As we discussed in our last blog on translations, the volume of non-English surveys are increasing in market research. That means researchers must take the steps to properly translate and localize surveys to ensure their studies get high respondent engagement and high quality results.
If you’re reading this and thinking that you don’t know where to start with translating your studies, you’re not alone! Many researchers need guidance when it comes to sending their surveys to a global audience. We’re here to help. This blog post will explain how a tech-forward approach can make the translation process easier, more efficient, and more affordable for you.
Where Research and Technology Meet
While terms like Machine Learning (ML) and Artificial Intelligence (AI) may sound intimidating or too “high-tech” for research, these technologies have made a tremendous impact on our industry – and they’ve made your research workflow much easier!
In fact, ML and AI have transformed many industries to-date from Manufacturing to Retail, and indeed Market Research has itself evolved in the face of such technological developments. Lucid introduced programmatic technology to the market research industry back in 2010. Today, programmatic technology has become the industry standard, and researchers can instantly reach survey panelists across the globe.
But what do such technologies signify for the translation and localization processes that are so essential to international market research?
Research Will Always Require a Human Touch
Where automated processes increase efficiencies in many industries, human skill and knowledge remains the powerhouse behind translation and localization. Without it, the cultural and linguistic nuances that make each market so unique are at risk of being unconsidered, impacting the quality of translation and subsequently the quality of insights derived.
Relying solely on machine technology in the translation process prevents you from obtaining a precise localized output. Technology does not yet have the ability to interpret context, tone of voice or sarcasm – a fact that is unlikely to change in the near future.
Such context is key to understanding, whether communication is written, verbal or visual. Take the word ‘stream’ for example; nowadays it is a verb related to accessing TV shows, films and music on-demand. Yet without awareness of context, ‘stream’ could just as easily be translated to ‘small, narrow river’.
This is the risk taken when relying on solely automated translations, hence at the very least a human audit stage is essential to any translation project, particularly when looking to achieve an intelligible localized output.
Efficiencies Achieved through Translation Memory
There are technologies that, when combined with core human translation, enable efficiencies in the translation process.
Translation memories are one example of such technologies, where previously translated content is stored and utilized in future projects to minimize the need to translate a word or phrase twice, which also means you won’t be paying for the same content twice.
The translation memory helps create efficiencies across cost and time, meaning the more you translate, the more you save. Utilizing an existing translation memory will also ensure consistency with historic stylistic preferences, creating uniformity in your research.
On average, a client working with a 15-minute (~3000 word) survey saves approximately 500 words (16% saving) on overall translation costs by employing a translation memory. These savings can be further augmented as new content is also eligible for savings via word or phrase repetitions to varying degrees.
Best Practices on Combining Human and Machine Translation
Here are some best practices that can be used to combine human and machine translation, especially in the context of market research.
1) Remember that human translators cannot be replaced with machines
The vocabulary of the entire market research sector is almost a language in itself. This is why it is so important that the translations are completed by a human. Machines cannot replicate style or tone, nor can machines relate words to context, or “slang” used globally. Humans are the only translators that understand how to translate your complex content and conversations to capture those cultural conventions of the markets focused on and into the target language.
2) Use machine translation for open end responses
Where you can leverage the machine translation – with human linguists applying an almost “Light Touch” approach to post-editing – is for open end responses. This level of service is aimed at providing accurate and readable translated content with the understanding that the grammar and flow of the target language may not be on par with human translation. Clients who are utilizing this dual-approach receive a reliable yet economical way to access translated content that would otherwise be deprioritized and or potentially very expensive with verbatim human translations.
3) Leverage the help of translation experts
As we mentioned earlier, not every researcher is an expert in survey translation – and that’s okay! When you are running multinational studies, don’t be shy about asking your sample or translation partners for help. They can help you understand how your studies will need to be translated and localized and provide the guidance you’ll need to optimize your research without breaking the bank. For more information on translation best practices, contact your Lucid representative or learn more at Language Connect.