One good workshop changes what people can do. It rarely changes what they keep doing. This programme exists for the second problem: spaced sessions, real work in the room every time, and a deliberate handover so the new way of working survives after I leave.
Shaped around the organisation — which teams go first, what gets built, and what gets measured are scoping decisions, not package contents.
The same three moves as every engagement, run at organisational depth. The methodology page documents how each session is designed; this is what the programme adds around them.
Before anyone books a room: a questionnaire, then a diagnosis of the real constraint. Is it knowledge, relevance, habit, or justified caution? Each needs a different response, and getting this wrong is how organisations end up with training nobody uses.
The output is a design: which teams go first, which of their real tasks come into the room, and what the programme should change. If training is not the right answer for a team, I say so here — before it costs you anything.
The training itself, in spaced sessions. People work on their own files for most of every session — try, inspect, judge, refine — because the aim is not a polished first attempt; it is the judgement to improve and safely use what comes back.
Between sessions, people use the workflows on live work. What broke, what stuck and what surprised them comes back into the next room. That gap is not downtime — it's where the adoption actually happens.
The phase most training skips, and the reason this programme exists. Every workflow that survives the Build phase gets an owner inside your organisation — a named person responsible for keeping it alive, improving it and onboarding the next colleague into it.
Adoption gets measured against what was agreed in Listen: who changed how they work, which workflows are in weekly use, and what quietly went back to the old way — because something always does, and finding it early is the point.
The handover standard: a workflow is not finished because it worked once. It needs named inputs, repeatable steps, an owner, a review gate and a next-use date. A colleague should be able to run it without the trainer in the room. Anything that cannot pass that test stays in development rather than being presented as adoption.
Choose the programme when the goal is changed behaviour across teams and someone will be accountable for it in six months. If one team needs a first supervised build, a single workshop is cheaper and honest about its limit: capability changes before habit does.
The return is ordinary arithmetic, not a multiplier — the methodology page shows the worked example and what quietly breaks it.
Thirty minutes, no deck. Tell me what your teams are already trying with AI, and I'll tell you whether a programme is the right response — or whether something smaller gets you there.