20 Percent Better Isn't Enough

The real moat in defense is not a modestly better widget, it is the ability to survive a slow, trust based buying process long enough to turn a breakthrough product into a funded program. Anduril won by selling products that were dramatically cheaper or more capable, starting with small contracts, then using live deployments and repeatable hardware plus software to climb from pilot budgets to programs of record. That is why slight performance gains rarely justify a switch.

• Defense buying behaves more like enterprise sales with extra gates. A startup needs an internal champion, working users, budget alignment, and a path through formal procurement. Even a clearly better system can take years to buy, so the improvement has to be big enough to justify the hassle.
  1 sacra 3 sacra
• Anduril paired step change products with a crawl, walk, run contract ladder. Early wins started in the tens or hundreds of thousands, then grew into a $12.5M Marine Corps contract and later a $1B scale program. That progression built trust as much as it built revenue.
  1 sacra 2 sacra
• This is also why product companies beat services firms in defense. Incumbent primes get paid to develop what the government specifies, which caps margins and slows iteration. Anduril self funded products, sold them closer to commercial rules, and scaled to an estimated $1B revenue in 2024 as counter drone demand accelerated.
  2 sacra 4 sacra

The next generation of defense winners will look less like niche component vendors and more like companies with a wedge product that is unmistakably better, then broad enough to expand across adjacent missions. As procurement slowly shifts toward buying finished systems instead of funding endless development, the companies that combine sharp product jumps with disciplined capture will keep taking share from the primes.