Lost in translation: Why global L&D programmes fail the people who need them most

What separates a global L&D programme that fails from one that succeeds? 

Only a quarter of companies say their training programmes measurably improve performance (according to McKinsey), which means the other three-quarters are spending money on training and still can't prove it works.

You might be familiar with the feeling of planning a programme that looked like a success on paper, but fell apart once it was rolled out. You find it hard to see what went wrong – you ticked all the right boxes, after all. Everything should be in order, shouldn’t it?

When the cracks begin to show themselves, they might look something like this: A manufacturing worker in England is reading compliance training that was written for a desk-based employee in Chicago, or a call centre agent in Lagos is doing a customer service module based around English behavioural norms. Both of them are extremely confused. It’s like a long, drawn-out game of telephone where everyone is speaking a different language and nobody is learning anything from it. 

Translation is a logistics problem, but localisation is a people problem. L&D has been solving the wrong one for too long. 

‍

The translation map

The L&D industry has treated language as the primary barrier to global learning, and spent enormous effort and budget solving it. Of course, language is a barrier to global learning – but it’s not the only one. There are important pieces of context it simply can’t account for. 

This is why elearning translation alone – however thorough – will always leave something behind.

• Regulatory and compliance differences between markets that the content doesn’t reflect: This is the most important sticking point. As any L&D professional will be aware, compliance is absolutely foundational to any learning strategy – but in order to work, it has to make sense to each individual person completing it. 
  ‍
• Cultural norms: This is a bigger piece of the puzzle than you might initially think. Training designed for your market but deployed outside of it will always feel alien, and in some cases, actively unsafe. In a lot of cultures, questioning a manager or admitting a gap in your skillset carries real professional risk.
  ‍
• Role-specific nuance: When one piece of content gets stretched across multiple different markets, customer expectations, and regulatory environments, it starts to snag and snap. It’s too generic to be actionable, and too specific to the originating market to feel relevant anywhere else. This is one of the core reasons cross-cultural training demands more than a one-size-fits-all approach.
  ‍
• References, metaphors, and examples: If you trotted out a Come Dine with Me reference to an American, do you think they’d understand it? Something that feels like a standard part of our own cultural lexicon is a completely foreign language to someone else, and the same goes for references used in learning. Ultimately, it’s a small but persistent sign that this content wasn’t created with them in mind. 
  ‍

Why the standard fixes don’t work

What are the repeated solutions your organisation keeps returning to, expecting different results each time? We’re willing to bet they look something like these: 

• Regional content teams: These are expensive, slow to scale, and still dependent on a brief that originated centrally. The local team adapts the edges, but the core stays the same. 
  ‍

• Learner feedback loops: They’re valuable in theory, but in practice, global employees tend not to flag training that felt off; instead they just disengage completely. 
  ‍

• Market-specific versions: The right instinct, but the resourcing requirement makes coverage impossible for most businesses. You end up with thorough localisation for your three biggest markets and token translation for the rest.

The common thread through all of these approaches is that they treat localisation as a content production challenge, which makes it expensive and slow enough that most organisations do it partially at best.

‍

The real cost of getting global L&D programmes wrong 

When training falls flat, people simply don’t engage with it. Or worse, they listlessly click through modules that make absolutely no impression on them, close the tab, and instantly forget what they just saw. 

The consequences of this are unevenly distributed, and it’s usually employees in markets outside of your company’s headquarters that will bear the brunt. They’re the ones who will receive learning that was designed with a completely different workforce in mind. It’s a workforce entirely unlike theirs, with different regulations and cultural expectations. That disconnect then has a knock-on effect on things like skills gaps and compliance failures – two things no L&D team wants to be answering for. 

Outside of these practical failures, it also sends the wrong message to your global workforce. It’s as though you’re saying, “Yes, we know you’re there, but we don’t really care enough about you to prioritise your specific development needs.” 

‍

What real L&D localisation needs to succeed, and how AI can help

AI can go beyond simply translating something, and do the important step of contextualising it. That gap between translating a module and actually adapting it for a wholly different workforce is exactly the kind of thing LLMs were built to close, and it’s where global L&D strategy has been stuck for years. 

In practice, that looks like: 

• Generating market-specific scenarios rather than translating generic ones
• Adapting tone and register to cultural norms around authority and communication
• Flagging where underlying assumptions about regulation or organisational structure don’t travel, before those assumptions undermine the learning

The more significant shift, though, is one of scale. L&D teams have always known what good localisation looks like — and what multilingual elearning actually requires. The blocker has been resources: building out adapted content for every market, role and context simply wasn't feasible. AI changes that. You define the parameters once – the market context, the regulatory environment, the cultural norms – and apply them systematically across your content library rather than manually, piece by piece.

For organisations evaluating a multilingual LMS or comparing global training solutions, the ability to localise at this level – not just translate – is worth treating as a non-negotiable.

Thrive's AI Content Creator is designed for exactly this kind of output. Using OpenAI and Google Gemini, it instantly turns prompts, PDFs and rough ideas into structured learning content like videos, quizzes, and written copy. For global compliance training, that means compliance documents and policy updates can be converted into localised, market-ready learning content without a production backlog. And because content is generated fresh each time, it can be adapted to reflect the specific context of each market rather than cloned from a master version.

Kiki, Thrive's AI agent, sits alongside that. Where the AI Content Creator handles production, Kiki handles the layer underneath – surfacing skills gaps by team or region, flagging who's behind on compliance before it becomes an audit risk, and coaching learners through content in the flow of work rather than routing them back to a module. For a deskless workforce in a market where L&D resources are thin on the ground, that's what it looks like when learning reaches the people it was meant for.

Global L&D has always had the right ambition: to develop every person in the organisation, wherever they are, in a way that's actually useful to them. The tools are finally catching up.

‍

Beyond translation

Translation was always the easy part. What the industry hasn't solved, until recently, is everything underneath – the layer that determines whether content actually lands for the person reading it.

That's what real localisation requires, and it's what's been out of reach for most global L&D teams. AI makes it possible to deliver the version of global L&D that organisations have always wanted.

See how Thrive helps global teams get learning that actually lands. Book a demo.