A drone completes a successful demonstration. It identifies a target, transmits its coordinates and returns to land. For the startup that designed it, the hardest part may be just beginning: securing an order, financing a production line and delivering equipment that can be used far from the testing ground. In defense, technical performance opens a door; it guarantees neither recurring revenue nor profitability. The real step change in scale comes when the government moves from testing technology to buying it.
Looking ahead to September 2026, this question shapes the sector’s prospects. This analysis draws on facts known through June 2024; subsequent developments are considered as scenarios, not established events.
Dual-use technology opens doors, but not every budget
Since Russia’s invasion of Ukraine in February 2022, drones, analytical software, satellite communications and electronic warfare systems have become central to military discussions. The conflict has demonstrated the usefulness of equipment originating in the civilian sector, but also its limitations: jamming, vulnerable communications links, component availability and an adversary’s constant adaptation. A relevant innovation can quickly become inadequate if it fails to keep pace with battlefield practices.
This overlap attracts investors. A thermal camera can be used for industrial inspection as well as military surveillance; autonomous navigation software can support logistics as well as a reconnaissance robot. Dual-use technology thus promises access to several markets and, in theory, reduces dependence on a single buyer. But a shared technology does not mean an identical product. Ruggedization, cybersecurity, confidentiality and integration with existing systems create specific costs.
The civilian market therefore does not always provide a safety net. It has its own prices, competitors and sales networks. Maintaining two versions of a product can stretch a small team thin rather than secure its revenue. The right question is not simply whether a technology has two uses, but whether the company can serve both without funding two almost separate organizations.
A pilot project is not yet a market
Initiatives to bring startups and the military together proliferated before June 2024. In France, the Defense Innovation Agency, established in 2018, supports innovative projects, among other activities. At the European level, the European Defence Fund supports collaborative research and development. NATO has also launched the DIANA accelerator and a separate innovation fund. These initiatives facilitate access to funding, testing and military contacts.
They do not, however, replace an order for operational use. A grant pays for work; a trial tests a use case; a procurement contract commits a buyer to specified deliverables. Confusing these stages leads to an overestimation of a startup’s commercial strength. A prototype well received by a military unit may go no further because of a lack of available funding, shared priorities or a suitable procurement process.
Military sales involve several different timelines. Users want a solution to their immediate problem. Buyers must comply with applicable rules. Technical officials verify security and interoperability. Funding follows planning cycles and allocation decisions. For entrepreneurs, this sequence means paying salaries while the decision moves from one department to another.
Public procurement makes factory financing possible
A firm order changes the conversation with financial backers. It provides revenue visibility, helps in negotiations with suppliers and can make inventory financing easier. By contrast, a political announcement or a framework agreement with no guaranteed volume does not carry the same value. What matters is the portion firmly committed, its timetable and its payment terms.
Imagine a young company capable of assembling a few drones a week. To deliver several hundred aircraft, it must reserve components, buy tooling, recruit staff and document its processes. These expenses often precede final payment. Even with an order book, it can run short of cash if advance payments are insufficient or acceptance procedures are delayed.
This is where the structure of the contract becomes as important as its value: milestone payments, acceptance procedures, price adjustment provisions, maintenance and follow-on orders. A credible multiyear relationship can justify investment in manufacturing. A succession of small trials, each funded separately, risks keeping the company confined to a prototype workshop.
Produce, repair, endure: the other selection test
Software can be replicated at low cost; a complete military system is a very different proposition. Batteries, optics, motors and radio components come with their own lead times. Companies must also ensure consistent quality, maintain batch traceability and plan for substitutions when a part becomes unavailable. A component easily sourced for ten units can become a bottleneck when production scales up.
The true cost therefore exceeds the equipment’s sticker price. Training, spare parts, updates, documentation and repairs determine its operational availability. For armed forces, a less spectacular but maintainable device may be worth more than a brilliant solution that cannot be supported. For the startup, these services can generate regular revenue, but they also create lasting obligations.
Partnering with a major industrial company offers a route in: production capacity, knowledge of procedures, integration and support. It comes at a price, however. The startup may lose some of its margin, access to the customer or commercial freedom. Intellectual property, data rights and exclusivity must be negotiated before this dependence becomes irreversible.
Export decisions are not made in a spreadsheet
Selling to several countries seems a logical way to recoup development costs. Yet the addressable market is not simply the sum of military budgets. Exports of military equipment require authorization. Certain dual-use goods are also subject to controls, notably under European rules. The destination, end user and intended use can determine whether a sale can proceed.
The supply chain adds another constraint. Depending on their nature and legal framework, foreign technologies or components can trigger additional obligations, particularly US requirements. An architecture chosen to save time during prototyping may therefore complicate certain sales. Compliance must be considered from the design stage, not just when it is time to sign.
Government support then matters in two ways: as an initial reference customer and as the authority regulating access to markets. This sets defense fundamentally apart from consumer software. Growth depends as much on sovereign decisions as on sales effectiveness. A sound startup must incorporate this uncertainty into its forecasts rather than treat every export opportunity as likely revenue.
What next?
Looking ahead to September 2026, the decisive scenario would be less about multiplying demonstrators than improving continuity between trials, orders and production. More incremental procurement, backed by credible prospects and rapid operational feedback, could help new entrants without sacrificing oversight. For investors, the test would remain simple: who is buying, how firm is the commitment, and who is financing delivery? In defense, the winning startup will not necessarily be the one that impresses most, but the one that knows how to deliver, maintain and do it all again.
