Let’s start with the basics: what exactly is a translation memory?
A translation memory (TM) is a database of translated texts, divided up into sentences or segments. This memory helps translators by giving them suggestions based on previous translations.
For example: you might need a recipe to be translated every week. Sentences like ‘finely chop the onion’, ‘season with salt and pepper’ or ‘combine the ingredients’ recur time and time again. Saving them in a translation memory means that the translator doesn’t have to translate them from scratch every time or search previous translations to find out how they have translated them in the past.
A segment that is exactly the same as a previous one is called a 100% match. A partial match (where a lower percentage of the content is the same) is called a fuzzy match. For example: ‘finely chop the onion’ and ‘finely chop the shallot’ is a fuzzy match. Here, the translator only needs to change one word from the existing translation.
When is a translation memory useful?
- For large amounts of text. In other words, there’s not much point using a translation memory for a one-off invitation.
- For texts containing a lot of repetition. We mean documents like user manuals, instructions, contracts and so on.
- For texts with fixed terminology. Less ‘creative’ language, as it were.
All the advantages of using a translation memory
More consistent
With a translation memory, you can rest assured that the translations will be consistent, irrespective of which translator does your assignment.
Faster
The more text the translation memory contains, the less work the translator needs to do themself and the faster you will receive your translation.
Cheaper
When we set up a new translation assignment, we analyse the number of matches with the translation memory. The more matches, the more attractive your quote will be!
Bonus: a terminology list gives you greater control
Do you already know how you want certain (subject-specific) terms to be translated? Do you have a specific tone of voice that includes fixed translations? We can include all these in a terminology list or ‘term base’, which supports the translator in much the same way as a translation memory, by offering suggestions.
Do you want a discount based on the number of matches in the translation memory?
Of course you do! As we explained, the number of matches has a direct effect on the cost price of your translation. We calculate the discount as follows:
| REPETITION TYPE | DISCOUNT |
| Context match* | 100% |
| 100% | 50% |
| 95-99% | 25% |
| 85-94% | 15% |
| 75-84% | 0% |
| 0-74% |
* A 100% match with exactly the same sentence before and after it.
Want to find out more about translation memories or translation software in general?
Read the interview with our computer-aided translation (CAT) expert Kobe or contact Margo for a chat!