The Proof Gap

Healthcare isn’t short on good science. It’s short on affordable ways to prove the science works.

I spoke on a panel at Massive Bio’s “Massively Insightful” conference recently. Different questions, but the answers kept landing in the same place. How has funding changed? What defines the next generation? Where are the opportunities? On the surface, these sound like separate conversations. They’re not.

The bottleneck has moved. It used to be discovery. Now it’s validation.

The funding math broke

After 2022, venture capital pulled back hard. There’s been a resurgence lately, but it’s concentrated at the top. Firms like Andreessen and Lightspeed are absorbing the majority of venture deals, and when you’re deploying out of a $10 billion fund, a seed check into a $15 million valuation doesn’t move the needle. The math pushes them toward later, larger bets.

Meanwhile, emerging managers… the ones historically willing to take early-stage risk… are fewer and more cautious. The result is a widening gap at the bottom of the funnel. Early-stage healthcare companies are competing for a shrinking pool of capital from investors who actually understand the space and are willing to sit with the timeline.

This is where it gets ugly. Something like 90% of startups never raise enough money to get through clinical trials. The science may work. The math doesn’t. Getting a novel therapeutic to market can cost billions, and most companies flame out somewhere between “promising preclinical data” and “we need another $200 million to finish Phase III.”

The proof gap isn’t a funding gap, exactly. It’s a mismatch between how much it costs to validate something and how much early capital is available to attempt it.

The regulatory pile-on

The clinical trial system was designed for a world where therapies were mass-produced and deployed broadly. That makes sense when you’re approving a drug that millions of people will take. But the same regulatory framework gets applied to early-stage trials with tiny patient populations, often for diseases where the alternative is death.

That feels like a calibration problem, not a safety one. If someone has a late-stage disease with no approved treatment options, the risk calculus is fundamentally different. The current system doesn’t reflect that. Regulatory requirements that exist to protect large populations end up blocking access for the patients who have the least to lose and the most to gain. Companies like Massive Bio are working on this from the patient side, using AI to match cancer patients with clinical trials they’d otherwise never find. That’s a meaningful step, but it also highlights how broken the discovery process is… patients shouldn’t need an AI platform to find a trial that could save their life.

I think we’re going to have to get more honest about this. Loosening requirements for early-stage trials and late-stage patients isn’t reckless. It’s acknowledging that a one-size-fits-all framework creates its own kind of harm… the kind where a promising therapy exists but can’t reach the person who needs it.

Where the proof gets cheaper

If the core problem is the cost of validation, then the most interesting opportunities are the ones that reduce it.

Reimbursement literacy is one. The best healthcare companies I’ve seen treat CPT coding and payer pathways as a core product feature, not an afterthought. A lot of these companies are built by brilliant scientists, and the science is genuinely impressive. But “our TAM is massive” is not a business plan. You need a pathway to revenue that a payer will actually recognize, and building that into the product from day one compresses the distance between validation and commercialization.

Decentralized clinical trials are another. Running trials internationally rather than concentrating everything in the US can dramatically reduce cost and timeline. Kazakhstan, for example… and I know that sounds like a punchline, but hear me out… is a tier-1 digital economy with a centralized healthcare system and a regulatory framework that was essentially modeled after the FDA. You can run trials there meaningfully cheaper and faster than in the US, with comparable rigor. The infrastructure exists. You just have to look for it. There are environments around the world where the conditions for running lower-cost trials are already in place, and the companies that figure out how to access them will have a serious structural advantage over the ones insisting on doing everything domestically.

The real filter

If you’re investing in early-stage healthcare, stop leading with science quality as your primary filter. The science is table stakes. It’s getting better almost everywhere. What separates companies that make it from companies that don’t is their validation cost structure… how cheaply and quickly they can generate the proof that unlocks the next dollar.

Every diligence conversation should start with “show me your path to proof and what it costs,” not “tell me about your mechanism of action.” The mechanism matters, obviously. But a company with solid science and no affordable path to validation is just an expensive lab. The discovery layer is thriving. The validation layer is where everything stalls… and where the best investors and founders are quietly building an edge.