Modularity Theory

The theory that spurs performance and explains adoption of a product or service within a market.

Gears illustration


Modularity Theory (or the Theory of Interdependence and Modularity) explains how different parts of a system’s architecture relate to one another and consequently affect the development and adoption of that system. 

According to the theory, a system’s architecture determines its components and systems and defines how they must interact—fit and work together—in order to achieve desired outcomes.

The place where any two subsystems fit together is called an interface. A system’s architecture is interdependent at an interface if one part can’t be created independently of the other. In other words, the way one is designed and made depends on the way the other is designed and made. Unpredictable interdependencies require the same person, team, or organization to simultaneously develop both components for the overall system to work. 

In contrast, a modular interface is one in which there are no unpredictable interdependencies between subsystems, people, teams, or organizations. Modular components and subsystems fit and work together in well-understood and highly-defined ways. A modular system architecture specifies the fit and function of all elements so completely that it doesn’t matter who makes the components or subsystems, as long as they meet specifications. In other words, to the extent that interfaces are specifiable, verifiable, and predictable, they are modular in nature. Systems that rely on modular interfaces allow people in separate teams or organizations to do their work with little to no effort spent on coordination.

Contrasting Product Architecture Types

Unique Insight

The implementation of either modularity or interdependence does not determine whether a product is adopted, but predicts how quickly adoption will occur.


What type of system architecture should you use to equip your business model? Whether it’s modular or interdependent, which one you choose as you enter the market predicts how quickly adoption of your product or service will occur. Check out our helpful video.

Case Studies

Apple largely builds integrated architectures—they have control of the full architecture of subsystems and interfaces that come together to serve their customers. As a result, they can offer a highly functional and reliable user experience. Apple’s AirPods bluetooth headphones switch effortlessly between an iPhone, a Macbook, and an iPad because Apple designs all of these devices and their interfaces. If devices malfunction, Apple’s Genius Bar has troubleshooting services and software optimized for its particular devices.

In contrast, devices that run on Microsoft Windows have a non-integrated architecture—modular interfaces specify the fit and function of all elements so completely that it doesn’t matter who makes the components or subsystems—as long as they meet specifications. This non-integrated architecture supports coordination between various independent providers. While Microsoft supplies the Windows operating system, the physical devices that run Windows hail from a wide array of suppliers like Dell, HP, Lenovo, Asus, and Microsoft. Whether you want a machine with a high-end graphics card for gaming or a cheap computer to run basic applications, you have many more options, design styles, and price points with Windows-based machines.

Computer blueprint

It’s important to note that the tradeoffs inherent in integrated vs. non-integrated systems aren’t static approaches for appealing to different customer segments. Innovation research shows that, over time, certain improvements in non-integrated systems can shift how customers weigh their tradeoffs.

In the 1980s when desktop computers were new, Apple—the most integrated company—made the best desktop computers. They were easier to use and crashed much less often than computers of non-integrated design. But, over time, as non-integrated systems improved, they eventually caught up to Apple’s functionality and reliability. In the 1990s, when the functionality of desktop machines became good enough, the non-integrated, open-standard architecture of Windows-based machines became dominant. Apple’s proprietary architecture, which in the not-good-enough circumstance was a strength, became a disadvantage in the more-than-good-enough circumstance. 

Fast forward to the present, and the newness of mobile and wearable devices has given Apple the advantage again. It leverages its integrated approach to deliver the cross-device functionality and reliability that consumers demand. But innovation theory predicts that, over time, less-integrated designers of mobile and wearable devices will increasingly appeal to Apple’s current customers. As the functionality and reliability of non-integrated suppliers gets good enough, demand shifts toward customization.

By and large, school districts are integrated education systems—they have control of the full architecture of subsystems and interfaces that come together to serve their students. For example, curriculum and instruction departments work out the interdependent interfaces between curricular materials and instructional practices. These departments ensure that the curriculum a district adopts or develops aligns with the districts’ instructional standards and pedagogical philosophy. Then, they ensure teachers receive the professional development needed  to use the curriculum effectively. That said, teachers know to effectively engage and instruct their students, they need to align their lesson plans and materials with their particular students’ needs. As a result, teachers adapt and supplement curricular materials.

As a district works to improve how it serves its students, if there are interdependencies in the interfaces between its various teams, departments, or schools, the district’s leaders have the authority to reorganize the district’s subsystems and their interfaces to resolve the interdependencies.

Drawing of a building project

In contrast, a learning ecosystem is an example of a non-integrated system because their modular interfaces allow providers to work independent of one another. In these ecosystems, students can move seamlessly across various learning experiences provided by an array of organizations that all connect to provide a cohesive form of formal education. In this scenario, the ecosystem would need rules and standardized processes to determine how learning providers join the ecosystem, how funding flows through the ecosystem, how credit for learning gets counted, how transportation between learning sites gets coordinated, how struggling students are identified and supported by the ecosystem, etc. 

These rules and processes enable providers to work at arm’s length with minimal coordination effort. But they also constrain the providers, requiring them to conform to pre-established rules and processes.

For now, the unique opportunities available through learning ecosystems appeal to a small subset of learners and families. They will never be viable mainstream alternatives to traditional schooling unless they get good enough at functionality and reliability. People generally don’t switch to non-integrated systems because their values change. Rather, they switch after non-integrated systems improve over time and start to get “good enough” with their offerings. 

Today, learning ecosystem proponents focus on increasing the supply, diversity, and quality of custom experiences. But they  may be inadvertently shooting themselves in the foot if the quest to increase customizability leads to routinely underinvesting in functionality and reliability.

These insights are part of Thomas Arnett’s blog series on learning ecosystems. Learn more here.

Driverless car illustration

Autonomous vehicle companies

Some companies like Zoox took an integrated approach to build the entire vehicle themselves, while others like Aurora took a modular approach by only building the “brain” of the vehicle and attaching it to traditional cars and trucks.

Car assembly illustration

Ford Motor Company

In the beginning, Ford was vertically integrated all the way to owning its own steel mills because it needed full control to improve performance in the early days of the industry. Now, the auto industry is fully modularized with nearly every part being made by a specialized supplier. Today, automakers focus mostly on design, marketing, and assembly.

Helpful Tools

Infographic: Definitions & conditions?

Modularity Theory explains how different parts of a system’s architecture relate to one another and consequently affect the development and adoption of that system. This theory is built on a number of components and conditions. Need a cheat sheet? We’ve got you covered. Click the image to download and check out our resource library for additional content.


  • Architecture: Determines a system’s constituent components and how they must interact—fit and work together—in order to achieve desired outcomes.
  • Interface: The place where any two subsystems fit together.
  • Interdependence: If the way one part the system’s interface is designed and made depends on the way the other is being designed and made.
  • Modularity: The fit and function of all elements are understood so completely that it doesn’t matter who makes the components, as long as they meet the specifications.
  • Integration: When a system has control of the full architecture of subsystems and interfaces that come together.
  • Non-integration: When systems depend on modular interfaces to coordinate the work of various independent providers

Conditions for Modularity

  1. Specifiability: The people or organizations working on both sides of an interface need to know which attributes of the component are crucial to the operation of the system, and which are not.
  2. Verifiability: They must be able to measure those attributes so that they can verify that the specifications have been met.
  3. Predictability: There can’t be any unpredictable interdependencies across the interfaces. People using the system need to get what they expect when interacting with the system.

At the Clayton Christensen Institute, we’re using Modularity Theory to:

Masked teachers and students in a classroom

Develop student-centered teaching competencies

Most K–12 educators today don’t have the skill sets necessary to run student-centered schools. This report helps dismantle that barrier by identifying specific student-centered competencies educators can stack to create customized teaching micro-credentials.

Students working together

Evaluate the growth of learning ecosystems

What if instead of formal education happening through schools, students could stack together various learning experiences from across a learning ecosystem? Integrated and non-integrated systems each have particular advantages and disadvantages when it comes to serving students. Understanding the trade-offs with each approach can help us better understand who learning ecosystems would be best suited to serve and the runway they would need in order to become viable alternatives to conventional education.


Imagine better systems of support for postsecondary alternative credentials and credit transfers

An era of increasing data interoperability,new postsecondary entrants are changing the fundamental architecture that connects higher ed students to institutions in the workforce. As a result, the very categories and boundaries in place for  transfers, as well as work and learning, are being challenged.

Get more research at your fingertips

Check out our latest Modularity resources: