Our board asked for our AI strategy three meetings straight. Meanwhile our firmware team ships with coding assistants every day, nobody asked permission. I read only 37% of companies see EBIT impact from AI. That gap bugs me. I could write the top-down plan with milestones and owners. Or I could stay out of the way and let bottom-up adoption run. Which one actually closed the gap at your shop?
Or I could stay out of the way and let bottom-up adoption run.
I would vote for the plan. The structural force is local optimization, and the concrete consequence is fifteen tools with fifteen bills that never touch what the company actually sells. Each team gets faster at its own tasks while the org stays as fragmented as before, like adding lanes without fixing the interchange. A strategy is mostly a list of what not to do. No second pilot until the first moves a number. One data surface. Build versus buy decided once. Without that list the 37% gap stays permanent.
Usage moves quota, your no-list just kills weird bets. Name one number a deck ever moved?