There's never been a more important time for European science.
Solutions derived from years of lab research are rapidly starting to solve problems in the physical world. Increasingly, science-based companies, led by scientist founders, are securing VC investment as funds look to moonshots to find the next wave of innovation.
Politicians are also promising more deeptech, more industrial sovereignty and more breakthrough companies built at home. So for investors and policymakers alike, backing the “right” research — research that is externally validated — is ever more critical.
But when it comes to financing those ambitions, we keep reaching for the wrong tools. If we are serious about building globally relevant science companies, we need to say something uncomfortable: startups should not begin with public or research grants.
That sounds counterintuitive in Europe, where the instinct is that if something matters enough, the state should step in early with funding.
Yet this well-intentioned impulse is distorting both how companies are formed and end up performing.
The validation problem
The earliest stage of a startup is about one thing: discovering whether the outside world cares about what you are building. A company only becomes viable when a customer or a partner, and then an investor, are willing to back it.
When founders rely on grants too early, that process changes. The question stops being whether the technology solves a real problem and becomes whether it satisfies a grant funding programme’s criteria.
Instead of building around a commercial bottleneck, teams begin building around eligibility requirements and funding cycles designed for research projects, not companies.
Over time, that shift changes behaviour. Scientists who might have become ambitious founders learn to behave like applicants, prioritising grant proposals over products, and programmes over markets.
We see the consequences across Europe. In some ecosystems, generous early grants mean founders remain part-time in academia while their startups drift.
They optimise burn rate instead of urgency, often reaching the market years late and untested as leaders. The pressure that produces strong companies is missing.
When public money goes wrong
In many cases, the system itself absorbs a large share of funding before it ever reaches the company. At some universities, overheads on research grants can reach as high as 50%, with significant portions diverted to administrative structures rather than scientific work.
The researchers who generated the idea often see only a fraction of the capital intended for it. What is positioned as support for innovation can quickly become administrative drag.
The damage is not limited to founders. When governments fund unproven companies too early, taxpayers become the risk capital.
Public funding is supposed to reduce risk for society. Used too early in startups, it can end up concentrating that risk on the public, a contradiction at the heart of many grant-led ecosystems. When governments become the first and largest backers of unproven companies, they often end up underwriting risk that the market itself has not yet validated.
Take Orbex, the Scottish aerospace company. It began with private backing, but as progress slowed, public funding increasingly filled the gap. Launch timelines slipped, and key milestones were missed. When it entered administration in February, taxpayers had become part of the risk capital.
The strongest science companies tend to emerge through the opposite sequence. They secure private capital first, which forces them to answer hard questions about markets, customers and execution. Public funding can then play a valuable role later, supporting research and scaling technologies that matter to society.
Europe’s obsession with match-funding schemes is another problem. Many founders approach investors with public funding already promised, asking for private capital simply to unlock it. In practice, this reverses the logic of venture building: companies seek markets only after programmes have validated them.
None of this means public support for science should disappear entirely. Governments have an essential role in funding research, infrastructure and strategically important technologies.
Bodies like Innovate UK, UKRI, and Horizon have helped support pioneering companies such as Oxford Quantum Circuits, Nanopore and Ceres Power. Grant funding is critical to de-risk foundational research when the technology is still unproven, there’s no clear revenue model yet and private investors see the risk as too high for example, in fields like quantum computing.
But, that said, the first test of a startup should rarely be a committee.
Proxima Fusion, among Europe's best-capitalised fusion companies, shows what works. It raised private capital first to validate its technology and commercial potential, before securing significant public support once the potential of its technology was demonstrated. Public funding works best when it accelerates proven companies, not when it is asked to validate them.
Building real businesses
Europe does not lack scientific talent. What it lacks is a culture that pushes that talent to build companies quickly, decisively and in the open market. Too often, we confuse support with progress and programmes with outcomes.
If we want globally competitive science companies, we should stop teaching founders how to navigate funding systems and start helping them build real businesses. That means more private capital earlier, stronger founder ownership and much greater scrutiny before public money is deployed.
Private validation, whether from customers or investors, should come before public funding.
If private capital doesn't come first, we risk creating an ecosystem full of grant-supported projects and far fewer companies capable of changing the world.
