Some years ago, a large corporate organisation contracted me to design a programme for SMEs in its value chain. Similar programmes were already being run in some of its sister countries, so from the client’s perspective the assignment probably seemed straightforward: design the course and get on with it.

Naturally, one of my first responses was that I would need to conduct a needs assessment.

That was when things became interesting.

As I began asking questions about the SMEs, their current capability, the challenges they were facing, what the organisation wanted them to do differently and what success would look like, I could sense the head of the department becoming increasingly uncomfortable. Eventually, his frustration became clear. Trainers in sister countries already had the course, he said, while here I was asking questions and, in his view, wasting everybody’s time.

He more or less suggested that perhaps I did not know my job.

I remember thinking how ironic that was.

I was not asking questions because I did not know how to design the course. I was asking questions because I did.

That experience has stayed with me because it captures one of the biggest tensions in Learning and Development. Clients often want the course quickly. Good learning design first wants to understand the learning.

And those two things are not always moving at the same speed.

A course is a container. It has a title, duration, content, activities and materials. Learning is the change we are trying to create in the person who experiences that container and, ideally, in what that person can subsequently do.

That distinction matters because we often start designing in the wrong place.

Someone says, “We need a two-day leadership programme,” and almost immediately the machinery begins. We think about modules, models, videos, case studies, activities, handouts and PowerPoint.

Before long, we may have 70 beautiful slides and still be unclear about the most important question:

What should these people be able to do differently when the programme is over?

For me, that is where design should begin.

If managers are struggling to give performance feedback, for example, the answer is not automatically to load a course with definitions, models and ten tips for effective feedback. If the real performance requirement is that managers should be able to conduct difficult feedback conversations with employees who become defensive, then the design must give them an opportunity to practise exactly that.

They need to rehearse the conversation, get it wrong, receive feedback and try again.

Because the difficult part does not happen when a facilitator asks, “Why is feedback important?”

Most managers can answer that.

The difficult part happens when the employee folds their arms, looks offended and says, “So you think I am incompetent?”

That is the moment we should be designing for.

This is why good instructional design starts with performance and works backwards. What must people be able to do? What knowledge do they need? What practice will help? Where are they likely to struggle? What feedback will help them improve? What should happen during the learning, and what must happen later in the workplace if the learning is to stick?

ATD describes instructional design as a systematic process for creating learning experiences that support real-world skill acquisition, and it explicitly places needs assessment before course design. That is important because it reminds us that asking questions is not delaying the work.

The questions are part of the work.

There is another trap we need to watch: our love affair with content.

Subject-matter experts naturally know a great deal and often want learners to know a great deal too. I understand the instinct. When you care about your subject, everything feels important.

But the designer has to ask: important for what?

What does the learner actually need in order to perform the task? What is useful background? What is nice to know? What could be provided as a resource later? What are we including because the learner needs it, and what are we including because the expert enjoys teaching it?

Those are not always the same thing.

I sometimes look at programme designs and think the poor participant is being treated as a storage facility. More slides, more models, more definitions, more frameworks. We keep pouring information in and then seem surprised when very little comes back out.

Learning requires more than exposure. It requires retrieval, practice, feedback and application.

A major meta-analysis by Steven Pan and Timothy Rickard, published in Psychological Bulletin, reviewed 122 experiments involving more than 10,000 participants and found that retrieval practice could support transfer beyond the original material, including to application and inference tasks.

That matters because learning is not demonstrated by how much content was covered.

It is demonstrated by what the learner can now do with it.

The same applies to activities.

I love a lively learning environment, but engagement and learning are not the same thing. An activity should not be included simply because it gets people moving, makes them laugh or produces a lot of noise in the room.

Every activity should have a job.

Does it help participants practise? Does it reveal an assumption? Does it allow them to experience a concept? Does it generate material for reflection? Does it produce feedback?

If the answer is no, then perhaps the activity is entertaining rather than instructional.

There is nothing wrong with entertainment.

I just would not invoice it as learning design. 😄

And then there is the part we often forget: what happens after the programme.

We can design every minute of a two-day workshop and still give very little thought to what happens when the participant returns to work. Yet if a manager learns to coach but never coaches anyone, if an employee learns a new sales technique but gets no chance to practise it, or if a leader learns to delegate but works in a system that rewards control, then much of the design has been left unfinished.

The workplace is not simply where learners return after learning.

It is part of the learning design.

This means thinking about manager support, follow-up practice, workplace assignments, coaching, job aids and opportunities to apply new skills while the learning is still fresh.

And sometimes, after asking all the right questions, the answer may be that a course is not what people need at all.

Perhaps the issue is a broken process. Perhaps employees already know what to do but lack the authority to do it. Perhaps a job aid, coaching conversation, demonstration or redesigned workflow would solve the problem more effectively.

Good instructional design is therefore not about proving how clever we are at building courses.

It is about identifying the most appropriate learning or performance solution.

That matters even more now because AI can create courses faster than ever. ATD's 2025 research on AI in Instructional Design reported that 80% of instructional designers surveyed were using AI tools. That is exciting, but it also introduces a new possibility:

We can now produce the wrong course much faster.

If the diagnosis is poor, AI can help us build a beautifully structured response to the wrong problem. If the learner need is misunderstood, AI can generate more content for the wrong audience. If the workplace will not support application, AI can produce excellent follow-up materials for learning that still goes nowhere.

When content becomes easier and cheaper to produce, professional judgement becomes more valuable.

That is why I think the future of instructional design is not about who can produce the most polished course fastest. It is about who can ask the best questions, diagnose the real need, design meaningful practice, anticipate transfer barriers and decide what learning solution actually fits.

Which brings me back to that corporate client.

The sister countries may already have had a course.

But my job was not to prove that I could copy one quickly.

My job was to understand whether that course made sense for the SMEs in front of us.

“Do not design what you want to teach. Design what people need to learn, practise and eventually perform.” Margaret Jackson

So the next time someone asks you to design a course, resist the urge to open PowerPoint first.

Start with the learner. Start with the performance gap. Start with the business need. Ask what must change and what evidence will tell you that it has changed.

Then decide whether the answer is a workshop, digital learning, coaching, workplace practice, a job aid or some intelligent combination of them.

Because our job is not to manufacture courses.

It is to design learning.