Every AI new-revenue initiative dies in the same room. The CFO’s business case.
The pattern is always the same. Product or GTM proposes an AI-enabled offering. The commercial team builds the model. Finance stress-tests the assumptions. The forecast does not stand up because you cannot forecast a new revenue stream with the same certainty as an efficiency saving. The board sees the numbers, hears the doubt, and defers. Nine months later the initiative is dead and someone else has shipped it.
The CFO is not the problem. The business case is the problem. It is being asked to do a job it cannot do, which is to forecast paying customers for a category that does not exist yet.
The fix is not a better model. The fix is a different funding shape.
Fund it in stages. Frame, Probe, Pilot, Scale. Each stage has its own budget, its own team size, its own evidence gate and its own kill criterion. You do not ask the CFO to approve a five-hundred-thousand-pound Pilot on faith. You ask the CFO to approve a fifty-thousand-pound Probe with a written kill criterion on day one. That is a business case the CFO can defend to the board.
The point of the staged model is not that it is more disciplined. The point is that it is fundable. It moves the decision from “back the whole thing” to “back the next fifty thousand”. That question the CFO can answer.
Ortent published a diagnostic on this last month. It sits at ortent.co/tools/reinventors.
Three stages. Pick which of four patterns your idea fits. Score six axes on a 1 to 5 scale. Get a starting stage on the funding ladder and a draft brief for your CFO. Ten minutes end to end. A Claude project if you would rather run it conversationally.
The paper walks through a case study of a real CEO and CFO who broke a stalemate on a 1.6 million-pound investment by running the Playbook and shipping the offering for 445 thousand instead. Not the whole thing. The next stage of the whole thing.
If you have an AI new-revenue idea that has not made it out of the business case, this is worth ten minutes.
One warning. The Playbook does not work if the CFO does not run it with you. Score independently. Compare. The axes where you scored 5 and the CFO scored 2 are the ones that need evidence, not argument.