Executive boards are lighting $10M on fire buying AI they don't understand.
Last month, I watched a CEO approve a "revolutionary personalization platform" that would've nuked their unit economics. How'd I know? I'd already stress-tested that exact tool on platforms with millions of DAUs. Overpriced junk with a Silicon Valley bow.
Here's what's driving me crazy: 90% of enterprise AI purchases are expensive theater. Companies buying solutions to problems they can't define, from vendors whose biggest win is raising a Series B.
The executives actually winning? They spot BS because they've built the real thing.
After 20 years scaling products across platforms from startups to 50M+ DAUs and $500MM+ ARR, the pattern is obvious. Leaders crushing it work with people who've been in the trenches.
Here's what separates winners from the walking dead:
Two-velocity organization - Velocity squads ship AI features biweekly. Platform crews build infrastructure to scale those innovations. When competition launches breakthrough capabilities every other week, quarterly roadmaps are death certificates.
The reset question - Before any major AI investment: "If we launched today, how would we solve this?" Most legacy approaches crumble. Thriving companies rebuild core assumptions, not bolt AI onto old processes.
Skills without silos - Your best PMs prototype with no-code tools. Top engineers understand conversion psychology. Designers grasp technical constraints. When everyone speaks multiple languages, you ship solutions instead of compromises.
Build, don't brief - I've worked with teams building weekend prototypes that destroy million-dollar "solutions." Stop PowerPointing AI strategy. Build working demos. The gap between vendor promises and reality becomes crystal clear.
Curiosity over compliance - Teams that break through don't just "evaluate AI tools" - they obsess. They tear apart model releases, reverse-engineer competitors, constantly ask "what if we tried this differently?"
The uncomfortable truth: You can't evaluate technology you don't understand. You definitely can't lead a transformation you haven't lived through.
Companies dominating the next decade won't have the biggest AI budgets. They'll be led by people who know venture-backed snake oil from actual innovation.
Everyone else gets acquired. Or becomes Harvard case studies on how established companies got disrupted by startups with 1/10th their resources but 10x their execution speed.