This is part of my ongoing series exploring Innovation, Lean, Lean Startup, and Agile.

A few days ago I wrote about how the Lean Product Development method at Toyota essentially builds a Skunk Works around a new technology. This concept was popularized by the Lockheed Martin Skunk Works, which produced some of the most amazing aircrafts ever created, including the fastest plane ever made, the SR-71 Blackbird, which was designed in the mid 70’s.

In a similar fashion to Toyota, Lockheed relied heavily on existing technologies to drive down the cost of a new break through innovation. This allowed the Engineers to focus more energy on the few systems that truly required breakthrough engineering and thinking. Existing components for the bulk of the design also significantly reduced costs as the engineers did not have to design everything from scratch and could rely on standardized or consistent components at interfaces. This means the cost was lower using an already designed part, the part could be purchased at economies of scale, and the engineering work on new components was reduced because of they didn’t need to design interfaces.

This approach is seen as a best practice not just by Toyota and Lockheed, but Christensen argues this approach should be applied for any sort of disruptive innovation at a large company. This is required because of the risk of cannibalization of existing sales, lack of alignment with the needs of existing customers (doesn’t meet performance needs), and wins that might be small for the overall organization are huge for the new team. The alignment between wins and team help cement the team as successful and continue to shelter it from unrealistic demands from the broader organization.

In the Innovation Machine the author argues this is the approach that the most successful companies at innovation use. It helps companies become serial innovators and continually disrupt their markets and industries. Furthermore, the small wins for a new team doesn’t mean that these companies can’t have extremely aggressive goals for these internal startups. In fact, it’s likely many of them have billion dollar goals within 5 years for them, such as IBM has for their internal startups. However, the funding and the organizational structure is sheltered for a short amount of time and protected from the normal funding cycle that other portions of the organization are subject to. This is to ensure investment in future innovations that have a great deal of uncertainty.

The added benefit of these internal ventures is that the leadership of the entire organization can decide to use them as a means to change how the organization operates. For example, in the case of the Macintosh, re-integrating the Mac mini-company back into Apple forced massive changes in the way everyone worked. If the work processes are more efficient in the Skunk Works it is key that those processes are adopted by the business unit that will eventually house that new innovation. If the Skunk Works is using Lean processes and Lean startup practices integrating into the rest of the organization is going to be difficult unless those leaders are helped to get on board with the methodological approaches. This will take time but will certainly be worth it.

This is the approach that GE is taking for their new engine development. They essentially used the Lean Startup to develop an MVP by using a Skunk Works/Lean approach where they leveraged an existing engine to move the technology in a new direction with customer feedback the whole way.

For companies to be successful in deploying these approaches it will be key to to develop a safe place for them to experiment. Learn and grow away from the unrealistic expectations of a mature idea, when the idea is living in a space of extreme uncertainty.