Don't Try To Be Anduril
The biggest mistake defense startups make
The real point is that Anduril was not just a good defense startup, it was a rare attempt to build a new prime from scratch. That required a founding team that could invent multiple products, get them in front of buyers, survive long procurement cycles, and keep shifting bets until one scaled. Forterra and Saildrone look more repeatable because each starts with a narrower wedge, one autonomy layer for vehicles, or one long endurance maritime platform, instead of trying to replace Lockheed across categories.
-
Anduril got its early traction through Homeland Security and base security, not by immediately trying to win the Pentagon's biggest programs. It used smaller contracts to prove towers and counter drone systems, then climbed into larger programs. That crawl, walk, run path is very different from a startup claiming it will be the next prime on day one.
-
The hard part was not only building better hardware and software. It was packaging products in a way the government could actually buy, then funding years of iteration before budgets caught up. Ross describes Anduril as near n of 1 because most startups today are built around one program and one product, while Anduril ran a broad discovery motion across many possible products.
-
Forterra and Saildrone show the alternative playbook. Forterra sells one autonomy stack across military and commercial trucks, using partners instead of building whole vehicles. Saildrone operates unmanned ocean vessels for both science agencies and defense customers. Both are strong businesses, but neither is trying to become a full spectrum prime across land, air, sea, sensors, and command software.
The next wave in defense will come from companies that rhyme with Anduril, not copy it. The winning pattern is to enter through one concrete workflow, one buyer, and one budget line, then expand from there. As Anduril itself gets larger and more entrenched, the bar to building another broad defense prime only gets higher.