The Course Is Not the Product

Most learning projects begin with a request for a course. Someone has noticed a gap — onboarding is slow, a product launch is confusing, a team keeps repeating the same mistake — and the organizational reflex is to produce material. A deck, a module, a curriculum. Something visible.

The reflex is understandable. Content is the part of learning that can be scoped, budgeted and shipped. It has a start and an end. It looks like progress. But content is rarely the thing that changes how work gets done.

The product is a change in behaviour

If we are honest about what an organization is buying, it is not a course. It is a different outcome in the work: fewer errors, faster ramp-up, better conversations with customers, more confident decisions under pressure. The course is one instrument among several that might produce that outcome.

This distinction matters because it changes what counts as success. A course is successful when it is finished. A capability shift is successful when someone does something differently a month later, in a situation nobody scripted.

Completion tells us that someone reached the end of a sequence. It says almost nothing about whether they are more capable in the work.

What content can and cannot do

Content is very good at a small number of things:

• making shared language available;
• explaining a model or a procedure clearly;
• providing a reference people can return to;
• setting a standard for what "good" looks like.

Content is poor at other things that matter just as much:

• creating the pressure that produces judgment;
• rehearsing a difficult conversation;
• giving feedback on an attempt;
• making a new behaviour socially acceptable in a team.

Those require conditions, not material. Someone has to design the practice, the feedback, the manager's involvement, the moment where the new behaviour is expected and supported.

Designing conditions instead of modules

A more useful starting question is not "what should this course cover?" but "what would have to be true for someone to work differently?"

That question usually surfaces four things:

1. A clear picture of the target behaviour. Not a topic list — a description of what a capable person does in a specific situation.
2. A place to practise safely. Simulation, role-play, a real task with a safety net, a peer review, a low-stakes first attempt.
3. Feedback that arrives quickly. From a system, a peer, a coach, or a manager who knows what to look for.
4. A working environment that permits the change. If the incentive, the tooling or the workload contradicts the training, the training loses.

Content still appears in this design — but it appears in service of practice, not as the deliverable itself.

AI has made this more urgent

Generation cost has collapsed. Producing plausible learning material is now trivial, which means material is no longer scarce and therefore no longer valuable on its own. What remains scarce is design judgment: knowing what to build, what to leave out, where practice belongs, and how to tell whether anything actually changed.

This is good news for anyone who cares about capability. The volume conversation is over. The remaining conversation is about usefulness.

A practical reframe

When the next request for a course arrives, it is worth spending an hour on three questions before scoping anything:

• What decision or action are we trying to improve?
• Where will people practise it, and who will respond to their attempt?
• How will we know, in the work, whether it improved?

If those three questions have answers, the content almost designs itself, and it tends to be shorter than expected. If they do not have answers, more content will not help — it will only make the gap harder to see.

The course is not the product. The product is a person who can do the work.