Insights.

Thoughts on innovation, leadership & technology

• The Pilot Worked. That Was the Problem.The slide says “Phase 2 — roll out to twelve sites.” I have watched that slide bury more good work than any budget cut — because a pilot that works at one site proves the solution, not the transfer. What actually made it work never makes it into the handover doc. With a small flip you can feel.August 9, 2026 · 7 min read
• Not Everything Needs AIWe've started reaching for a model the way we used to reach for Google — by reflex. But much of what we hand to AI is work a plain rule does faster, cheaper, and more reliably; in one study a regex matched an LLM at around 28,000x the speed. The scarce skill was never using AI. It's knowing what doesn't need it. With a small switch you can feel.June 20, 2026 · 6 min read
• Stop Choosing Between Efficiency and EmpathyThe oldest turf war in how work gets done: the efficiency people (Lean, process, metrics) against the empathy people (Design Thinking, the user, the workshop), each certain the other is the problem. I was trained by both camps — and the secret neither tells you is that there was never a choice. Lean’s own foundation is two pillars, one of them Respect for People; Design Thinking’s own model is three lenses, two of them feasibility and viability. Each tribe amputated half its own doctrine. Why the split persists, where each half fails alone — and a small thing to try.June 20, 2026 · 7 min read
• Cascades: The Shape of Everything That ScalesFrom the shopfloor to the boardroom, from Lean huddles to agile Flight Levels, I kept meeting the same move — and it turns out your lungs, the internet and every army use it too. Cascades: break the whole into nested levels, push each decision to the lowest one that can make it, let only the exception travel. Why it's how complexity gets tamed — and how it rots into bureaucracy. With a small thing to try.June 13, 2026 · 8 min read
• AI Made Building Free. Taste Is the New Bottleneck.Andrej Karpathy can build an app in an hour, in a language he doesn't know. When anyone can build anything, why is so much of it still worthless? Because building was never the hard part — and now that it's free, the one scarce thing left is taste: knowing what's worth building, what "good" is, and what to cut. With a small test you can feel.June 6, 2026 · 7 min read
• When "Fail Fast" Is Malpractice"Move fast and break things" is the best advice in the world — until the thing you'd break is a patient, a balance sheet, or a number nobody can take back. Why "fail fast" is only half a sentence, what Bezos's one-way doors have to do with it, and how the best high-stakes teams keep the speed without the recklessness. With a small experiment you can feel.May 30, 2026 · 7 min read
• Your Innovation Playbook Doesn't Survive the FlightI've run the same workshop in Berlin and Jakarta, same slides, and watched it electrify one room and flatline the other. The method travels; the unwritten rules that make it work don't. What splitting my life between two very different rooms taught me about why your facilitation playbook breaks at the border — with a small experiment you can feel.May 23, 2026 · 7 min read
• A Budget Is a Bet, Not a PlanMost companies decide their budget once a year, top-down, behind a closed door — and quietly anchor their whole strategy to last year. What agile budgeting really means, why it decides whether you can innovate, and how a game once turned it around. With a small experiment you can feel for yourself.May 16, 2026 · 6 min read
• The Plug Nobody PullsSome projects get worked on for months — even years — after everyone privately knows they're dead, because nobody wants to be the one to pull the plug. Why that happens, why it quietly poisons an innovation culture, and how to make stopping safe. With a small experiment you can feel for yourself.May 9, 2026 · 7 min read
• Micropolitics: When It Stops Being About the IdeaThe best idea in the room rarely loses to a better idea. It loses to someone's quiet calculation of what it would cost them. That calculation has a name — micropolitics — and it is the most underestimated reason good ideas never happen. Here is how to see it, and why the fix is not better people but a better game.May 2, 2026 · 7 min read
• The Skill Isn't Using AI — It's Reading the MomentThree weeks, two continents, two innovation sprints — one in aviation, one in MedTech. One taught me to build fast; the other, to slow down and stay in the conversation. The real skill with AI isn't using it. It's reading the moment.April 25, 2026 · 4 min read
• Shikake: Designing the Trigger, Not the MandateShikake: a Japanese idea for designing cues people are glad to follow — no mandate, no penalty, just an invitation they can't resist. Why it's really a lesson about design thinking, and why understanding behavior has to come before designing any solution.April 18, 2026 · 7 min read
• What Learning Bahasa Indonesia Taught Me About CuriosityFor 15 years I've told people mistakes are good. Now I'm learning a language — and finally understanding what that actually feels like.April 11, 2026 · 6 min read
• Stop Delivering RecommendationsThe gap between insight and implementation is where good ideas die. It doesn't have to exist.April 4, 2026 · 8 min read
• Design Thinking Is Not a WorkshopThe workshop is the spark. The practice is the fire. What companies get right and wrong.March 28, 2026 · 7 min read