Translation Memories: What are they, and why do I need them?


Win-win situations are ideal for everyone. For translators and their customers, a Translation Memory is something worth celebrating! A Translation Memory helps translators work faster and more efficiently. Not only does this improve the workflow for the translator, but it helps the client by making sure the results are of a consistent and high quality. And last, but not least, the client can save a lot of money. 

What is a Translation Memory?

In a nutshell, this is a software that makes sure you never have to translate the same sentence twice. This means that the time spent working on a translation is significantly reduced, especially when it comes to revisions or content that contains a lot of repetition. A translation memory works like a database that stores previously translated texts. Words, sentences – even entire paragraphs – of a text is divided into two parts: source language and target language. The source language is the original language of the document you want to translate. The target language is the translation of this text. Together these pairs are known as a “translation unit” or “TU”.

“Maybe it is time to use a modern tool that ensures consistency, lets your company communicate what is important, creates believability and drives more sales?”

How does a Translation Memory work?

Translation Memories are a part of larger systems known as CAT systems (Computer Assisted Translation). A CAT system is a software that is used by professional translators to increase their productivity, eliminate mistakes, and improve the workflow. 
Translation Memories are an important part of CAT. The system starts a translation by opening a document, often referred to as the “source file”. It then proceeds to check if any of the content already exists in the Translation Memory. The tool searches for “100% matches” (identical matches) and “fuzzy matches” (similar matches). A “fuzzy match” means there are no identical matches for a term or sentence, but that there are segments in the translation memory that are similar to the text you are working on. 

When you work with a Translation Memory, human judgement is incredibly important. The translator can accept, change, or modify the system’s suggestion. Even if a Translation Memory finds something that corresponds with the translation, the translator needs to verify how well the match fits in the given context. If a segment does not correspond (this happens if it has not been translated previously), the Translation Memory will store the new TUs, and this makes it so that the Translation Memory can keep growing and improving. 

Example of a Translation Memory


Using a Translation Memory helps aid localization, and improves quality and consistency. It also helps translators work faster and more efficiently. To work with a Translation Memory is not the same as using a machine translation tool (or “MT”). In contrast to MT, which is an automatic translation without input from a translator, a Translation Memory is built by professional translators. A Translation Memory helps you tailor the content to a specific language, culture, and audience, so that the content feels natural and is easy to read. 
Nothing communicates better than a well-written text. It is the text that generates knowledge, sales and understanding. Large marketing campaigns are not worth much if the text is full of mistakes or does not reach your audience. 
Maybe it is time to use a modern tool that ensures consistency, lets your company communicate what is important, creates believability and drives more sales?