The Real Work Begins After Go-Live

This part has four chapters, and together they answer one question: once a system has "shipped" in the technical sense, how do you make it "live" in the organizational sense — used by people, backed by people, and still running once you walk away.

Part Three wrapped up with the data connected, actions under control, outcomes provable, and bills accountable. But a production system's real enemy isn't only engineering — it's the organization. A system that scores full marks at acceptance can die quietly within three months, and the cause of death is never in the code. It's in the incentives.

Chapter 15 dissects this silent death first: going live is not the same as activation, and all three failure modes happen only after "successful launch." The antidote is three questions about incentives, embedding change gradually instead of all at once, and cultivating two kinds of allies.

Chapter 16 hands the key to its rightful owner: AI adoption is inherently cross-departmental, and budget, process, data, and benefit are four cards only an executive sponsor can hold at once. Executive sponsorship isn't a slogan — it's three kinds of backing (budget, incentives, failure), two gates the project has to pass through, and a flywheel that gets heavier the more it spins.

Chapter 17 defines where delivery actually ends: after acceptance and adoption there's a third stage — handoff. The three exit conditions, the handoff ladder, three anti-patterns — all boiling down to one line: a system that can't survive without its FDE was never delivered. It was parasitic.

Chapter 18 turns the lens back to China: procurement structure, trust structure, and who defines the last mile all work differently there. Four structural differences give rise to two homegrown forms — and because China's constraints are tighter, they make an ideal stress test for the methods in this book.

The four chapters form a single progressive chain: first get the system used (adoption), then make sure an organization is backing that use (executive sponsorship), then, once it's stable, hand it off (handoff) — and how hard that handoff turns out to be is finally decided by the soil it's planted in (China). The primary reader for this part is whoever is on the hook for "the system is actually alive": the delivery lead running the project, the customer's IT or informatization owner, and any manager stuck staring at a system that shipped and nobody uses. Engineering takes half a step back here, and the organization takes half a step forward — because every step in this part starts from the same line, said in a conference room: "Why won't anyone use this thing?"