Localization strategies for L&D content: the human element
16 Sep 2024
6 mins
This is the third in a series of blogs I'm writing following a survey by RWS and Training Industry of more than 300 learning and development (L&D) professionals, exploring their localization strategies for L&D content.
In the first of the series, we saw that their top localization challenge is how to incorporate cultural nuance into training materials. In the second of the series, I looked at concurrent authoring as one way to address this challenge. Now I want to look at the role of human language specialists and their ability to help you address the challenge of cultural relevance in training.
Approaches to translation
The survey asked which translation methods the respondents are using for localization (see the chart below – and for regional differences, download the full report).
Because they could pick multiple options, those using a combination of methods had several ways to reflect this fact. So although you can't see it in the diagram, we do know from further analysis that only 9% are relying entirely on human translation and only 8% entirely on machine translation (MT) alone (i.e., without human review and refinement).
This means the vast majority are making choices to use both human and artificial intelligence to some extent as part of their localization strategy for L&D content. They may only be using AI-powered MT with human review and refinement, also known as post-editing (in fact, 19% are using only this method of translation). Or they're using some mix of human-only, MT-only and/or post-editing methods across their portfolio of courses.
It's a balancing act
This tells us two things:
1. The value of machine translation is fully recognized. More than 90% are using it to some extent, with or without post-editing – probably using both options at different times. And the reason is obvious enough: AI-powered MT is fast and cost-effective, and it's one way to combat a lack of internal resource or expertise for localization (a challenge covered in the first blog of this series).
2. But machine translation also has shortcomings – at least in the context of L&D – requiring human input of some kind. This is why more than 90% are using human translation or post-editing – and, again, probably both at different times. Bringing humans into the loop signals either that the content requires a level of nuance that MT isn't yet up to all by itself. Or it signals that the value of the content is too high to trust it entirely to AI, even if the risk of error or poor translation with MT is small.
We know that both reasons for human input apply to localization strategies for L&D content. I've already extensively covered the fact that incorporating cultural nuance is the top L&D localization challenge for L&D. And if you read the full survey report, you'll see clear evidence for the value that enterprises place on training and its localization. They're engaged in a constant balancing act to be as efficient as possible with the help of AI, without sacrificing the nuance and quality that only human intelligence can add.
The limits of traditional localization
If the choices of these enterprises suggest that they understand the importance of human input to create culturally nuanced content, it's natural to wonder: why is incorporating cultural nuance still their number-one challenge?
One of the reasons is that many enterprises still interpret localization too narrowly as translation – or perhaps as translation+, where the 'plus' acknowledges that non-literal language, examples and images may need to be adapted for different cultures. If 'translation+' is the extent of your localization brief, the ability of your post-editors or translators to deliver culturally relevant content will always be limited.
Sometimes, localization needs a less limiting approach involving what we call concurrent authoring. This allows course development to address every element that may be affected by cultural differences, including formats, structures, styles and narrative approaches that suit different prevailing learning preferences.
Goodbye translators, hello language specialists
Another challenge faced by enterprises is finding the right people to take localization beyond 'mere' accurate translation. Enterprises need their translation teams to step up and become language specialists who are:
- Creative linguists and cultural experts able to craft content that resonates with the target audience and makes them feel that a course is created specifically for them
- Smart technologists able to harness AI and other technologies to elevate their craft more efficiently, while continuing to adapt as new tools emerge and existing tools advance
- Content optimizers able to work with content in all its forms, for any channel or learning purpose
Not all translators or translation teams are able to step up in these ways, and this can make finding the right localization partner for L&D content challenging. To learn more about what to look for in a training localization partner, along with training content localization tips, read the survey report, Learning across borders: a survey of localization practices in corporate training and development.
Isaura Saguer Colomé
All from Isaura Saguer ColoméRelated Articles
