Here is Trae Stephens, partner at Founder’s Fund, talking with Vago Muradian on the Defense and Aerospace Podcast.
We found that selling software was incredibly difficult. There’s almost a moral aversion to paying software margins inside of the DoD. Our focus was putting the thing we know DoD really needed, which was the software system/autonomy capability, and wrapping it in metal because it is significantly easier for the customer to buy metal than it is to buy software. We use that as a Trojan horse to get software in the front door. It significantly simplifies the sales process.
I think what might be going on is that Anduril does not want, nor expects, a major software development contract to build to DoD requirements and get paid by the number of source lines of code (SLOC) written. But it can’t build the software upfront and reap 50-80 percent margins on the backend to earn a return on all those sunk R&D, sales and marketing, general and administrative expenses. So I speculate it builds those software margins into hardware instead.
Traditional primes would spend tons of money building a new sensor and integrating the capability with some simple software. DoD is used to paying high prices for a piece of hardware. Anduril is banking on doing the hardware for significantly less cost using off-the-shelf tech, but building enough margin into the price to pay for the real capability which is all the software that creates new behaviors. A price analysis by a contracting official would find Anduril’s price fair and reasonable compared to existing hardware, even though Anduril’s value add is in the software.
Anduril’s Lattice surveillance system essentially pulled together off-the-shelf devices and sensors, integrating that with sophisticated AI/ML for target recognition and tracking. The system is extensible, adding elements like counter-UAS effects, base defense, and battlefield management capabilities. But that doesn’t mean Anduril would be well positioned to write code for bespoke requirements like the F-35 logistics system or satellite downlinks from SBIRS High.
By contrast, traditional defense companies respond to siloed DoD requirements, building things from scratch and often without utilizing enterprise capabilities. There is no shared code base. Interfaces are not well documented because they are one-offs. That’s what makes them so expensive. It is the reason traditional companies plead with government: “Just tell us your requirement and we’ll IRAD that!” But that means they have no strategic vision of their own.
One of the major cultural differences between the market economy and DoD is that DoD focuses totally on process. It believes strategic vision for weapon systems comes from a step-by-step process where people are interchangeable. As long as enough functional offices review and add to the plan, it will be optimal. That’s simply not how important companies and capabilities are created in the private sector, as Trae notes:
The number one thing in my mind is allowing the companies in private industry to bring you a vision for how to address the critical problem in the Department, rather than offering a highly spec’ed out requirement that they would then need to go build from scratch.
If you look at any commercial innovation that changed the world, things like the I-Phone and Tesla automobile, these did not result from a requirements process. They result from the brains from entrepreneurs that came forward and said, “based on where technology is today, the physics and software that has been solved today, things available at scale from existing manufacturing in other sectors, we can take those things and turn them into a really clever product that addresses a core problem you’re having.”
Believing that government officials are the ones coming up with these visionary concepts is fundamentally wrong. You need to be open to, and have the ability to buy, things that you couldn’t even dream of existing. We spend too much time talking about what the Department’s vision of the future is. That matters less than what exists and what people are able to build.
I would quibble that folks inside government can have strategic insight. It has been demonstrated several times in the past. But in DoD acquisition, people take a back seat to process. That tacit knowledge could reside with civilian or military folks, but often isn’t cultivated to the same degree. If people’s strategic vision and ability to execute was a flower, DoD would be trampling it with process and frequently transplanting it to inhospitable conditions.
What DoD needs is founder-led programs, rather than consensus-based programs. Founders can lead with context and judgment, where speed and delivery of operational capability is the focus. By contrast, what we have today is paralysis-by-analysis, where managers execute the cost-schedule-technical baseline imposed by ivory-tower planners.
Leave a Reply