Executive Decision Guide
Strategy First,
or Pilot First?
A ten-minute self-assessment for leaders who cannot yet tell whether AI activity is producing progress.
Inside
- The one question that can settle it immediately
- A 12-point diagnostic across direction, selection, and execution
- A practical recommendation matched to the layer you are missing
- Two executive questions that reveal when activity is being mistaken for progress
A defensible next move: build direction, strengthen selection, or learn through bounded pilots, without mistaking activity for progress.
The question that overrides the rest.
Step 1 of the guide is a single gate. Answer it before anything else. If the answer is yes, the sequence is already decided, and the rest of the guide becomes confirmation rather than diagnosis.
Would this AI system touch a decision that a regulator, auditor, or court could later ask you to justify?
Step 1, page 4. Examples in the guide include credit and lending decisions, clinical or diagnostic support, hiring and promotion, safety-critical operations, and anything producing a record you must be able to defend.
If the gate does not apply, the guide moves to a 12-point diagnostic across three layers: direction, selection, and execution. The layer you are missing determines the recommendation, which is why two organizations with identical pilot counts can need opposite next moves.