The fastest way to build something is to deconstruct it. Think about all those companies that produce complicated things on an insanely large scale: Toyota, McDonalds, Boeing. If you want to build enough cars to move the entire world or sell a trillion burgers or create an aircraft with millions of moving components, you need to first prototype the final product and then systematically divide it into its individual pieces. If you look at each part separately, you can develop a product faster and more flexibility. I call this process modular design and development: breaking down complex systems into parts that can be developed and tested independently.

This strategy works best on products or processes that are highly complex and contain identifiable subsystems. For example, think of the intricacies of a new surgical methodology or the construction of a eco-tech water filtration system. It also works best in situations of high volume or scope–when you want to take your product or service to scale.

Modular design and development can help lead you to any of these desired outcomes:

• Methodical development of products and services with reduced risk
• Outlining of clear methods for designing quality
• Adapting and customizing products or services to keep up with a changing marketplace
• Reducing wasted effort by avoiding reactive behaviors

Modular design and development is not about radical innovation. It’s about efficiency and reliability. It’s excellent for those who desire incremental innovation, essentially building off existing sciences, technologies, or systems. Testing early and often is crucial to eliminating bugs on individual pieces before they are integrated. With this technique, you can experiment with numerous variations to find the optimal design with the lowest risk.

Be methodical when you perform modular design and development:

• First, clearly define the current specifications of the product.
• Next, create concrete standards for development.
• Then, estimate and assign time and resources.
• Build prototypes, models, simulations, and proofs of concept.
• Identify and communicate lessons learned from experimenting with these variations.
• Finally, return to the first step and begin the next iteration of the product.

The goal is not to avoid failure but to accelerate the failure cycle with low risk of impact. Most early simulations and prototypes will inevitably fail. But these will move you closer and faster toward the product that will work.