Website translation is complicated. It often requires a time-consuming, labyrinthine process that includes:
- Gathering all the right content for translation
- Sending it to a translator team
- Overseeing the iterative translation and quality control process
- Getting the translated content back
- And then publishing it online where needed
And if you create new content, you have to repeat the process all over again…which costs time, and in most cases, more money.
Fortunately, translation memory technology can help the process, and drastically simplify website translation projects.
What Is Translation Memory?
Translation memory is a database that stores all of a website translation project's translated content. This content is often stored in word-, phrase- or sentence-length chunks called segments. Segments are used by translation management software to help human translators, and speed up the process of publishing content that's previously been translated.
For example: Let’s say you’ve translated your website for French-speaking customers, but add a new section to your site at a later date. Your translation team can use the translation memory to find phrases within the new content that have been translated before.
They can manually publish this localized content, preventing duplicate translation efforts … and save time and money in the process. Alternately, these translated segments can be auto-populated with the help of software.
How Does Translation Memory Work?
When you translate your flagship website for the first time, great website localization vendors will help identify the content that repeats often—think phrases like “product description”—or is core to your brand, like a company description.
Your vendor’s technology team will identify and mark those segments of content with a special code. The translation memory database will capture those translated segments in your flagship site’s origin language, along with their translated counterparts. These pairs are known as “translation units.”
As you publish new content on your flagship site, your translation teams can access the translation memory and ask it to process the new content for your localized site. Translation memory will analyze the new content, looking for previously-translated phrases.
The database will show exact match segments—those that are identical to previously translated content—as well as fuzzy match segments that appear to be similar to content that’s been translated before.
Translators can elect to deploy the existing translations, rather than putting the new content through the translation process again.
The Benefits of Translation Memory
Having a database of previously-translated content can help you and your teams with:
Speed
Translations are completed and deployed more quickly, saving teams valuable time and effort, and preventing duplicate work
Cost-Effectiveness
Scrupulous website localization vendors won't charge you to translate this previously-translated content more than once. That means your translation spend is minimized, while the impact of your translations is maximized as it can appear on your localized site dozens (or hundreds) of times at no additional cost
Consistency
Translation memory ensures that translations are accurate and preserve the brand voice, cultural nuance and technical detail of previously-translated content
Plus, when you work with a fully turn-key website localization provider, translation memory is a seamlessly integrated part of the translation workflow. You'll never have to bother with building the translation memory the first time, nor managing its use and deployment in later projects. It's all included, with no effort required from you.
Translation memory is a handy and powerful tool for companies that need to quickly and easily deploy multilingual websites, and is a critical component to scaling your growing global business online.
Last updated on July 03, 2018
