
Most often referring to “content management systems” or CMS, this term incorporates all the systems your company may have in place that host content. That includes software such as Zendesk, Drupal, and Adobe Experience Manager.
Application Programming Interface (API)
A tool that enables connections between your content management systems and your translation/localization software. With API integration, these systems can “talk to each other” automatically—eliminating the need for files to be emailed or manually exported. Taking it a step further, some LSPs let you set up a custom CLI to deploy APIs with even more efficiency, control and flexibility.
The localization industry standard exchange format. In the absence of API integration, XLIFF files are often used as a container for full bilingual sets of content, almost replicating a TM. Thus, they’re one more tool for companies to use in the pursuit of more efficient translation.
LQA / FQA
“Linguistic quality assurance” and “Functional quality assurance”—two processes you absolutely want your localization partner to be using with regards to your content. LQA involves viewing translations in context to ensure no errors or misconceptions are present in the final product. FQA involves testing all linguistic elements to see whether or not they “break” when tested in other languages.
Endeavor to find out as much as possible about how your potential localization service uses the tools on this list. Figure out which CAT tools they’re using. Be aware that most LSPs will claim they can connect with a seemingly endless list of CMS through API, but few can actually deliver on that promise. Which integrations matter most to you, and how quickly can they have those up and running after you sign the contract? While integrations are the “bells and whistles” of the localization industry, don’t forget to ask questions about the less-flashy QA process—and make sure any potential partner of yours can provide comprehensive details about how they’re managing translation quality.
Translation Memory (TM)
Perhaps the most crucial tool in the localization manager’s toolbox. A TM is an ever-growing database of translated strings that automatically updates with each translation completed by your team. LSPs can offer translation discounts based on the degree of similarity between new sentences and sentences that are already translated in the TM. Linguists can leverage a well-maintained TM while working within a CAT tool to save time and effort. TMs can suggest already-translated strings for one-click implementation. Linguists can also reference TMs to ensure that new translations are consistent with older ones. They’re most often exported as .tmx files.
How closely a new string matches what has already been approved in the TM. A 100% match would mean you already have the entire phrase translated and saved in the TM, so the translator only has to click “approved” to complete the translation work on that string. Good localization companies don’t charge anything for translations that are a 100% match. The best will even offer you a discount for matches that are considered “fuzzy” or partial.
The TM was able to pull previous translations for some of the words in the string, but not the whole phrase. For instance, consider the phrase “We offer the best technology on the market.” If the TM already has a translated version of “We use the best technology in the United States,” that could be considered a fuzzy match. The TM did 80% of the work, and the translator has only to edit the remainder of the translation, not start from scratch. The best localization services monitor the percentage of each fuzzy match and offer discounted rates accordingly—meaning you pay less for matches as they get closer to 100%. This is precisely how using a well-maintained TM can save your company a lot of money.
Similar to a TM, a terminology database (also called “term base” or TB) is a collection of mission-critical terms that have been translated by a linguist to serve as an authoritative reference. These terms are intended to be applied consistently across your content and don’t need to be re-translated each time. TBs include terms related to product names, slogans, UI elements, and other specific terminology that must remain identical after translation to uphold content integrity. Term bases also indicate certain terms with the DNT / “do not translate” distinction; for example, you likely won’t translate brand names into the target language. Term bases are exported as .tbx files.
A style guide, also referred to as a “brand guide,” indicates the elements of tone, voice, and word choice that should be observed in each target language. This guide helps translators create content that maintains a consistent brand voice. With a style guide in hand, you can be sure that your entire website, application, or elearning suite conveys your brand’s characteristic tone, whether formal, edgy, or business casual.
You can learn a lot about a localization company by observing how they maintain and control terminology assets. Really home in on your discussion of translation memories when interviewing potential services. Ask how meticulously they’re maintained. Ask how much they charge for certain percentage matches. And ask if they’ll let you retain your own TM after the contract is over. Chances are, if they answer these questions positively (and transparently), they’ve got a pretty good grasp of localization best practices.
Making the Localization Glossary Work for You
If your company is going global and you’re looking for localization services, you’ll start to recognize these terms on websites across the industry. Use the Ultimate Localization Glossary to identify each company’s unique claims and to challenge them to back up those claims. With a firm grasp of the jargon, you’re less likely to be led astray by shiny marketing and bold promises. You’ll have the basic knowledge to weed out the empty potentials, to keep your larger localization ROI in focus, and to find the partner that’ll have your back.
Bureau Works is a localization management platform that harnesses the best of automation and human innovation to create content that converts around the world. We’re proud to run our business on the basis of radical transparency. If you’re looking for a localization service you can trust, contact our team. We’d be happy to serve as your first (and potentially your last) prospect.