Modularity

Clayton Christensen called it the Law of Conservation of Attractive Profits: when one layer in a value chain becomes modular and commoditised, the adjacent layer becomes integral and profitable. The profits don't disappear. They migrate. The strategic question — the only one that matters — is whether you are building in the modular layer or the integral layer, because the economics of each are diametrically opposed and the wrong choice is terminal.

The PC industry provides the canonical demonstration. In the 1980s, IBM owned the entire personal computer stack: hardware design, operating system, application software, distribution. The value chain was vertically integrated, and IBM captured value at every layer. Then IBM modularised the hardware architecture — open standards, interchangeable components, commodity manufacturing. The hardware layer became modular: any manufacturer could assemble a PC from standardised parts. The profits didn't vanish. They migrated to the layers that remained integral: Microsoft's operating system (the software layer that tied the modular hardware together) and Intel's processors (the component that determined performance across all modular hardware configurations). "Wintel" captured the value that IBM's modularisation released. IBM's PC division, which had created the market, was eventually sold to Lenovo for $1.75 billion — a rounding error compared to Microsoft's and Intel's cumulative profits.

Apple understood Christensen's law and made the opposite choice. While the PC industry modularised, Apple kept its value chain integrated — designing the hardware, the operating system, the silicon, and the retail experience as a single interdependent system. The integration captured value because each layer was optimised for the others in ways that modular competitors couldn't match. The M1 chip was designed specifically for macOS. macOS was designed specifically for Apple hardware. The retail experience was designed specifically to demonstrate the integrated product. The result: Apple captured 80%+ of the smartphone industry's profits with roughly 20% of its unit share. Integration in a modular world is the highest-margin position available.

AWS provides the most consequential modern example. Amazon modularised computing infrastructure — servers, storage, networking, databases — into standardised, interchangeable services available via API. Before AWS, every company that needed computing infrastructure had to build its own integral stack: buy servers, hire operations teams, manage data centres. AWS modularised that stack into commoditised components. The infrastructure layer became modular. Christensen's law predicted what happened next: profits migrated to the adjacent integral layer — the application layer. Companies like Stripe, Snowflake, Datadog, and Twilio built integral products on top of AWS's modular infrastructure, capturing value because their software integrated data, workflows, and user experiences in ways that modular infrastructure alone could not.

The pattern repeats across every industry. Smartphone hardware is increasingly modular — Qualcomm's chips, Samsung's displays, and Foxconn's assembly are available to every manufacturer. The integral layer is the software ecosystem: Apple's iOS and Google's Android capture value because they integrate the modular hardware into a coherent user experience. Semiconductor fabrication was modular until TSMC made leading-edge manufacturing integral — and now TSMC captures value that the entire fabless semiconductor industry depends on. Media distribution modularised (anyone can publish content to the internet), and the aggregation layer became integral — Google, Facebook, and YouTube capture the value because they integrate modular content into a coherent discovery experience.

The law is not a theory. It is a description of how value moves through technology stacks, and it has predicted the winner in every major platform shift of the past forty years. The leaders who understand it position their companies at the integral layer before the modularisation of the adjacent layer becomes obvious. The leaders who miss it build better commodities — and watch their margins evaporate as substitutes proliferate.