Running the Practice

By this point, we can explain several ways to make a plan feasible. The harder conversation begins when somebody has to accept the delay, pay for the reserve, or change a commitment they have already made. A technically sound option doesn’t become an operating plan simply because the capacity team has finished the calculation, and those conversations go better with people you already talk to. If you’re not having a face-to-face meeting with your internal customers regularly (Zoom is fine), you’re not actually running capacity. Capacity is running you.

This is the third level of the work, shaping the system: once we understand what drives consumption, which interventions actually change it? Donella Meadows, who spent her career studying how systems change, ranked the places to intervene in one and put numbers last. “Probably 90—no 95, no 99 percent—of our attention goes to parameters,” she wrote, “but there’s not a lot of leverage in them.”¹ A spending target is a parameter. The greater leverage lies in who knows what and when, in the rules for who can agree to what, and in the goals the system is built to serve, and that is where the following chapters spend their time.

We will follow that work through relationships and specialist investigations, reserve decisions, risks and anomalies, and commitments whose dates have to join up. Then come the additional constraints of operating in several markets with different oversight, privacy, and sovereignty requirements. The monthly review asks what actually happened; the chapter after it examines how the practice changes when the bill is ten times larger. The book closes with the markets forming around compute itself. This is where the enterprise learns whether its process helps people make difficult decisions or merely helps them describe those decisions after the useful choices have run out.

Notes and sources

1. Donella H. Meadows, Thinking in Systems: A Primer, ed. Diana Wright (White River Junction, VT: Chelsea Green, 2008), 148. Her earlier essay, Leverage Points: Places to Intervene in a System (Hartland, VT: Sustainability Institute, 1999), checked October 4, 2026, ranks twelve places to intervene: the goals of the system rank third, its rules fifth, and the structure of its information flows sixth, while parameters rank last.↩︎