Welcome to the 86th edition of Deep Tech Catalyst, the channel from The Scenarionist where science meets venture!
Today, we’re joined by Tom Miller, Founder and Managing Partner at GreyBird Ventures!
Together, we unpack what separates good ideas from investable companies—and why most business plans in diagnostics get it wrong.
In this edition, we explore:
Why the diagnostic opportunity starts with clinical problems—not markets
What early-stage founders must prove before talking about TAM
How to test for real-world clinical validity—not just performance curves
Why capital efficiency and risk reduction matter more than projections
What founders get wrong about adoption, KOLs, and exit dynamics
Whether you're a scientist translating lab research, an operator in the trenches, or an investor navigating clinical risk, this conversation offers a practical blueprint for building companies that matter—at the frontier of diagnostic innovation.
Let’s get into it 🧬
✨ For more, see Membership | Deep Tech Briefing | Insights | VC Guides
BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE
The Diagnostic Opportunity Is Not About Markets, It’s About Solving Hard Clinical Problems.
When evaluating early-stage diagnostics, most people begin with market categories. They talk about liquid biopsy, omics platforms, AI-powered imaging, or molecular diagnostics. These categories, while useful for mapping the field, are almost never the right starting point when it comes to building an investable company.
A more grounded approach begins with the patient—not with the technology. The first and most important question is: what is the clinical problem being solved, and how important is that problem in the real world of care delivery?
Some of the highest-potential opportunities in diagnostics have nothing to do with category hype and everything to do with solving specific, high-impact bottlenecks in clinical workflows.
The examples are concrete. Stroke triage in the emergency room, where every minute of delay results in irreversible brain damage. Sepsis detection in hospitalized patients, where rapid pathogen identification can mean the difference between life and death. Early cancer detection—real early detection—with enough specificity to avoid overtreatment and false positives.
These aren’t just interesting ideas. They are clinical pain points where a better diagnostic could redefine outcomes.
In this view, technology serves the problem—not the other way around. It's not about whether a new assay uses CRISPR, microfluidics, or AI. It’s about whether it enables a faster, more accurate, more actionable decision in a real clinical setting. If it doesn’t meaningfully improve the patient journey, it’s unlikely to matter.
The Shift from Technology to Unmet Clinical Needs
This is also where investor priorities shift. Rather than chasing broad market opportunities with theoretical value, the focus moves to what some define as “grand challenges” in diagnostics—problems whose solution would generate immediate clinical, operational, and economic impact.
The emphasis is on specificity, speed, and clinical utility—not just technical novelty.
In diagnostics, this mindset changes how founders need to frame their value proposition. Instead of leading with “what the test does,” they need to articulate what changes when the test exists. A good list of questions to prioritize could be:
What is the clinical decision that shifts?
What downstream actions are enabled or avoided?
Who benefits, and how quickly?
Many diagnostics can produce elegant results in the lab but fail to meet a clear decision point in the clinic. If a test does not change what a doctor does next, it risks being irrelevant—regardless of its scientific merits.
The strongest investment cases emerge when a founder starts with a well-defined clinical outcome, identifies the diagnostic gap that blocks it, and demonstrates how their approach reduces time, cost, or uncertainty in reaching that outcome. From there, everything else—technology, business model, regulatory plan—flows as a means to that end.
Founders Must Start by Building a Risk Map, Not a Revenue Model
In early-stage diagnostics, most business plans fail not because they aim too low, but because they assume too much. The projections are almost always wrong—in both timing and magnitude. That’s not a sign of incompetence. It’s the nature of the space. What separates viable ventures from speculative ones is not their ability to predict the future, but their ability to de-risk it.
A realistic investment case in diagnostics is not based on spreadsheets that show $200 million in revenue five years out. It's built on a credible narrative of what technical, clinical, and regulatory risks will be reduced with each tranche of capital. Investors want to understand what will be proven—not just what is promised—by the time the next funding round is needed.
You need to show real-world performance, clinical relevance, and a viable path to revenue—all with far less capital than therapeutic ventures typically raise. Capital efficiency isn’t optional. It’s fundamental.
Here are the questions every founder should be able to answer:
With the money you’re asking for, what specific uncertainties will be eliminated?
Can you demonstrate analytical validity with real biological samples?
Can you replicate performance outside controlled lab conditions?
Can you simulate how the test will behave in actual clinical settings, with human variability and imperfect workflows?
Early traction isn’t about users, revenue, or partnerships. It’s about reducing uncertainty.
If a test is still technically fragile, clinically unvalidated, or based on idealized sample cohorts, it’s not ready—no matter how impressive the technology looks on paper.
That’s why founders must spend disproportionate time on verification under real-world conditions. The most common failure pattern is a technology that performs well on paper but collapses under clinical pressure. Samples drawn from different institutions, handled by different people, stored in subtly different conditions—these variations are where diagnostics either prove themselves or unravel.
The role of capital at this stage is not to scale, but to validate.
Founders should use early investment to clarify the limits of the technology and stress-test the assumptions that underpin their thesis.
What breaks when the environment gets messy?
What assumptions no longer hold when the device is in the hands of an actual clinician, not a PhD in a lab?
This is also the phase where clarity of purpose matters. It's not enough to develop a technology that can do many things. The key is to focus relentlessly on the single most valuable problem it can solve first—and build a roadmap from that point outward.
If a founder can articulate how capital will turn unknowns into knowns—and how those knowns unlock the next stage of progress—they’re far ahead of most. That’s the foundation of an investable company.
The Power Law of Matter: Why Advanced Materials are Eating Digital Margins
From SaaS Logic to Commodity Logic. From advanced materials in photonics to fusion energy — the real profit pools are moving back into the atomic layer.
Technology Must Survive the Chaos of Real-World Conditions
Building a diagnostics company isn’t about showing that a test can work—it’s about proving that it does work, reliably, in the “unpredictable mess” of the real world. That gap between technical promise and clinical validity is where most ventures stumble. And it’s where capital is either well spent—or wasted.
A common trap among early-stage diagnostic startups is overconfidence in lab results. It's easy to get strong performance metrics when testing is done in-house, under controlled conditions, with carefully curated sample sets. But those results are often meaningless once the test leaves the lab and enters the clinic.
Founders routinely present data from “perfectly clean” comparisons—hundreds of healthy samples on one side, clearly diseased samples on the other. In these scenarios, almost any machine learning model or signal processing tool can achieve strong separation. But those signals may have nothing to do with the biology of disease. More often than not, they’re reflecting batch effects, storage artifacts, or site-specific biases.
When performance drops in real-world deployment, the issue isn’t the algorithm—it’s the assumption that early data was representative. Robust validation means testing on biologically diverse, operationally messy samples. Different patient populations, different institutions, different handling protocols. If a test can’t survive variability, it’s not ready.
Analytical validity—proving the test actually measures what it claims to measure—is just the beginning. The harder challenge is demonstrating clinical validity: that the test correlates with a meaningful clinical condition, in the settings where it will actually be used.
This is why validation must include behavioral proof, not just statistical performance.
Does the test change what a clinician does?
Does it lead to faster diagnosis, better triage, or different treatment choices?
Without that impact, even the most accurate test becomes irrelevant.
Capital Efficiency
Capital efficiency doesn’t mean cutting corners—it means focusing on what matters most. Founders should resist the urge to pursue broad claims, multiple indications, or platform generality in the early stages.
The goal is not to show how many things the test could do, but to prove one thing it does—better, faster, or cheaper than the current standard of care.
This focus is also what attracts the right investors.
Diagnostics is not a “big spend” field. It’s a space where outcomes must be proven incrementally, with lean resources and smart design. Investors want to see how each dollar reduces uncertainty and moves the company toward commercial validation.
A strong early-stage plan doesn’t promise exponential growth. It shows how clarity will be gained, one milestone at a time, until the path to adoption is undeniable.
Know the Physician. Know the Workflow. Know Who Loses if You Win.
Many diagnostics startups approach clinical adoption as if it were a purely rational process. They assume that if the test is accurate, it will be used. If it saves time or reduces costs, adoption will follow. But that’s not how healthcare works. In reality, the inertia of clinical behavior is one of the most underestimated barriers to success.
Changing Clinical Practice Is Harder Than Building the Technology
Physicians don’t change how they practice medicine just because a new test is available. They change only when doing so becomes easier, safer, more aligned with their workflow, or when not changing carries consequences.
Clinical behavior is deeply ingrained. It’s shaped by medical training, institutional norms, reimbursement constraints, and legal liability. The majority of practicing clinicians are not early adopters. They are risk-sensitive professionals with limited time. That’s why a diagnostic must not only be better—it must be frictionless.
Founders who don’t account for these dynamics risk building a solution in search of a user. It’s not enough to prove technical superiority. The diagnostic must integrate cleanly into existing pathways. It must reduce cognitive burden, not add to it. It must fit into the patient flow, not disrupt it. In short, it must work for the physician—not just for the data.
Key Opinion Leaders Aren’t the Market
Early-stage teams often rely heavily on key opinion leaders (KOLs) for feedback and validation. While valuable, KOLs represent the most progressive segment of the market. They thrive on novelty. They publish. They present at conferences. They are not typical.
Worse, KOLs tend to validate based on the potential of the technology—not its operational reality. They may support a diagnostic that requires a complex workflow, extensive training, or institutional change—things that mainstream clinicians will resist.
To build for real-world adoption, founders must go beyond expert enthusiasm and study the middle of the bell curve. Here is a list of thoughtful questions:
How do average physicians behave?
What are they worried about?
What are they incentivized to do or avoid?
These are the users that drive scale. These are the decision-makers who determine whether a diagnostic becomes standard—or stays niche.
Every Innovation Creates Resistance Somewhere
A diagnostics startup doesn’t operate in a vacuum. Every successful test disrupts an existing process, provider, or incumbent technology. Understanding who loses if your test wins is critical.
Will a new test reduce imaging volumes?
Undermine a pathology department?
Shift revenue away from a current vendor?
If so, founders should expect pushback—from clinicians, administrators, and competitors alike. Adoption requires navigating this resistance.
The same logic applies to strategic exit planning.
If a technology threatens an incumbent without offering them a path to participate, acquisition becomes less likely. On the other hand, if it makes them more competitive—by adding revenue, reducing cost, or filling a strategic gap—it becomes a natural target. In some cases, fear is a stronger motivator than opportunity. Knowing this dynamic in advance is part of building a company designed to exit.
The companies that succeed are the ones that understand how healthcare systems behave. They know who benefits, who loses, and what it takes to shift behavior at scale.
Why GPT-5’s “Underwhelm” Is Bullish for Deep Tech & more... | Deep Tech Briefing #73
Weekly Intelligence on Deep Tech Startups and Venture Capital.
The Founder Mindset: Obsession and Listening to People Who Disagree
Technology alone doesn't build a diagnostics company. The differentiator—especially at the earliest stages—is the mindset of the founder. This is not about charisma or technical brilliance. It’s about the mental posture required to navigate a long, uncertain, and unforgiving journey. And in diagnostics, that journey requires a very specific kind of cognitive duality.
Vision Is Essential. But So Is “Constructive Paranoia”.
The best founders operate in two modes at once.
On the surface, they communicate a clear and compelling vision. They can articulate how their technology could radically improve patient outcomes, reshape clinical practice, or shift the economics of care. This ability to project belief is essential—not just for fundraising, but for recruiting talent, building partnerships, and engaging stakeholders.
But behind that confidence sits a different force: strategic paranoia. The most capable founders are quietly obsessed with everything that could go wrong. They are constantly asking:
What if I’m missing something?
What if this assumption doesn’t hold?
What if this feature fails in the real world?
This dual posture—belief in the vision, fear of the risks—is not a contradiction. It’s a requirement.
Validation Should Be Earned, Not Sought
Founders under pressure often seek affirmation. They want to hear that their idea is strong, their data is good, and their pitch is compelling. Especially when fundraising, it’s tempting to listen only to those who agree. But this creates a dangerous filter—one where critical feedback gets drowned out by what sounds reassuring.
To build something real, founders must deliberately seek out dissent.
Talk to skeptics. Study the people who don’t believe in the approach. Find the clinicians who wouldn’t use the test. Learn from the investors who passed. These voices often surface issues that happy feedback will never reveal. And those issues—if ignored—become the reasons companies fail.
This mindset also accelerates decision-making. By surfacing objections early, founders can refine their strategy or walk away from dead ends before burning too much capital. In a space where capital is limited and timelines are long, knowing when to stop can be just as valuable as knowing when to push forward.
Some Ideas Don’t Deserve to Scale
Diagnostics is full of “zombie companies”—ventures that should have shut down years ago, but continue to limp along by raising small amounts of bridge capital or chasing grants. These companies often have talented teams, good intentions, and technically sound products. What they lack is traction.
The longer they linger, the more capital they consume and the harder it becomes to reset. In many cases, a faster failure would have been a better outcome—for founders, investors, and the market.
The truth is, not every diagnostic is commercially viable. Not every test can change behavior, command reimbursement, or achieve scale. Recognizing that early, and moving on, is a sign of maturity—not weakness. The ability to “fail well” and redeploy learnings into the next venture is often what defines long-term success.
Focus Is the Antidote to Uncertainty
In the end, the founder’s job is to reduce uncertainty, one decision at a time. That means knowing exactly what kind of problem they are solving.
Is it an analytical problem, where the test needs to measure something more precisely than existing tools?
Or is it a clinical workflow problem, where the goal is to influence decision-making regardless of the technology used?
The answer to that question shapes everything else—go-to-market strategy, pricing, validation design, and even exit dynamics. Founders who can answer it clearly—and build around it with discipline—are far more likely to reach the thresholds that matter.
They don't just build tests. They build companies that matter.
Deep Tech Investment Secrets: What Top VCs Do Differently - Part 1
Strategic patterns used by leading deep tech investors to get better deals, faster access, and lower downstream risk.
