How I Lead

01Hold the heading, not the wheel.

I go deep on the pieces that are expensive to change later: the architecture, the positioning, what the product refuses to be. And navigation never stops. Direction is a living thing, not a slide from the kickoff deck. It shifts as the market does, and keeping it current and visible is my job every single week.

What I don't hold is the wheel. When everyone knows where we need to go and the foundation under them is solid, each person knows better than I do how to do their own job to get us there. The destination can change often; that's fine, as long as nobody has to guess what it is. That is how you scale: a team at ease at the wheel, not a leader who can't let go of it.

Micromanaging is a symptom, not a cure. The urge to grab the wheel means something upstream is unclear: the heading, the foundation, or the hiring. Fix that instead.

02Every seat is a product seat.

Hyperfocus on users, product, and positioning is not the product manager's private job. It belongs to every seat, and engineering most of all. An engineer who knows exactly who the user is and why the product wins makes a hundred better decisions a day without asking anyone's permission: API names, error messages, what to optimize, what to leave alone.

The best engineers I've worked with could pitch the product cold. That's no accident: teams that split into "builders" and "product people" bleed energy into translation, and the loss compounds. Now that AI writes much of the code, an engineer's product sense is the multiplier that separates output from progress.

03Teach fishing.

Give a person a fish and you feed them for a day. Teach a person to fish and you feed them for a lifetime.

Handing someone the answer solves today's problem and guarantees they bring you tomorrow's. I try to hand over frameworks instead: how to break the problem down, what to measure, when to escalate. It is slower the first month and dramatically faster every month after, and it grows people instead of dependents.

04Hire self-directed people.

I favor people who pull work toward themselves over people who wait for tickets, every time I can. Self-directed people turn context into action; my job becomes supplying the context and clearing the road. One of them is worth a great deal of management overhead you never have to spend.

05Data over opinions. Prototypes over debates.

Opinions are how meetings expand. When there is data, we follow it, including when it disagrees with me. When there is no data, the move is not a longer debate; it is the cheapest prototype that produces some. An afternoon of testing settles what a week of argument cannot.

06Stay close to the code.

AI made this easier for leaders, not optional. Models write the bulk of the code now, which means someone senior has to read it, question it, and own what ships. I still write and review code, and I think product and engineering leaders who let that muscle fade are handing their standards to an autocomplete.