Peter Drucker, Management: Tasks, Responsibilities, Practices — Chapter 47: Relations-Focused Design: The Systems Structure
Introduction
Hyundai Motor rose to a global top-three position over the past fifty years, working with hundreds of Tier 1 suppliers. Now it must move toward autonomous driving and Physical AI — not only with those same suppliers, but with an entirely new kind of partner.
For some time, I have argued that the closed supply chain model must give way to an open ecosystem. Yet I could not precisely define the difference myself. Beyond the visible contractual relationships, what really separates the two? The question lingered.
Drucker’s Chapter 47 on systems structure gave that question its shape. A systems structure is not simply a collaborative relationship. It is the most difficult and most expensive form of organization — one that allows each participant to operate by its own culture and values while moving toward a shared direction.
Whether large corporations are prepared to adopt this structure may be the most consequential question facing them in this era of transition.
Tight Control, Tight Integration: How the Old Supply Chain Worked
According to Drucker, a systems structure becomes necessary when “diversity of cultures and values must be integrated into unity of action.” Each member operates by its own logic, its own norms of behavior — and yet all members move toward a common goal.
NASA is Drucker’s primary example. It bound together military culture, German-trained scientists, large private corporations, and individual university researchers. Owning these entities, or binding them through simple contracts, would have stripped them of the very excellence each brought. What made them move as one was a single, audacious goal: “Get a man on the moon by 1970.”
Hyundai Motor’s supply chain has worked in a different way. Most suppliers shared the same culture. They invested in technology and infrastructure. In return, the headquarters guaranteed several years of purchase volume. This pattern repeated across nearly every component category.
It worked not because it was exceptional, but because the problem of that era demanded exactly that level of solution.
The Era of Convergence: Why Systems Structure Becomes Necessary
The automotive industry now faces a fundamentally different problem. Electrification, AI, robotics, content, financial services — diverse industries are converging into a single mobility ecosystem. And each of these industries operates by sharply different cultures and values.
Drucker identified precisely this condition as the case for systems structure. Each member must move in its own way, by its own logic, to be effective. Yet all must move toward a shared goal. When both conditions hold at once, no other structure works.
Control a software company through manufacturing’s methods, and the excellence it brings disappears. Bind an AI startup to traditional supply chain logic, and the speed and creativity that defined it suffocate. The point of systems structure is to allow each participant to keep its own culture and values intact while still moving in one direction.
And here lies the hardest truth about systems structure. Drucker never praised it. He called it “fiendishly difficult.” He warned against using it whenever a simpler structure could do the job. Systems structure is the last resort — chosen only when nothing else will work.
The Hardest Task: A Goal Both Large and Specific
For Drucker, the single most important requirement of systems structure is clarity of objective. “Get a man on the moon by 1970” worked because it was large enough to inspire and specific enough that each participant could derive its own role from it.
This is where reality becomes difficult. The old supply chain made such goals easy. “Global top five.” “X million units per year.” Quantitative, contract-ready, capital-investment-ready. A supplier could look at the number and decide on a five-year investment.
But how do you concretize “autonomous driving” or “Physical AI” sufficiently for a supplier to commit capital?
Suppliers still wait for the kind of quantitative target large corporations used to provide. But that target can no longer be made in purely quantitative terms.
Creating a new kind of goal that is both measurable and vast: this is the hardest task left to those who lead today.
Closing
Two questions remain.
Are you trying to control the cultures and values of your partners, or to integrate them? And to integrate them, are you prepared to create a shared goal that is both large enough and specific enough to anchor their decisions?
If you cannot answer these, systems structure is premature. The question to ask first is what simpler structure might still do the job.