An under 10-minute read for change leads, programme sponsors, operational leaders, and L&D professionals who have watched a well-funded training programme quietly achieve nothing.
If you have worked on a transformation programme, you have probably seen this film before.
A major programme reaches its final phase.
The system is built. The process has been redesigned. The business case is intact.
Then someone raises the question of training.
The programme machine shifts into a familiar gear.
A training needs analysis is commissioned — or more often dusted off and updated. A provider is engaged. Content is developed. A schedule is issued. Hundreds of people attend sessions: classroom, e-learning, or a hybrid of both.
Attendance is tracked.
Evaluation forms are collected.
A satisfaction score above four out of five is quietly celebrated.
Six months later, the new system is being used to replicate the old process.
The redesigned workflow has been bypassed by teams who found it easier to keep doing what they always did. The benefits that depended on behaviour change have not materialised, and nobody can quite explain why — because the training data says it went well.
This story is not unusual.
In most organisations, it is the default outcome.
And the reason has very little to do with the quality of the training itself.
It stems from a fundamental misunderstanding of what training is actually for.
Most transformation programmes treat training as a handover activity.
Deliver the change.
Then train the business to operate it.
It feels logical. It fits neatly into programme structures. Training can be scheduled near the end of delivery, keeping costs deferred until late in the lifecycle. It creates a satisfying milestone:
Training complete.
Go-live authorised.
Programme closed.
But this model assumes something about human behaviour that is demonstrably wrong.
It assumes that knowledge transfer produces behaviour change.
That if you show people how to work differently, they will.
In reality, behaviour rarely changes because someone attended a session.
People change how they work when the environment around them — their colleagues, tools, incentives, measures and managers — makes the new behaviour easier and more rewarding than the old one.
Training can support that shift.
But it cannot create it on its own.
When training is treated as the mechanism of change rather than one component of a broader behavioural environment, it fails — reliably and expensively.
Research in organisational psychology and adult learning has pointed to the same conclusion for decades:
Only a small proportion of what is covered in formal training is consistently applied back in the workplace.
The precise percentages vary by study, but the directional message is clear: the learning transfer problem is enormous.
What predicts whether learning translates into changed behaviour is rarely the quality of the training event itself.
It is the environment people return to afterwards.
Manager reinforcement.
Opportunity to practise.
Peer norms.
Structural barriers removed.
Psychological safety to learn through mistakes.
These are not primarily L&D issues.
They are leadership and programme design issues.
No amount of investment in training quality can compensate for an operating environment that quietly pushes people back toward established habits.
Across transformation programmes, the same patterns appear often enough to be worth naming directly.
Training is scheduled dangerously close to go-live.
Programme delays compress preparation windows. Sessions become rushed. Participants arrive already overwhelmed by communications, system access issues and operational pressure.
Training becomes associated with stress rather than capability.
Retention is poor. Confidence is low. The first weeks of go-live become a survival exercise where workarounds emerge — and those workarounds often become permanent.
Training programmes are frequently designed for an imaginary “average” user.
In reality, change lands differently across roles, teams and regions.
Some people need deep process understanding.
Others need system navigation.
Some are change-ready.
Others need reassurance and repetition before they trust the new approach.
A single curriculum designed for everyone rarely works well for anyone.
Attendance is measured. Completion rates are celebrated. End-of-course assessments are passed.
Then the programme moves on.
But nobody measures whether behaviour actually changed.
Three months later, nobody checks whether the new process is being followed or whether people quietly reverted to the old way of working.
If you cannot measure behaviour change, you cannot manage it.
And if your success metrics stop at completion rates, you are measuring the wrong outcome.
If training is not the mechanism of change, what actually works?
The answer is not “better training”.
It is recognising that behaviour change must be designed into the delivery architecture of the programme itself.
Several principles consistently show up in successful transformations.
The most effective capability building happens during the change, not after it.
Involve operational teams in pilots, testing and early adoption programmes. Let people help shape the solution rather than presenting it fully formed at the end.
People who helped build the solution understand it differently.
They also become natural advocates for it.
The training plan should not end at go-live.
In reality, the 90 days after go-live are when behaviour patterns are formed.
This period should include:
Capability building must continue while the organisation is actually using the new system.
If the new process is harder than the old one, people will find workarounds.
Every workaround that becomes habitual is a piece of behaviour change that has failed.
Effective change teams focus not only on explaining the new way of working, but on removing the friction that makes the old way more attractive.
Programme sponsors often disengage at go-live.
That is precisely the moment when leadership visibility matters most.
The weeks following go-live are when resistance emerges, narratives form and habits stabilise.
Visible sponsor engagement during this period has a disproportionate impact on whether change actually lands.
There is an uncomfortable truth underneath many of these programmes.
Training failure is rarely owned.
When transformation programmes fall short of their expected benefits, the analysis usually focuses on technology, scope or governance. Training may be mentioned, but it is rarely examined with the same rigour.
Partly this is because training happens near the end of delivery, when teams are already disbanding.
Partly it is because the operational teams living with the consequences are rarely invited into programme retrospectives.
The result is predictable.
The same mistakes are repeated in the next programme.
Breaking that cycle requires behaviour change outcomes to be treated as programme delivery metrics, not just L&D metrics.
Not just “Did the training happen?”
But:
Did people actually start working differently?
Before the next transformation programme commissions its training plan, pause on a question that most programme teams never formally ask:
What would need to be true about the post-training environment for this learning to actually translate into changed behaviour?
The answer usually surfaces the real risks:
manager capability gaps
structural barriers
conflicting incentives
weak sponsorship
All the things that no amount of classroom time will fix.
And it reframes the purpose of training entirely.
Training is not the change.
It is only one part of the environment that makes change possible.
What was the last training programme that genuinely changed behaviour in your organisation?
