The Quality Standard for Professional Learning

Have you successfully measured the delivery of your training yet had no idea if any learning had actually occurred? This one's for you.

Come with me for a moment. We're in a wash-up meeting for a big leadership programme we’ve designed. The Head of L&D is beaming. She puts a slide up showing a 97% completion rate and a Net Promoter Score in the high seventies. By any conventional measure, a resounding success.

Then she asks for my thoughts. What would you reply if this was you? Personally I'd reply with another question.

“Has anyone seen the managers behaving differently since they did it?”

I imagine the air in the room would change. The beaming stops. Why? Because the likelihood is that this is a question nobody has the data to answer. And deep down, they knew it was the only one that truly mattered.

Do you relate to this?

The Seductive Lie of the Dashboard

This isn't a unique story. It’s the default story for much of corporate learning and development. We are addicted to metrics that measure the process, not the outcome. Completion rates, learner satisfaction, time spent on platform. The things a Learning Management System can easily spit out into a tidy report. Things that look good on a slide for senior leadership.

There’s a reason for this. These metrics are clean. They are quantifiable. They provide a sense of control and accomplishment. And asking about actual behaviour change? That’s messy. It’s qualitative, expensive to track, and it often surfaces uncomfortable truths – like the fact that your £50,000 investment in a course didn't change a thing.

So we collectively agree to accept the proxy. We pretend that a completed module is equivalent to a learned skill. That a happy learner is an effective one.

It’s a comfortable lie, but it’s the single biggest reason why so much professional training fails to land. We’re optimising for the wrong thing. It’s like a golf pro judging his coaching success by how many pupils finish a bucket of balls, rather than whether their handicap ever comes down. The activity is not the result.

Everyone in L&D knows Kirkpatrick's Four Levels of Evaluation. We can all talk a good game about aiming for Level 3 (Behaviour) and Level 4 (Results). But look at the dashboards. Look at the annual reports. Look at the metrics that determine bonuses. We’re still living at Level 1 (Reaction) and Level 2 (Learning/Recall).

From Knowing to Doing

What does ‘professional learning that actually works’ mean, then? It’s simple. It means the person who took the course now does something different, and better, when they get back to their desk.

That’s it. That’s the standard.

A financial adviser who can explain the risks of a new investment product more clearly to a nervous client. A care worker who spots the subtle signs of patient distress they might have missed before. A software developer who writes more secure code not because the policy document tells them to, but because they’ve internalised the principles.

This is about performance, not just knowledge. And you can’t build for performance by just giving people information.

We've all seen this play out, haven't we? Let's take compliance training. An organisation spends a fortune on an anti-money laundering course. It's polished, it's interactive, it has a quiz at the end. Everyone passes and gets a certificate. Six months later, the regulator runs a spot check and finds the exact same procedural gaps they found two years earlier. The training had a 100% completion rate. The training was also a complete failure.

Why? Because the assessment tested recall of the rules, not the judgement required to apply them to a messy, ambiguous, real-world situation. The staff knew the what. They had not practised the how.

Designing for behaviour change means focusing on practice. It means creating authentic scenarios where people have to make decisions. It means giving them a safe place to fail. It means your assessment doesn't ask “Which of these is the correct procedure?” but “Here is a complex situation with conflicting information. What do you do now?”

Historically, this has been difficult and expensive, which is why it's rare. Creating one good, specific scenario is hard. Creating fifty, so that learners don’t all just memorise the 'right' answer to the same one, is an order of magnitude harder. It's far easier to write a few multiple-choice questions.

This is one of the few areas where I’m genuinely interested in the role of AI. Not to replace the craft of the designer, but to augment it. To help a skilled professional create a dozen varied, plausible, specific scenarios in the time it used to take to create one. To help scale the part of the design process that actually builds for performance, not just for recall.

It’s about using the technology to finally bridge the gap between knowing and doing.

The real quality standard for our work isn’t found on a dashboard. It’s visible in the confident and competent actions of the people we’ve trained, long after the course is complete.