As someone who has tested new military technology, both on the front line and in more benign training environments, I know that it doesn’t always work the first time. Or sometimes it works, but not in the way you expected it to. Or wanted it to.
I’ll come back to that idea in a moment.
Over the weekend I read an interesting piece by Chris Miller – A Competitive Assessment of US Software — from 1984. He points out that the top software companies from that year are almost unrecognisable to the modern reader.
That’s because the industry’s explosive revenue growth was being driven by a major business model shift. The leading software players of 1984 were losing ground to new upstarts.
Before the late 1970s, software was largely sold to a specific company to enable its unique computing needs. Software consultants would write bespoke programs to manage, for example, airline reservations or banking transactions. They would sell them to one, or at most a handful of companies. The revenue would cover the costs of a well-paid professional workforce, plus some margin on top.
That pricing model may sound familiar to people who work in defence procurement. It’s called ‘cost plus’ and it’s the way that defence primes price their work for Ministries of Defence. Put simply, they’re permitted by the government to earn a certain amount of profit above the costs they incur. Primes have adapted to that model because that’s how the government has required them to work.
A cottage industry has even been built around winning the tenders that MOD puts out – writing a proposal in just the right way and pricing the work at just the right price are skills acquired after years of experience. People who understand that process are coveted employees within the industry.
However, new defence tech startups are inverting the model. First they raise money, largely from venture investors like my own firm, Expeditions, then they use that to develop products that they subsequently sell to a wide range of customers. They design and develop the products first, then they find customers.
Just like packaged software replaced bespoke software in the 1980s, defence tech startups are creating a new class of product which can be sold en masse. They’re taking the software development approach to new weapon systems.
The idea is to benefit from economies of scale, create a standardised product – cheaper than a bespoke solution – and get that product to market quickly. That last part is crucial.
In recent days the European defence tech market has been reacting to a damning article in the Financial Times which criticises German defence tech startup Stark. Its flagship drone, Virtus, is accused of failing to hit a single target in four attempts. For Stark’s competitors that may make for enjoyable reading. But the article fails to understand a broader shift that is taking place.
Before I go any further, for avoidance of doubt, I myself am not an investor in Stark, but I do back many companies in the same space. My opinions here, my stance on the situation, and my main motivation for investing in defence tech all come from my years of service in the military – an experience which has shaped the way I view the world.
The mentality of packaged software developers has always been ‘if you’re not failing, you’re not moving fast enough’. Their methodology is to test and iterate. Ideally in as realistic a setting as possible, to simulate real world environments. If the product isn’t perfect then they send an update.
Similarly, the new breed of defence tech startups is in the business of testing and iterating. And sometimes failing.
That is possible because companies like Stark are developing ‘software defined hardware’. In classic terms that refers to physical hardware components, like computer chips, whose function and internal logic can be reconfigured or changed after they have been manufactured, using software instructions.
In defence tech terms that means the hardware is built and then gets better as the software is updated. Just as Teslas routinely connect to a data network, download new firmware, then install updates, so defence tech companies push improved software to existing hardware. In some cases it will already be in use on the front line.
Why don’t they perfect it before selling it? The goal is to get product in the hands of end users as fast as possible. The reason for that is largely commercial – to beat their competitors. But there is also a geopolitical narrative behind the urgency. Defence ministries across Europe are looking to buy this new technology, born in Ukraine, after decades of underinvestment in military kit. Drones won’t replace everything. But they will form an important part of a modern capability.
Governments will need to get more comfortable managing the messy unknowns of innovation, where iteration and imperfection are part of the process. At the same time, startups must adopt what their predecessors understood well: the slow, deliberate work of building trust, reliability, and customer confidence. Innovation cannot mean abandoning rigour; it must mean expanding it to include speed.
Europe now finds itself on two front lines at once: one against a resurgent Russia, and another in the global race to redefine defence technology.
Mistakes will be made and innovation is never without cost. But if this new model can be managed, and if speed can coexist with trust, then Europe has the chance not only to rearm, but to rebuild an industrial base capable of producing the next generation of global leaders in defence technology, just as America once did in software.
Samuel Burrell is a Partner at Expeditions. He started his career in the military, where he was an officer in the Royal Marines. After leaving the military he joined a startup called ZipJet, before studying for an MBA at INSEAD. He has since been in venture capital, at Octopus Ventures, NKB Group and the National Security Strategic Investment Fund (NSSIF), a UK government fund focused on investing in national security and defence technologies. He joined Expeditions at the beginning of 2025 and is the lead for UK and US investments.
