How I work.
A working map of what I have learned, what I have watched fail, and what I still try to live by. None of it fixed. All of it updated when reality disagrees.
01
Build a relationship with reality before a relationship with your opinion of it.
1.1
Treat every belief as a hypothesis.
Every strong view has a weak point. If I have not found mine yet, I have not looked hard enough.
1.2
Never speak beyond your receipts.
Early on I watched a lot of senior people talk confidently about things they had never touched. I decided I would not join them. If I have not lived it, I do not speak on it.
1.3
When the data says otherwise, update.
An expert treats their knowledge as fixed. A scientist treats it as a hypothesis still in motion. The rooms that compound over time are run by scientists.
02
Earn your seat. Credentials arrive afterward.
2.1
Do the unglamorous work first.
At fourteen my internship was throwing out trash, painting apartments, and working with the construction crew. And sitting in board meetings. The first three opened the fourth.
2.2
Show up before the work is assigned.
Ownership opens doors that applications do not. The best assignments of my career were not in my contract when I started.
2.3
Investigate before you are briefed.
Every room runs on shared assumptions. The unspoken question is usually the one that decides whether the work ships. The operators I respect go looking for it before anyone hands them the brief. That habit has opened more doors for me than any line on a resume ever did.
03
Surround yourself with people whose track record in their domain exceeds yours. Then listen.
3.1
Humility is a hiring strategy.
The founder who taught me this built a company from thirty people to eight hundred. His one rule: leverage on what other people know. You will never have the time or the energy to learn it all.
3.2
Twenty focused minutes beats six months of solo study.
The shortest path to fluency is a clear-minded subject matter expert willing to give you an honest opinion.
3.3
When you hire smart, get out of the way.
Telling a capable person what to do is the fastest way to lose them.
04
Leadership is a posture, not a title.
4.1
Meet each person as a person.
Hierarchy flattens people into roles. I learned to see past it in a Swiss boardroom, watching one HR director command a room of senior executives through how completely she met each one. Their thinking style. Their rhythm. What they needed to feel heard. It costs nothing, and it changes everything about how the decision actually lands.
4.2
Take responsibility first. Assign blame last.
I learned this at eighteen, commanding senior soldiers in a culture I was still adapting to. Take responsibility when the unit makes a mistake. Share the credit when it lands. People work through hard nights for the leader who does this. They update their resumes for the one who does not.
4.3
Trust your team on the ground more than they trust themselves.
The strongest units I have been part of were the ones where the people below me knew I believed in them more than they believed in themselves. Not coddling. Not flattery. Earned trust, given early. It multiplies what the team is actually capable of.
05
Stay a builder. Stay a beginner.
5.1
The day you stop building is the day your opinion stops being worth much.
The people who shaped me most were builders first, commentators second. When I notice myself talking more about work than doing it, I treat it as a warning. Expertise that does not touch a keyboard, a client, a product, or a team decays faster than anyone admits.
5.2
Pick one thing a year you are not already good at.
The French was not on my curriculum. The SharePoint platform was not in my contract. The AI work started before anyone paid me for it. Every chapter I care about opened because I was willing to be bad at something first.
5.3
If I am comfortable, I am probably behind.
The moments I was most at ease in my career, I was usually falling out of pace with the people around me. Discomfort is data. It is the earliest signal that the next thing I should learn is already here.
What I do not believe.
Lines I will not cross. The opposite side of the principles above.
No. 01
I do not believe technology wins adoption. Engagement does.
I have shipped solutions that nobody used. Shipped was not the victory. Adopted was. Every product I have built since starts with who the user is, what their Tuesday looks like, and what they have stopped hoping will change. The build follows that.
No. 02
I do not believe AI replaces humans. It sharpens the operators paying attention.
The frontier labs themselves, Anthropic, OpenAI, Google, are the first to insist that humans in the loop are non-negotiable. AI is a collaborator, not a vending machine. The projects I watch fail are the ones where the operator expected output without judgment. The ones that ship are the ones where the operator treats AI as a colleague whose work still needs verification.
No. 03
I do not believe relationships are transactional.
Most business schools teach relationships as leverage. What you give, what you get, measured in months. Saudi taught me a different arithmetic. Relationships earned over years do not work that way. Trust is given, slowly. Respect is returned to those who offer it first. The relationships I lean on now were built long before I needed them.
No. 04
I do not believe hierarchy makes better decisions.
In Sweden, the idea speaks. No matter who brings it, no matter their rank or title. That wiring is why Klarna, Spotify, Volvo, and IKEA exist. I have tried to bring it into every room I operate in since. The best decision is not the loudest voice. It is the clearest thinking, from whoever in the room has it.
No. 05
I do not believe in charm as a substitute for work.
Charm opens the door. Work keeps it open. I have watched polished operators lose clients inside a quarter because the delivery never caught up with the pitch. I have also watched quieter operators keep those same clients for a decade. The best version of this job is the one where the work speaks louder than the speaker. I am still working on it.
These evolve. When they do, I update them here.
Last reviewed · April 2026