I posted yesterday about the shift from studio to system in L&D.

That post diagnosed the problem; this post attempts to respond to it.

If the organisation is already producing learning without you, the answer is not to compete with it, ignore it or drag everything back into the formal offer. The answer is to help useful knowledge move through the organisation more safely, clearly and effectively.

Six practical shifts you might consider:

1. From commissioning editor to knowledge shepherd
Stop waiting for every request to become a course. Look at where knowledge is already moving through Teams chats, shared folders, peer advice, manager shortcuts and AI summaries. What is useful? What is wrong? What is out of date? What needs pinning, correcting or removing?

2. From Hollywood director to creator coach
Not every useful resource needs high production values. Sometimes the best answer is a short explanation, screen recording, worked example, checklist or common mistake from someone close to the work. L&D’s job is to help people capture that knowledge clearly and safely.

3. From prime-time programming to point-of-need access
Stop thinking only in launches, campaigns and calendar slots. Ask where the need actually appears: at the point of decision, error, handover or friction. This is where “just for me, just enough” learning earns its place: not as content convenience but as support that meets the work when the work creates the need.

4. From box office numbers to usage signals
Attendance and completion still tell you something, but they are weak signals of usefulness and even weaker signals of validity. Look at what people search for but cannot find, which resources are reused, which questions keep coming up, and where informal workarounds are spreading.

5. From single-channel control to managed distribution
Not everything belongs in the LMS. Some learning needs to sit in Teams, workflow tools, searchable guidance, manager routines or short peer resources. The question is not “How do we get people into our platform?” It is “Where does this need to live so people can use it properly?”

6. From scriptwriting to AI guardrails
Employees are already using AI to summarise, explain, draft and troubleshoot. L&D does not need to own every answer, but it does need to help set standards: what sources to use, what claims to check, what prompts are safe, and where AI creates risk.

This is what moving from studio to system looks like in practice.

Less focus on producing everything centrally and more focus on helping people find, test, share and use knowledge well enough to improve the work.