Why generic data & AI literacy programmes fail at scale

Most enterprise data & AI literacy programmes that are managed by external vendors have the same problem: the content is fine, there are competent facilitators in the room and the slides look professional.

Yet, after getting people to take their precious time to attend, there’s no clear road from learning to application.

This doesn’t point to their motivation or level of engagement. The big reason why this model doesn’t work is a lack of relevance. When someone spends two hours learning data concepts through examples that have nothing to do with their job, their industry, or their actual systems, their mind has to work overtime to translate what they’ve learned on their own. And that translation is cognitive overhead they can't afford when they're back at their desk with real work in front of them.

Generic training teaches concepts while contextualised programmes build capability.

One unfamiliar thing at a time

This is the principle that guides how we build every customised programme at Data Literacy Academy.

When someone is learning something genuinely new, like how to interrogate data quality, how to structure a data request, how to challenge an insight before acting on it, that's already asking a lot. The last thing they need is to be doing that whilst simultaneously decoding fictional scenarios, unfamiliar job titles, and systems they've never heard of.

Our approach is simple: make sure the only unfamiliar element is the data concept itself being taught. The language, the examples, the systems, the workflows: all of it should feel immediately recognisable. This results in people learning within their context from the outset.

The effect is measurable:

• Reduced cognitive load
• Faster comprehension
• Direct application to daily work, instead of theoretical future application

What this looks like in practice

The gap between generic and contextualised is most visible in the detail.

Generic training says: "Understand who owns your data."

Contextualised training names their business’ governance structure, distinguishes the relevant roles, explains the escalation pathway, and connects it to the system of record employees use every day.

Generic training says: "Name your files clearly and consistently."

Contextualised training references the organisation's own naming conventions, links to internal policy, and explains the downstream consequences when those conventions aren't followed, including the real business risks that people will immediately recognise.

Generic training says: "Ensure data accuracy in your reports."

Contextualised training identifies the high-stakes output formats that matter in this organisation, explains what happens when accuracy fails, and connects data quality to accountability at a leadership level.

These aren't cosmetic differences. They're the difference between a learner who nods along and one who actually changes how they work after absorbing the learning.

The compounding value of continuity

Now let’s consider another essential point. The value of a customised programme isn't just the first cohort. It builds as you work your way through your organisation.

Educators develop deeper contextual familiarity. Feedback accumulates across cohorts, enabling pattern recognition and relationships with internal stakeholders mature, improving how quickly refinements get validated. The content itself becomes more precise, more relevant, more impactful, because it's being sharpened against real-world evidence, not kept generic enough to sell to anyone.

Each cohort that goes through a well-customised programme makes the next one better. That compound effect is significant. And it resets to zero if you switch provider.

Strategic alignment isn't a feature. It's the point.

The reason executives commission data and AI literacy programmes isn't to improve survey scores but to change how the organisation makes decisions.

That goal requires content that connects individual behaviour to organisational priorities, explicitly, not implicitly. Learners need to understand why their data quality choices affect teams downstream. They need to see how the metrics they track connect to the strategic goals their leadership is accountable for. And they need to practise on scenarios that reflect the actual decisions they face.

This can't be retrofitted onto a generic programme, it has to be built in.

The organisations that get lasting ROI from data and AI literacy investment are the ones that treat programme design as a strategic decision instead of a procurement one. The question isn't which provider has the best off-the-shelf catalogue. You need to look critically at which partner will build something that makes the learning stick.

A final note on scope

Customisation exists on a spectrum. Deep contextualisation (building entirely bespoke modules, integrating enterprise-specific governance frameworks, developing scenarios from the ground up) is reserved for engagements where the strategic stakes warrant it. That level of investment compounds in value over time, but it requires the right scale and commitment to make sense.

Yet for most enterprise programmes, the principle still holds: the more familiar the context, the faster the capability develops. Even targeted contextualisation, with the right terminology, systems references and organisational examples, meaningfully outperforms the generic alternative.

The goal isn't training or tracking completion scores. It's delivering behaviour change at scale.

‍