Anduril's Build-First Defense Model

Anduril is trying to turn defense from a custom engineering business into a product business. Instead of waiting for the Pentagon to pay for years of development and then earning a capped fee, it spends its own money to build drones, towers, and autonomy software first, then sells them as finished systems at fixed prices. That lets it move faster, bid lower, and keep much higher margins if the product works.

• The old prime model rewards spending. In cost-plus contracts, the government reimburses costs and adds roughly a 7% to 9% profit layer, so contractors are paid to manage giant bespoke programs. Anduril and similar product companies invert that by carrying the development risk themselves and only getting paid when a usable system is delivered.
  1 sacra 2 sacra
• This works best in categories where software and cheap hardware matter more than exquisite platforms. Anduril used gaming GPUs and off the shelf components to build border systems quickly, then showed it could cover the full U.S. southern border for about $250M versus Boeing's earlier $1B program that covered only 53 miles.
  1 sacra 2 sacra
• The tradeoff is brutal discipline. Early defense startups can get trapped in small services and research contracts that generate revenue but do not scale. Anduril's playbook was to start with small contracts, prove trust with operators, then climb toward larger programs while keeping the company centered on repeatable products rather than custom work.
  3 sacra 4 sacra

Going forward, the winners in defense will look less like giant one off programs and more like companies that can ship repeatable systems across drones, counter drone, maritime autonomy, and missile defense. As budget pressure pushes the Pentagon toward cheaper, faster, fieldable tools, Anduril's build first model becomes not just a pricing advantage, but a procurement advantage.