Matthew Steckman on Anduril’s Defense Business, Procurement and Product Bets

Harry Stebbings welcomes Matthew Steckman, president and chief business officer of Anduril, to discuss defense procurement and company building. Steckman recalls knowing Anduril co-founder Trey since their first year at Georgetown. Stebbings identifies himself as an investor in defense companies, a relevant interest as he questions a defense-industry executive.

Steckman says Anduril’s leadership team has remained together since the company began. He attributes its resilience partly to friendship and partly to different backgrounds: technical outsiders bring fresh ideas, while people with customer and defense experience understand how the system works. This is his account of Anduril’s development, not an independent assessment.

Stebbings asks whether a founding team can succeed without sales and procurement experience. Steckman says winning a major defense program requires multiple disciplines, including acquisition, budgeting, technology and operational analysis. Because relatively few people have already won a large program, he says companies must assemble complementary experience rather than expect one hire to provide everything.

When advising teams, Steckman says he recommends finding someone with procurement experience if the founders lack it. Stebbings notes that he sees many European drone-company pitches and asks what warning signs Steckman notices. His investment activity gives him a commercial interest in the market they are discussing.

Steckman identifies two recurring mistakes. First, founders may assume a technology is novel when an equivalent has already been developed; he recalls encountering a pitch for something he says existed decades earlier. Second, a startup may mistake a narrow need or one potential contract for a sufficiently large, enduring market.

Asked whether a European company can become a major defense supplier without US sales, Steckman argues that a substantial American business is usually necessary. He estimates the US accounts for half of defense spending and says European purchases are fragmented among sovereign countries. Both the percentage and his conclusion are his own estimates and judgment.

Stebbings asks what Anduril’s announced $20 billion US military contract actually means. Steckman compares it with a credit-card limit: the government may spend up to that amount on a class of Anduril technology, but, he says, no money was obligated under the vehicle at the time. The ceiling is not a $20 billion order or guaranteed company revenue.

According to Steckman, the vehicle consolidates some contracting and evaluation steps that individual offices would otherwise repeat. He says individual purchases still require work and that orders and funds flow as products are delivered. His description explains the company’s understanding of the arrangement; it does not establish that the full ceiling will be used.

Stebbings asks whether the deal replaces spending with established primes. Steckman does not quantify substitution; he says the vehicle reflects enough existing federal business to justify a sophisticated arrangement. He estimates Anduril will make a couple of billion dollars in revenue that year and compares it with a much larger incumbent, but neither figure is independently verified in the interview.

Steckman says Anduril offers customers several ways to purchase: software can be packaged with hardware, and some products can be delivered as a service. The approach responds to government funding categories that restrict what a particular budget can buy. He describes an early counter-drone service for US special forces as an example of providing an outcome under contractual performance measures.

The hardest part of government business, in Steckman’s view, is the gap between rare large wins. He estimates that Anduril will have about 600 contracts during the year, only roughly 20 of material revenue size. Those are his prospective company figures; he says the concentration makes each major opportunity consequential.

Steckman says customers do not always specify what they will need five to seven years from now. He combines government rhetoric and budgets, military analysis and his view of future technology to make product decisions. He describes the process as educated guesswork, acknowledging that the underlying system is difficult to predict.

Looking back, Steckman says Anduril should have entered offensive cyber years earlier, although it was probably too small at the time to understand the opportunity fully. He says the company may try to catch up. This is an executive’s assessment of a potential market in which his employer could seek business.

Steckman describes cyber operations as potentially asymmetric: a relatively inexpensive action may create a large effect. He distinguishes non-kinetic activity from an explosion and says that ambiguity can make attribution and response difficult. His discussion stays at the level of policy and risk, not specific operational evidence.

In his account, leaders confronting a cyber incident must ask who was responsible, what response is proportionate and whether a physical response would escalate a conflict. He also asks whether a country can respond in kind without escalation. He argues that these choices deserve more public debate; he does not claim there is a simple answer.

Asked which assets matter, Steckman names critical infrastructure, energy, finance, health data and military systems. He says an attack on military systems could affect the ability to defend or mount operations. These are examples of potential consequences, not a report of a specific verified incident.

Stebbings asks why Anduril should enter cyber rather than concentrate on its strongest products. Steckman says the company deliberately went wide because individual technology categories can have very few large programs. In his view, missing one large award can leave a narrowly focused supplier without enough business.

He predicts that small-drone markets will produce very few material programs and, in the European example, says there will be one winner. That is a competitive forecast from an Anduril executive, not an established market outcome. It also sits alongside Stebbings’s disclosed interest as an investor in defense startups.

Steckman describes Lattice as a shared software platform for consuming data, making sense of it and controlling robotic systems. He says the company has about 20 separate profit-and-loss businesses applying common technology to different problems. The business count and claimed advantages are his description of Anduril.

As an illustration, Steckman says some code from an early sensing tower is present in a later autonomous aircraft, though much has changed. Reusing components, he argues, can reduce cost and development time. He says Anduril generally enters a new market when it believes it already has applicable technology or a relevant technical advantage.

For a new idea, Steckman says Anduril forms a small cross-functional team and spends internal research funds to build a quick demonstrator. The team then tests its assumptions with potential government customers: a prototype might be modified or abandoned if it misses the need. He looks for the point where customer enthusiasm begins to match the team’s own.

He says an individual government champion often helps shepherd a promising idea. Early contracts function like stepping stones, allowing the company to learn whether its approach works before seeking a much larger program. Steckman presents this as his experience, not a guarantee that a champion or early contract will lead to a major award.

Asked when to commit more capital, Steckman says teams gather evidence from formal and informal customer feedback, operators and the legislative branch. An internal investment committee reviews what a team learned, its next milestones and the anticipated cost. He believes Anduril can fund and move faster than peers once convinced, a company claim the interview does not independently test.

Steckman names the missile portfolio as a recent upside surprise. He says Anduril entered it expecting demand for lower-cost systems, including the Barracuda family, and believes later geopolitical events reinforced its thesis. His interpretation connects commercial opportunity with conflict; the interview does not establish independent product or market performance.

He refers to public reporting on missile use during the conflict involving Iran and says replacing some existing systems can be costly and slow. His stockpile and replenishment claims are not independently verified in the interview. He argues that mixing expensive high-performance systems with larger quantities of lower-cost products may be useful.

Steckman says one Anduril cruise-missile airframe uses a manufacturing process he compares to making a bathtub, allowing access to a broader pool of contract manufacturers. He claims this design can make production more responsive to demand. Those are company claims about manufacturing capacity, not independently measured performance.

Asked about margins, Steckman says Anduril has gross margins above 40% across the business, while individual products vary. He says higher-volume products can carry lower margins because of customer expectations, even if scale later offers cost savings. The aggregate figure is self-reported and the interview does not provide financial statements or a product-level breakdown.

Stebbings asks whether the company has ethical questions about how missiles are used, including in relation to Iran. Steckman says he trusts democratically elected governments to set rules and regards the supplier’s role as complying with them. That is his ethical position as a defense-company executive, not a settled answer to debates over weapons, civilian harm or accountability.

Stebbings asks whether war is good for business. Steckman rejects that framing and argues that defense companies should think over decades rather than become dependent on one conflict. His caution about short-term concentration is his business perspective; it does not establish the financial effects of any particular war.

Steckman says Anduril was much smaller when the war in Ukraine began and had some equipment there. He estimates the company has grown to around 8,000 people and says its participation in the Middle East is mainly defensive. He declines to identify specific systems in specific operations; the interview does not independently verify deployment claims.

He says Anduril makes air-defense systems intended to counter incoming threats and names the Shahed drone as an example of a threat. He does not disclose which weapons were used in particular incidents. His account should therefore be read as a broad company claim, not a verified inventory of battlefield deployments.

Given an unlimited acquisition budget, Steckman says he would buy several venture-backed defense companies whose technology he admires. He says their valuations currently make purchases hard to justify, while Anduril has acquired nearly a dozen or more mostly traditional defense businesses. The acquisition count and assessment of valuations are his own estimates.

Steckman cautions that highly valued companies can continue operating longer than outsiders expect, so he cannot say when attractive acquisition prices might appear. He estimates Anduril trades at about 10–14 times forward revenue and cites higher multiples elsewhere; Stebbings says he has seen still higher figures. These are conversational estimates, not verified transaction data.

Steckman identifies space as an area of interest, spanning ground systems and orbital technology. He sees a gap between traditional defense programs and commercially adjacent offerings from SpaceX, and says Anduril already uses partnerships, in-house work and integration in the area. His market-gap thesis reflects the company’s strategic interest.

Asked about competitor Helsing, Steckman says all competitors deserve respect and warns that technological hubris can cause a company to overlook serious rivals. He says he tries to provide a check on that tendency within Anduril’s leadership team. The exchange follows his earlier warning against assuming a technology has never existed.

Stebbings asks why Anduril does not raise more money at a higher valuation and buy the companies it wants; he refers to reported fundraising figures rather than confirmed terms. Steckman says the team wants to build an enduring public company, protect its valuation and avoid taking risks it cannot support. He says a mistake can cost Anduril both money and contracts.

On a possible listing, Steckman says public-company status can bring additional trust in national security work. He expects an IPO to be a couple of years away, not immediately, and says Anduril is investing in internal systems. His comments express an aspiration and approximate timing, not a promised transaction.

Steckman says about a quarter of Anduril’s 20 products are in meaningful production and generating money for the business, while others remain in development or low-rate production. He says more must move up the product curve before a listing. The proportions and economic condition of the products are company-reported and not independently audited here.

He estimates a traditional defense product may take seven to ten years to reach production, whereas he says Anduril aims for three to five. He cites Roadrunner as taking roughly 24 months from concept to a fielded system. These are his estimates and company claims; no comparable methodology or independent performance evidence is provided.

Steckman says Roadrunner required more than $100 million before generating revenue. Because development becomes expensive quickly, he says the company tries to stop weak concepts before entering the steepest part of the spending curve. He describes large sums as bets that products will eventually generate returns, not guaranteed outcomes.

As an example of a mistake, he recalls an early aerial system that consumed an estimated $20 million to $40 million before the company realized it lacked the right building blocks, team and understanding of how a customer would buy it. He thinks the loss was too large but also believes the experience contributed to subsequent aircraft work.

Asked about decisions he never made, Steckman says he has difficulty separating failed decisions or indecision from what came later. He wants to learn from errors and make the next decision rather than dwell on regret. His retrospective account cannot establish that the later project depended on the earlier failure.

In the quick-fire round, Steckman says he believes autonomous systems could eventually perform every military mission now served by existing technology, while emphasizing that current technology remains early. That is a long-term prediction by an executive whose company builds autonomous systems, not proof of present capability or a judgment that every use would be lawful or appropriate.

He returns to the risk of startups built around a single major program: if they lose the award, the business may struggle. He believes defense companies need breadth because of how procurement works. As with his earlier drone prediction, this is his commercial judgment rather than a universal rule for every supplier.

For people entering the workforce, Steckman emphasizes the community around them and the colleagues who provide opportunities and responsibility. He advises those leaving military service to speak with several companies, choose a starting point and change course if it proves unsuitable. His advice reflects his personal experience, not a prescribed career path.

Stebbings asks which Anduril co-founder Steckman would take to a hypothetical next company. Steckman jokes that he has no intention of founding another business and eventually gives a lighthearted answer. He says Palmer Luckey’s public profile helps build trust and relationships in a highly regulated industry, a positive assessment from a colleague.

The interview closes with Stebbings thanking Steckman and saying the discussion differed from what he had expected. Steckman says he listens to the podcast and appreciated the conversation.