Part II: The Alchemy of Organization

Modified

May 24, 2026

Organizations change through four levers: strategy, structure, people, and process. The levers are not equal. Strategy commands the highest leverage — it sets the direction of all the energy in the system. Structure determines what is possible before anyone shows up to work. People carry the capability — or the limitation — of the organization over time. Process coordinates the work that happens after the first three are in place.

The hierarchy matters. Optimizing process in a badly structured organization is motion without progress. Fixing structure in service of a bad strategy is precision in the wrong direction. And the most common mistake in engineering leadership is spending the majority of attention at the bottom of this hierarchy — on process — because process changes are safe. No one loses their job. No hard conversations. No teams get ripped apart. It is also the least powerful place to spend political capital.

When dysfunction appears, look up the hierarchy before you act. Two teams that can’t collaborate look like a people problem. It is almost always a structure problem. The interpersonal friction is real, but it is a symptom. Fix the structure and the friction mostly resolves. Coach the people without fixing the structure and the friction returns with different names attached to it.

The alchemy of organization is learning to reach for the lever that actually transforms — not the one that feels safest.


Strategy: The Decision-Making Razor examines what makes a strategy useful. Not as a slide deck, not as an aspiration — as something that cuts. A principle clear enough that when two options are on the table, it tells you which one to take. This chapter covers how to build strategy downward from organizational purpose, and how to recognize the kind that looks like strategy but can’t cut anything.

Structure: The Architecture of Delivery addresses the organizational design decisions that determine what delivery can and cannot happen. Structure is scaffolding — it creates the conditions for every other lever. This chapter covers the tenets of organization design that hold across contexts, the specific pathologies that emerge when structure is wrong, and how to navigate structural realities you didn’t design and can’t immediately change.

People: Gardening Your Teams uses the gardener’s frame because gardening is honest about what the work actually involves: growing, matching, and pruning. You cannot manufacture a strong team on a fixed timeline. You can create the conditions for one. This chapter covers the full arc — how to grow people toward the work, how to match their current capability to what the moment requires, and how to prune when you must, including the cases where you can’t.

Process: Outcomes Over Rituals makes the distinction that most process improvement efforts miss. Process is a means, not an end. When the ritual starts serving itself — when the team performs the process rather than uses it — the process has become a cost. This chapter is about designing and maintaining process that stays connected to the outcomes it is supposed to produce.

The Art of Judgment closes Part II with the disposition that makes the four levers above it actionable. The right lever isn’t always obvious even when the framework is clear. Sometimes two levers both need pulling. Sometimes the highest-leverage intervention is the one your organization will reject. Judgment is what you exercise in the gap between seeing the right answer and having the standing or the courage to act on it.


Political capital — accumulated in Part I — is what you spend here. The alchemy of organization is what that capital can buy when deployed with precision. Spend it on process and you get incremental improvements. Spend it on structure and you change what’s possible. Spend it on strategy and you change the direction of everything.

The question is never whether to spend it. You are always spending it. The question is whether you are spending it where the return justifies the cost.