Why most training doesn't change anything - and what actually does

Every year, businesses invest significant time and money in training their people. Workshops get booked, facilitators get briefed, calendar invites go out. People show up, engage reasonably well, and leave with good intentions.

And then, a few weeks later, not very much has changed.

If you've ever commissioned a training programme and quietly wondered why the impact didn't match the investment, you're not alone. In my experience, it almost never comes down to the quality of the training itself. The problem runs deeper than that.

We confuse learning with behaviour change

There's an important distinction between someone knowing something and someone doing something differently as a result. Training can achieve the first. The second requires something more.

This is a distinction occupational psychology has been making for decades. Knowledge transfer and behaviour change are not the same thing — and treating them as if they are is the reason so many well-designed programmes fail to deliver.

A useful way to think about this comes from the Kirkpatrick model, a framework for evaluating learning impact. It describes four levels: how people felt about the training, what they learned, how their behaviour changed, and what impact that had on the business. Most organisations only ever measure the first — a feedback form at the end — and assume the rest has followed. It rarely has.

What gets in the way

If the training was good and people were engaged, why doesn't behaviour change? A few reasons come up again and again.

No reinforcement. Most of what we learn in a single session is forgotten within days without practice and repetition. One workshop is not enough to change a habit.

The environment doesn't support it. People change behaviour in the context of their team, their manager, their day-to-day pressures. If those things aren't set up to support new ways of working, the training fades — however good it was.

No clear picture of what change looks like. If people leave without a specific, concrete idea of what they'll do differently, the intention stays abstract. Good intentions are not the same as changed behaviour.

AI is a perfect example

Right now, many organisations are investing heavily in AI training — and a lot of them are going to be disappointed with the results. Not because the training was bad, but because training alone was never going to be enough.

Getting people to genuinely change how they work requires reinforcement, a supportive environment, a clear picture of what good looks like, and accountability over time. The organisations getting AI adoption right aren't necessarily the ones with the best training content. They're the ones that have thought carefully about the conditions around the training — how managers are reinforcing new behaviours, how performance conversations connect to it, how progress is measured in terms of what people actually do, not just what they've sat through.

That's a learning culture problem, not a training problem.

What actually works

Training works best when the pieces around it are also in place — what happens before (setting clear expectations), during (building in reflection and application), and after (reinforcement, coaching, opportunities to practise).

It means equipping managers to keep the learning alive. It means defining what change actually looks like in behavioural terms. And it means measuring the right things: not just satisfaction scores, but evidence that behaviour has genuinely shifted.

This is harder than booking a workshop. But it's the only approach that reliably moves the needle.