Healthcare venture is not one asset class but several, each with a distinct value engine, regulatory shape, and capital rhythm. Therapeutics, meaning drugs and biologics, win or lose on FDA approval built from years of clinical trials, and are the most capital-intensive. Medical devices reach market through the risk-tiered 510(k), De Novo, or PMA pathways, generally on shorter timelines and lower capital. Diagnostics create value by changing a clinical decision, and today many reach patients as laboratory-developed tests overseen under CLIA rather than through device approval. Regulated digital health, meaning software that makes or informs a medical claim, is regulated as a device when it does. A founder’s modality choice sets the timeline, the capital plan, and the definition of a fundable pre-seed round before the first hire is made.
Why does the therapeutics, device, diagnostic distinction matter before anything else?
Founders often describe their company by the disease it addresses. Investors, particularly at pre-seed, first sort by something more structural, which is the modality. A company is building a therapeutic, a device, a diagnostic, or a piece of regulated software, and that choice, more than the therapeutic area, determines how long the road is, how much capital it will consume, what evidence proves the company works, and what a responsible first round should buy.
The reason is that each modality answers to a different regulator, or a different arm of the same regulator, using a different logic. A drug must prove it is both safe and effective in humans before it can be sold. A device must clear a risk-tiered gate that often turns on comparison to something already on the market. A diagnostic frequently reaches patients through the laboratory system rather than the device-approval system at all. Software is regulated by what it claims to do, not by the fact that it is software.
At Sonnerie we read the modality first because it tells us what a dollar at pre-seed is actually buying. The same check that funds a decisive early experiment in one modality funds only a fraction of a preclinical program in another. The map below is the one we use, and the one we think founders should use when they decide what kind of company they are starting.
What are therapeutics, and how does a drug create value?
Therapeutics are products that act on the body’s biology to prevent, treat, or cure disease. The category spans small molecules, biologics such as antibodies and proteins, cell and gene therapies, and newer modalities such as RNA. Value in therapeutics is created almost entirely by de-risking, meaning each experiment that survives moves the asset from a speculative idea toward a probable medicine, and the value inflects sharply at the moments where human data replaces animal data and where efficacy replaces mere safety.
The regulatory path is the most demanding in medicine. A therapeutic moves from discovery through preclinical work into an Investigational New Drug application, then through Phase 1 for safety, Phase 2 for early efficacy and dosing, and Phase 3 for confirmatory efficacy, before a New Drug Application or Biologics License Application is filed for approval. There is no substantial-equivalence shortcut. The asset is approved or it is not, and industry data has long shown that the majority of candidates entering clinical testing never reach approval.
Timelines are long, commonly a decade or more from discovery to approval, and capital intensity is the highest of any modality, because clinical trials at scale are expensive and cannot be meaningfully compressed. This is why therapeutics companies are typically financed in large, milestone-gated tranches, and why the equity story is written around a small number of value-inflecting readouts rather than around near-term revenue.
What does a fundable pre-seed therapeutics spinout look like?
Because the road is long, a pre-seed therapeutics round is not buying a product. It is buying a decisive early experiment. The strongest university therapeutics spinouts we see have a clear mechanistic hypothesis, a defined lead or tight series of candidates, a credible line of sight to the clinic, and clean intellectual property, usually an exclusive license from the originating institution with the composition-of-matter position understood.
What separates a fundable asset from an interesting paper is the quality of the decisive experiment the round funds. We want to know the single result that would most cheaply confirm or kill the thesis, and we want the round sized to reach it. A strong pre-seed therapeutics founder can name that experiment, its cost, and its timeline without hedging, and treats a clean negative as a legitimate and valuable outcome rather than a threat.
What are medical devices, and how is the regulatory path shaped?
Medical devices are instruments, implants, and machines that act through physical or mechanical means rather than through chemical action on the body. The category runs from tongue depressors to surgical robots, and value is created through engineering, clinical validation, and, critically, adoption into a clinical workflow. Unlike drugs, devices iterate, and a strong device company often improves the product across generations while the core clearance stands.
The regulatory path is risk-tiered into three FDA routes. The 510(k), or premarket notification, is for devices that are substantially equivalent to a legally marketed predicate device, and it is generally the fastest and least resource-intensive route. The De Novo pathway, established under the FDA Modernization Act of 1997, is for genuinely novel low-to-moderate-risk devices that have no predicate, and a De Novo grant establishes a new Class I or Class II classification that later devices can use as their own predicate. Premarket Approval, or PMA, is the most rigorous route, reserved for high-risk Class III devices, and it requires clinical evidence of both safety and effectiveness alongside manufacturing inspection.
Timelines are generally shorter than for drugs, often measured in a few years rather than a decade, and capital intensity typically sits well below therapeutics for the 510(k) and De Novo routes, though a PMA program with a pivotal trial can approach drug-like cost and duration. The pathway a founder targets is therefore a first-order financial decision, not a regulatory footnote.
What does a fundable pre-seed device spinout look like?
A strong early device company is defined as much by its regulatory strategy as by its engineering. The best founders can state, at pre-seed, which pathway they intend, what predicate they will cite if any, and what evidence the FDA will expect, and they have pressure-tested that view with regulatory counsel rather than assuming the easiest route.
Beyond the pathway, we look for a real clinical need that the current standard of care handles badly, a working prototype or a clear path to one, and early signal that clinicians will actually change behavior to use it. Reimbursement is the quiet killer in devices, so we value founders who are already thinking about how the device gets paid for, because a cleared device that no payer covers is a science project, not a business.
What are diagnostics, and why is their regulation different?
Diagnostics are tests that detect, measure, or characterize disease, from a molecular assay to an imaging-derived score. Their value is created not by treating anything but by changing a clinical decision, so a diagnostic is only as valuable as the action it enables, and the sharpest question for any diagnostic is what a clinician does differently because of the result.
The regulatory picture is distinctive and, at the moment, unsettled. Many diagnostics reach patients as laboratory-developed tests run within a single certified lab, and those labs are overseen under the Clinical Laboratory Improvement Amendments, or CLIA, administered by CMS, which governs laboratory quality rather than approving individual tests. The FDA issued a 2024 final rule to bring laboratory-developed tests under its device framework, but a federal court vacated that rule in 2025, the FDA did not appeal, and the agency subsequently rescinded it, leaving oversight with CLIA for now. A separate legislative proposal, the VALID Act, has been discussed in Congress as a possible future framework. Diagnostics sold as packaged in-vitro diagnostic kits, by contrast, are regulated as devices through the 510(k), De Novo, or PMA pathways.
The practical consequence for founders is that a diagnostic can often reach the market faster and more cheaply than a device or drug by launching as a laboratory-developed test through a CLIA-certified lab, while accepting real uncertainty about the future rules of the road. Timelines and capital intensity are generally lower than therapeutics and comparable to or below devices, but the reimbursement and clinical-adoption challenges are just as decisive.
What does a fundable pre-seed diagnostic spinout look like?
The best diagnostic founders lead with clinical utility, not analytical performance. Strong sensitivity and specificity are necessary but not sufficient, and we push hard on the decision the test changes, the clinician who orders it, and the evidence that acting on the result improves outcomes or lowers cost. A test that is accurate but changes nothing is not investable.
We also want a clear-eyed reimbursement plan, because coverage and coding are often the true gating step for a diagnostic rather than the science. Founders who understand the path to a CPT code and payer coverage, and who can articulate why the current laboratory-developed-test and CLIA route fits their launch while staying alert to regulatory change, are markedly ahead of those treating the test as a pure technology achievement.
Where does regulated digital health fit?
Digital health spans a wide range, and the regulatory question turns on a single distinction, which is whether the software makes a medical claim. Wellness and administrative software generally sits outside device regulation. Software that diagnoses, treats, or informs a clinical decision is regulated as a medical device, often called Software as a Medical Device, and travels the same 510(k), De Novo, or PMA pathways as physical devices, scaled to its risk.
Value in regulated digital health is created through clinical validation and integration into care, and the modality carries a distinctive risk, which is that software updates continuously while regulatory clearances are anchored to a specific version and intended use. The strongest founders treat the regulatory intended-use statement as a design constraint from day one rather than something to reconcile later, and they understand that, as with diagnostics and devices, reimbursement and workflow adoption usually matter more than the model’s raw performance.
How does Sonnerie evaluate across modalities?
We hold every company to the same underlying question, which is whether this is a genuine signal or noise dressed as science, but we apply it through a modality-specific lens because the evidence that constitutes signal differs by category. For therapeutics we weight mechanism, IP, and the decisiveness of the funded experiment. For devices we weight regulatory pathway clarity, prototype maturity, and reimbursement logic. For diagnostics we weight clinical utility and the decision the test changes. For regulated software we weight intended-use discipline and integration into real care.
Across all categories we look hardest at three things that do not depend on modality. The first is whether the pre-seed round is correctly sized to reach the next real inflection rather than merely to survive. The second is whether the founding team, usually emerging from a university lab, has the operating instinct to translate a discovery into a company, which is a different skill from making the discovery. The third is IP and institutional cleanliness, because a spinout with an unresolved license or a contested inventorship question carries a risk no amount of science can offset.
The map matters because the biggest early mistake in healthcare venture is applying one modality’s playbook to another, sizing a diagnostic like a drug or a drug like a device. Naming the modality honestly, and financing it on its own terms, is where the discipline of building in this field begins, and it is where we start every conversation.
Frequently asked questions
What is the main difference between therapeutics, devices, and diagnostics?
Therapeutics act on the body’s biology to treat disease and must prove both safety and effectiveness through clinical trials before FDA approval. Medical devices act through physical or mechanical means and reach market via the risk-tiered 510(k), De Novo, or PMA pathways. Diagnostics detect or measure disease and create value by changing a clinical decision, and today many reach patients as laboratory-developed tests overseen under CLIA rather than through device approval.
What are the 510(k), De Novo, and PMA pathways?
They are the three FDA pathways for medical devices, tiered by risk. A 510(k) clears a device that is substantially equivalent to an existing predicate device and is generally the fastest route. The De Novo pathway is for novel low-to-moderate-risk devices with no predicate and creates a new device classification. PMA, or Premarket Approval, is the most rigorous route, reserved for high-risk Class III devices and requiring clinical evidence of both safety and effectiveness.
Are laboratory-developed tests regulated by the FDA?
As of 2026, most laboratory-developed tests are overseen under the Clinical Laboratory Improvement Amendments (CLIA), administered by CMS, rather than through FDA device approval. The FDA issued a 2024 final rule to regulate these tests as devices, but a federal court vacated the rule in 2025, the FDA did not appeal, and the agency later rescinded it, leaving oversight with CLIA. Future regulation could change through legislation such as the proposed VALID Act.
Which healthcare modality is the most capital-intensive?
Therapeutics, meaning drugs and biologics, are the most capital-intensive and slowest, commonly requiring a decade or more and large milestone-gated financing to move through clinical trials to approval. Devices and diagnostics are generally faster and less capital-intensive, though a high-risk device pursuing a PMA with a pivotal trial can approach drug-like cost and timelines.
When is health software regulated as a medical device?
Software is regulated as a medical device, often called Software as a Medical Device, when it makes a medical claim, meaning it diagnoses, treats, or informs a clinical decision. In that case it follows the same 510(k), De Novo, or PMA pathways as physical devices, scaled to its risk. Wellness and administrative software that makes no clinical claim generally falls outside device regulation.
What makes a healthcare startup fundable at pre-seed?
Across modalities, a fundable pre-seed healthcare company has a round correctly sized to reach the next genuine inflection point, a founding team with the operating instinct to turn a discovery into a company, and clean intellectual property and institutional licensing. Beyond that, the specifics differ: therapeutics need a decisive funded experiment, devices need a clear regulatory pathway and reimbursement logic, and diagnostics need demonstrable clinical utility.