Great Science Isn't a Startup

You can publish in Nature and still not have a company.

That sentence irritates a lot of brilliant founders, so let me defend it. A startup is not a technology. It's a particular way of operating under uncertainty, a set of instincts about customers, money, focus, and learning that has very little to do with whether your science is sound. Eric Ries, in The Startup Way, catalogues these instincts as a kind of startup DNA: small cross-functional teams, customer-first thinking, metered funding, a culture that learns fast and treats failure as data.

Here's the part Ries doesn't dwell on, but every science-led founder lives: a university spin-off inherits only half of that DNA.

The vision is there. The team is small and forced to be cross-functional. But the commercial instincts, the half that decides whether the science ever reaches anyone, are usually missing. Not because the founders aren't capable. Because they were trained for a different game, and trained extremely well.

Lab DNA and market DNA are not the same code

A research career installs one operating system. It optimizes for being right: rigorous, defensible, peer-reviewed, repeatable. That is exactly the system you want pointed at a hard scientific problem.

A startup runs on different firmware. It optimizes for being useful, fast.

The two systems share some vocabulary: "experiment," "data," "results", which is precisely why the switch is so easy to underestimate. A spin-off founder isn't lacking intelligence or drive. They're running lab firmware in a market environment, and wondering why the machine stalls.

Four places the two systems diverge matter more than the rest.

1. The lab asks "is it true?" The market asks "who cares?"

This is the big one. Academic training makes you technology-first by design. The instinct is to perfect the thing, then look for where it might apply. So the spin-off pitch opens with the breakthrough, the mechanism, the efficiency gain, the publication, and treats the customer as a downstream detail.

The market runs the question in reverse. It doesn't ask whether your technology works. It asks whose problem it solves, and whether they care enough to pay. A spin-off can have a genuinely superior technology and still build a company nobody needs, because "superior" was measured against a benchmark in a paper, not against what a buyer is actually trying to get done on a daily basis.

The fix isn't to abandon the science. It's to put one customer conversation in front of every technical decision. Not a survey. A conversation, with someone who has the budget and the pain. You'll learn in a week what a year of refinement can't tell you.

2. The lab tracks readiness. The market tracks demand.

Ask a spin-off how it's doing and you'll often hear about technology readiness level, patents filed, papers accepted, the next grant milestone. These feel like progress. They're the deep-tech version of what Ries calls vanity metrics: real numbers that don't predict whether anyone will buy.

The metrics that actually predict a company's future are uncomfortable because they involve other people's behaviour. Did the customer come back? Did they pay? Did they introduce you to someone? Did the pilot convert? These are leading indicators, they move before revenue does, and they tell you whether you're building something the market will pull, or pushing something it tolerates.

You don't need a dashboard team to start. You need to write down, this week, the two or three signals that would genuinely change your mind about your direction, and then go generate evidence on them.

3. Grant funding trained you for the wrong financial game

This is the trap nobody warns European founders about.

Grant funding (Horizon Europe, EIC, regional instruments) behaves like what Ries calls entitlement funding. You compete hard, you win, and then the money arrives on a schedule against a workplan you wrote eighteen months ago. Milestones are deliverables-to-a-plan. The discipline is executing what you promised.

Venture and commercial traction run on the opposite logic: metered funding. You get a small amount, you're free to spend it, but the next tranche depends entirely on what you learned with the last one. The discipline is changing your mind when the evidence says so.

A founder fluent in grants carries the entitlement instinct into a metered world. They build to the plan, defend the plan, and treat a pivot as a failure to deliver, when in a startup the pivot is the deliverable. Public money is a genuine advantage for a science-led company; it can fund the years a startup spends without profit. But it quietly trains a reflex that the commercial side will punish. The two funding worlds need to be run together, on purpose, by someone who can hold both logics at once.

4. Academia rewards pedigree. The market rewards results from limited resources.

In research, signal comes from pedigree: the journal, the lab, the principal investigator's name, the institution. It's a reasonable proxy when you can't evaluate the work directly.

Buyers and investors use a different proxy. They infer your quality from what you can produce with almost nothing, the pilot you closed on a shoestring, the partnership you talked your way into, the early customer who vouches for you. A spin-off that leads with credentials in a room that's reading for traction is answering a question nobody asked. Your CV got you the science. It will not get you the company.

What you already have... and where it leaks

The honest news is that two pieces of startup DNA are already in you.

You have vision, often in surplus. That's a real edge. It's what lets a startup pivot strategy without losing its way. The risk is the inverse: a vision so fixed to a specific technical approach that you can't separate the mission from the method, and so you refuse to pivot when the market asks you to.

You have a small, cross-functional team by necessity, everyone does everything. But a small team only compounds when it's pointed at one or two ruthless questions. At the start, you really only have two: how do we get the next customer? and how do we get the next partnership that makes us credible? Most stalled spin-offs aren't short of talent. They're spreading a two-person team across ten priorities and calling it ambition. Focus isn't about clearer goals. It's about subtraction.

The DNA is installable

None of this means becoming someone you're not. The science and the vision are your genuine advantages, and a real commercial partner protects them rather than flattening you into a generic startup.

It means installing the missing half deliberately: putting the customer ahead of the technology, tracking signals that predict demand instead of readiness, running grant money and commercial money on their two different logics, and pointing a small team at the few questions that actually move the company.

That's the work. It's learnable, it's fast when it's done well, and it's the difference between research that stays in a building and a company that reaches the people the research was meant to serve.