Never Update Your Client’s TM…and Other CAT Tool Tricks

CAT (computer-assisted translation) tools are now a staple in the translation industry. If you’re not using one, I would argue that you are choosing not to take advantage of one of the best productivity tools available, regardless of the type of translation you do.

Even if you have been using a CAT tool for some time, you may not be taking full advantage for your bottom line. If you regularly work for agencies, it’s safe to say the agency is benefiting quite a bit from your use of a CAT tool. But agencies aren’t the only ones who can benefit from computer-assisted translation. Here are some of the top ways you can use your CAT tool to optimize your translations.

1) Never update your client’s translation memory (TM)

You may be screaming at the screen at this point, especially if you work for or are an agency, but don’t worry, it’s not what you think. Translators who work for agencies or large non-agency clients who manage their own TMs, are accustomed to receiving a translation memory from their client, which they are to use as they translate. But if that is the only translation memory you are using, you are not leveraging your own tool. Confused?

Have you ever come across a 100% match and thought to yourself “that’s completely wrong”? Translation memories are only as good as the translators feeding into them. If you are not the one in control of that, it can be frustrating trying to leverage poor translation strings to do your work. To help with this, always create a second translation memory just for you. Also make sure that your TM is prioritized over the one your client provided. This has the effect of inserting your translation first and if there are no hits in your translation memory, the tool will move on to the client-provided TM. This is especially helpful when you start realizing that 100% matches need to be proofread as carefully as fuzzy matches if you are not the one who translated them in the first place. Using this workflow will increase your quality by ensuring that your consistent and verified translations are prioritized.

Moreover, especially for large projects, this is very helpful in determining whether a given match is from your own TM or the client’s. While the tool may have the name of the user next to each match, this is a much faster way to ensure you’re trusting yourself first.

Now that I’ve just told you not to update your client’s TM, always remember that once everything is finalized, you need to update the client’s TM to contain your translation. But that doesn’t mean you can’t leverage the tool for yourself before you do.

2) Glossaries – not just for terminology

If you are not already using glossaries with your CAT tool, start today. They are very helpful for remembering how acronyms and other terms were handled earlier in the document, but they are useful in other ways as well.

Perhaps if I had more programming knowledge than I do, I would have a more in-depth understanding of the algorithms behind the various CAT tools and how they determine matches. However, I am still regularly surprised when I see what seems to be almost an entire sentence that I am certain I have translated before come up as a 0% match. Sometimes, this long phrase repeats throughout the document. Use your glossary to add the entire repetitive phrase so that even if the CAT tool does not consider this a match, your glossary will.

3) Quality assurance

 CAT tools typically have some sort of quality assurance checker integrated into the system. These checkers vary in usefulness. However, even if you are not interested in this feature, CAT tools still ensure quality in other ways. For example, it never ceases to amaze me how easy it seems for some translators to accidentally miss words, phrases, sentences or even entire sections. Using a CAT tool makes this far less likely. It’s simply impossible to miss a sentence or section using a CAT tool. Although I will admit that it does not protect 100% against missed words or phrases, I do think that the format of CAT tools (typically a table with the source on the left and target on the right with each row being one sentence or string) still improves the likelihood that all words will be translated.

Quality assurance checkers are also a very useful feature to leverage. For one thing, they can typically verify that your glossary has been used correctly. They can also check numbers, consistency, formatting, verify character limits, flag edited exact matches, and more.

CAT tools are not going anywhere, so embracing their many features to increase productivity and quality is essential for any translator on their road to success.

Avatar

Author: Jenae Spry

Jenae has been a French > English translator for over 10 years and a productivity and performance coach for freelancers for over 5 years. Jenae launched the Success by Rx blog to help freelancers achieve success.

Posted in Technology.

6 Comments

  1. I find the feature of checking numbers very useful too. Sometimes number should not be the same (e.g. a phone number may gain or loose a country code in the translation, and commas and dots could be used the other way around in the source and target language), but still it’s a very good overall check of consistency of often very important information in the document. Did you know that spinach is still thought to contain a lot of iron, just because the author (not translator, in this case) of a scientific article on iron content for several vegetables put the decimal dot at the wrong place? 🙂

  2. I use WordFast for translation, and so far I don’t think that it’s necessary to create a second Translation Memory. If I could make it, I woudn’t think that the both TM will work simultaneously. Or, if ‘ve second TM, does it mean that I should copy first the content of the client’s TM to the socond one, and then I might update it?.
    Thank you for this valuable post.

    • You can use multiple TMs in Wordfast (and other CAT tools). That way, I can prioritize my own translation memory. There’s no need to copy content from one to the other. At the end of the project, I simply change my client’s memory from “read only” to “write” and update it with my completed translation. Hope that helps!

  3. Thank you for the article! It’s an interesting idea about creating my own translation memory for a project. I will try that and check if it’s convenient. I use MemoQ – my favourite CAT with a good QA-tool.

Leave a Reply

Your email address will not be published. Required fields are marked *