Podcast: The next generation of defense primes with Matt Steckman and Trae Stephens

Matt Steckman and Trae Stephens joined me on the Acquisition Talk podcast to discuss their approach at Anduril Industries to scaling up in defense. Matt is the Chief Revenue Officer and a former guest of Acquisition Talk, while Trae is Anduril’s co-founder and executive chairman, as well as a partner at Founder’s Fund.

In the episode, we discuss:

• How Anduril is becoming a hardware rich company
• Reactions as to whether nontraditionals are helping in Ukraine
• Why software companies have larger margins than defense primes
• Whether Anduril will adopt DoD business systems as they scale
• How to improve competition through product over white papers
• Why new entrants need an effective government affairs engine

Download the full-text transcripts

Nontraditionals in Ukraine

At our recent GMU-DAU conference, USD A&S Bill LaPlante said that “tech bros aren’t helping us too much in Ukraine.” Ukraine isn’t holding their own against Russian with quantum or AI, “it’s hardcore production of serious weaponry.” I asked Matt and Trae for their reactions, because Anduril has deployed their Ghost VTOL drone for the ISR mission, as well as other equipment that is hard for them to talk about at this point. [I’ve compiled a good bit of their portfolio of systems here.]

Trae said that in many ways LaPlante was not wrong:

There’s a kind of a premium that needs to be put on scaling, manufacturing, delivering things into the field rapidly… We have had Anduril tech that has been deployed in Ukraine since the very beginning of the conflict. Is it at a scale that would rise to the Bill LaPlante level? Probably not. We’re not orchestrating multi-billion dollar arms deals as part of aid packages. But I think that critique goes in both directions.

Trae said that the issue is also a self-critique, because DoD has not been “buying things from new entrants to the market, so there was nothing in the inventory.” He stated that dual-use tech, like Microsoft Office or cloud compute, does have a way to push those commodities into government when they’ve reached a large scale. But for companies focusing on defense unique technologies, they have a much harder time and so they are often pushed towards commercial first.

Economics of a Nontraditional Company

Software is different than traditional industrial era products. The cost of deploying software and selling it to the next customer is basically zero. It’s simply to cost of hosting the data, plus some marginal services and support. Many commercial software products can be sold for 70 or 80 percent margins because the real costs incurred by the company are up front in research and development. The company owns the risk, and the margins provide not just a return on that investment but also more capital to put back into improving the product.

Large margins should not be considered predatory when they are smartly reinvested back into the business. Particularly if the product improve performance at a lower overall price. It represents cost avoidance to the customer. However, if the margins were simply returned to shareholders, then some other company will inevitably come and “gobble up your work.” That’s the beauty of free enterprise.

Those reinvestments into product are not whatever the company wants to do. Matt said that they are constrained by what the defense customers are telling them, and the company interprets those requirements in their product lines. Moreover, for products like counter-UAS, the threat is evolving at a fast pace. If a company is unable to continuously develop and release capability, then “you almost declare that you want to lose.”

Anduril is not targeting 80 percent margins like a pure software company. Anduril is targeting margins “somewhere in the 40 to 50 percent range.” Those margins are important because Anduril, on their own, spent almost one-quarter of what the top five primes spent in independent R&D in 2021 put together. 

That is an astounding fact considering their revenue was reported to hit just $150 million in 2021, a mere 0.2 percent of Lockheed Martin (that’s one-fifth of a percent). Of course, Anduril has raised over $2 billion in private funding. That outsized investment in R&D won’t continue into perpetuity. Eventually, they aim to go public with a much larger business. Investment in R&D will fall below 100 percent, but they do not expect it to fall to where the traditional primes are: roughly one or two percent of revenue (and much of that is directly reimbursed by government through indirect rates).

Doing Business with DoD

Given the business model Anduril pursues, a simple 10 or 15 percent profit over marginal cost would not suffice. It would literally be unfair, because a 10 percent profit might represent a very high return on investment for a defense prime which has low IR&D and outsources 70 percent of the work to subcontractors. However, a 10 percent profit could represent a loss on the contract for Anduril considering its private investments. Yet, as Matt pointed out:

That’s not the way that our defense acquisition system works. We don’t reward people for taking risk. We don’t reward people for reinvesting their money into research and development. The primes are all in the like low single digit percentage of IR&D, whereas tech companies tend to be in the like mid-double digit percentages for IR&D.

Interestingly enough, the aerospace and defense sector did invest roughly 19 percent of sales into IR&D in the 1950s. Yet that business model of risk in R&D and profits in production was intentionally killed by the systems analysts at RAND and the PPBE process. That has not only created perverse incentives, it also creates an antagonistic attitude in DoD toward market processes and new entrants, as Trae argued:

The closer I find myself engaging with senior DoD officials, it seems what they’re really trying to signal is that people that have found success in the tech industry by implementing new technologies in new sectors are like in some way reprehensible. It says something about the creation of wealth. What they’re saying is that we like working with the traditional defense industry because we have total control over their economics. We give them tiny margins. They’re basically just like an index fund for the defense budget… and in many ways they’re just subsidiaries of the United States government.

 

This comes out in all of these different ways: it comes out in calling the tech community tech bros; it comes out in critiques of the venture capital industry; it comes out in critiques of the patriotism of tech companies. But really, it’s a tremendous distrust in the allowing the capitalist system to work in the way that it should within an industry that desperately needs capitalism. We don’t like the idea that people could break into this and make money while saving money for the taxpayer. It’s destabilizing in some way.

There’s a lot to that statement. I think the uneasiness with the patriotism of nontraditionals and expectation that they indefinitely take self-funded risk was highlighted when SpaceX started to ask for compensation for deploying Starlink to Ukraine. But really, it is a distrust of the entrepreneurial spirit. DoD likes to build its requirements, budgets, and acquisition plans based on long-range analytical plans. The entrepreneur’s job, just like in the market, is to recognize a misallocation of resources and reallocate labor and capital to higher valued uses. This results in a better product at a lower price, which flies in the face of cost models performed by analysts and execution to baselines expected by oversight.

Scaling Up in Defense

Even though nontraditionals can make reasonable margins in commercial, competitive, and other transactions contracts, once they move into sole-source production there comes a structured approach to determining profit based on accounting cost. This is built into the regulation of FAR 15. For example, Anduril is competing on the American Rheinmetall team for the Army’s Bradley replacement program, OMFV. If they succeed and go into production, the profit guidelines will be flowed down to them, along with a slew of other business system and audit requirements. Here’s what Matt said to that:

If you want to play at the top tier level, you have to have the capability to create that spectrum of contractual structures, accounting structures, cost structures that the government transacts against.

 

… We have no choice but to do these contracts if we want to be a player in major defense programs, which we do and which we are in some cases. We have to do these things. We’re never going to move the government completely off those approaches, especially on these very large, ACAT I, II, and III-type programs. That would be silly and naive of us to think. And so again as the opportunities come up where it makes sense, we’re gonna pursue them with the full resources of the company behind it.

That’s probably a very wise stance to take. Anduril wants to be the next generation of prime contractors for defense. Despite all the talk of bringing in innovation from nontraditional sources, including a commitment to the “fast follower” approach in the 2022 National Defense Strategy, DoD cannot be expected to bend to new business models. Compliance and oversight is even starting to rain down upon novel contracting and acquisition approaches like Other Transactions and Middle Tier of Acquisition.

Matt and Trae provide several recommendations for how DoD can compete contracts more regularly based on product testing and comparative evaluations. They give reason to be hopeful, as there are signs that competitions are being run in a better way, including the Army’s Titan program with PEO IEW&S. But this also requires additional budget flexibility so that companies that perform with excellence can capitalize on those wins within the same year of execution.

As production contracts start scaling up, companies like Anduril that developed intellectual property on their own dollar will be willing to provide greater data rights. It doesn’t make sense to exchange IP for a $2 million contract when the company invested hundreds of millions into it. They government is a user, not an investor. But when scale up, then there are tons of hybrid structures of even full government purpose rights that could make sense. See more on intellectual property in defense contracts from my conversation with Anduril’s general counsel Babak Siavoshy.

Thanks Matt and Trae!

I’d like to thank Matt Steckman and Trae Stephens for joining me on the Acquisition Talk podcast. They both have had interesting blog posts on Medium, including Trae’s The Business of War is the Business of Deterrence and Matt’s The DOD Must Start Contracting for Excellence. Trae has been on a number of interesting podcasts, including the Realignment Podcast, the Aerospace & Defense Report, Eye on AI, and Venture Stories podcast.