The Diagnostic Discipline

You tell clients you run a rigorous process. Then you sit in the car afterward wondering what you are actually going to do next. That gap, between the process you describe in the room and the improvisation that follows, is what this paper is about.

The market has quietly conflated two different things: using AI to run an audit, and running a rigorous diagnostic. They sound identical in a sales call. They are not the same act. Dropping a client's material into a general model and prompting it to "do an audit" returns one ungrounded answer that cites nothing, compounds nothing, and falls apart the moment scope grows.

The record is blunt about where that leads. More than 80 percent of AI projects fail, and the leading cause is not the model. It is people, framing, and readiness (RAND, 2024). Nearly everyone has adopted AI; almost no one has scaled it (McKinsey, 2025). The gap between activity and results is judgment, and judgment is what you actually sell.

The paper makes one core claim and defends it: a rigorous diagnostic and an improvised one are different in kind, not in quality. It derives the five principles and the five-phase methodology from first principles, grounds them in more than 200 conversations across the market, and argues that AI raises the value of the method instead of retiring it.

The full 34-page paper is behind this one field.