Ignoring procurement and hardware realities
The biggest mistake defense startups make
The real reason defense tech looked uninvestable in 2017 was not lack of market size, it was a mismatch between venture math and defense reality. Startups had to spend heavily on hardware and product before revenue showed up, then survive procurement cycles that could take years, in a category with few visible startup scale outcomes. Early Anduril got funded because the market was enormous and the founding bench looked unusually capable of navigating that gauntlet.
-
Back then, defense tech was barely a recognized venture category. Palantir was still private, SpaceX had not yet created a broad startup playbook for defense, and many investors saw government procurement, hardware, and lumpy contract revenue as reasons to stay away entirely.
-
What changed was proof, not just sentiment. Anduril showed a startup could start with small contracts, turn field tests into paid programs, and compound into larger awards. That made defense feel less like custom services work and more like a real product business with venture scale upside.
-
The later hype cycle was reinforced by geopolitics and visible financial outcomes. Russia's invasion of Ukraine sharpened demand for drones, autonomy, and counter drone systems, while Anduril's rise from 2017 founding to $1B revenue in 2024 and $2.1B in 2025 gave investors a concrete winner to anchor on.
Going forward, more capital will keep flowing into defense, but it will concentrate around teams that can prove a repeatable path from prototype to program of record. The winners will look less like pure software startups and more like product companies that can finance upfront development, survive slow procurement, and turn battlefield relevance into durable budgets.