Architectural innovation
Architectural innovation occurs when familiar components are reorganised into new relationships. The novelty lies not in the parts but in the architecture connecting them, which is precisely what established firms tend to miss.
The components stay the same. The way they fit together changes. And the incumbent, whose entire organisation is built around the old arrangement, cannot see the shift until it is too late.
Architectural innovation occurs when the core components of a system remain relatively familiar but the relationships between those components are reorganised. The novelty lies in the architecture that connects the parts, not in the parts themselves.
Why established firms are vulnerable
Henderson and Clark's original insight was that organisations hold different kinds of knowledge. Some of it concerns the components themselves: the physics of a lens, the chemistry of a battery, the engineering of a motor. Some of it concerns the architecture that links those components into a functioning system: which parts interact, how information flows between teams, what trade-offs are built into the design.
When an architectural shift occurs, the incumbent's component-level expertise may remain intact. But its understanding of how those components should be integrated can become obsolete. The firm does not feel ignorant. Much of what it knows is still true. The motors still work, the lenses still focus, the chemistry still holds. Yet the system-level logic has changed enough to make that knowledge misleading.
The photolithographic alignment industry in the 1960s through 1980s provides the clearest example. Established manufacturers held deep expertise in optical components, light sources, and mechanical positioning. When the architecture shifted from contact aligners to proximity aligners to projection aligners, the core optics knowledge remained relevant. But the way components needed to be integrated changed fundamentally. Each architectural transition saw a new entrant displace the incumbent, despite the incumbent's superior component knowledge.
How the filters fail
Established firms miss architectural shifts because their internal structures mirror the old architecture. Information filters, reporting lines, performance metrics, and communication channels are all built around the existing design logic. The organisation continues to improve what it already knows how to improve while the new architecture gains traction elsewhere.
The people within the firm are working hard and producing good results by their own metrics. They cannot see the shift because their metrics, their organisational boundaries, and their design intuitions are all calibrated to the old configuration. The failure is structural rather than intellectual.
System design as competitive ground
Architectural innovation is a reminder that system design matters at least as much as component quality. Competitive advantage can move from invention at the part level to reconfiguration at the system level. Firms that want to avoid this kind of surprise need to ask whether the way the system is put together is changing, not only whether each component is improving.
The concept also sharpens the idea of core rigidity. The same deeply embedded knowledge that drives advantage under one architecture can become a liability under another. What changes is not the quality of the knowledge but its relevance to the new configuration. The best component expertise in the world is a poor defence if the architecture has moved on.