Reducing survey translation cost is on every researcher and end client’s mind, especially when it comes to high-volume content. But if you know about the project management triangle, you’ll know that cheap prices tend to lead to lower quality and/or slower turnaround.
The good news is that, especially when it comes to surveys, there are ways to get a bigger bang for your buck. Since translation cost is usually based on word count, it follows that a lower word count leads to an overall lower project cost. Fortunately, writing shorter surveys alone is not the only way to achieve a lower word count.
If you’ve been keeping up with the Language Intelligence blog, you’ll know that we bring up Translation Memory technology quite a lot. But that’s not all we use. We also use the tool replock™, which increases TM savings by “locking” repetitive text segments so that a translator doesn’t translate the same phrase twice.
The problem is, if there aren’t any repetitive segments of text, tools like this won’t be much of a help. That’s why we recommend making your surveys consistent wherever and whenever possible.
Computer Assisted Translation (CAT) tools like Translation Memory or replock™ work best the more consistent your surveys are. While certain questions may need to be worded differently to capture different nuances, other items in your survey, like instructions or screener questions, can be made consistent or boilerplate. This will reduce survey translation costs.
Making the language of your survey as consistent as possible will effectively reduce your “unique word count”—the words billed at the full per-word rate after Translation Memory analysis.
You can also reduce your word count, and thereby reducing survey translation cost, by using boilerplate text as much as possible. There might be certain text in your survey that can be reused in multiple places. Ranges or scales, for example, could be turned into boilerplate text.
Do you always use a five-point scale? Do you usually label all of the points the same way? If that sounds familiar, you can have your scales translated once and reuse them for future surveys.
You could also reuse demographic questions by making them boilerplate. Just keep in mind that you may have to formulate your questions slightly different depending on each target market.
In the US we tend to talk about annual income, whereas many European countries refer to monthly income. You may want to develop a standard set of demographic questions for the countries in which you most commonly field surveys.
By translating your boilerplate demographic questions once per country, you can take them out of the word count for later surveys.
Another way of reducing survey translation cost is by making sure that only text that needs to be translated is included in the word count. You’d be surprised at how much text in your file doesn’t actually need to be translated.
Programming code, notes about survey logic, or other behind-the-scenes text can sneak into your word count. Language Intelligence has an in-house tool called intellitranslate™ that helps your project manager “lock down” any text that shouldn’t be translated and make sure that it’s not counted as part of your quote.
That said, it’s always good to let your project manager know what does and doesn’t need translating.
If you’re interested in survey translation, you may also be interested in mobile survey translation, which you can read more about here. If you have any specific questions about survey translation, you can contact us by reaching out to us at firstname.lastname@example.org.
If you’re interested in learning about what we can do to help make your international survey project a success, check out our market research translation services page.