Companies that should exist

The structural gap between what venture funds and what the world needs and why philanthropy might not fix it.

There is a class of companies that should exist but don’t because the economics don’t fit neatly into any existing capital structure.

These are companies that would work on problems of enormous consequence: biosecurity, diagnostics, antibiotics, climate, small scale domestic manufacturing. They’d have real customers, generate real revenue, and compound real value over time - they will even generate real cashflow. They just won’t return a venture fund. And so they don’t get built, and the problems they’d address remain unsolved.

At Convergent Research, we found a parallel of this type of problem in science. Important research bottlenecks - technically hard, coordination-intensive, sufficiently big, public-good-oriented - weren’t getting solved because no existing institution was designed to solve them. Not startups, not academic labs, not government agencies. So we built a new organizational form: the Focused Research Organization. FROs are purpose-built teams with fixed timelines and dedicated funding, designed to de-bottleneck an entire field by producing open tools, datasets, and platforms. Think - mini Hubble space telescope. We’ve proven that the model works. But FROs are designed for a specific shape of problem - one where the output is a public good.

I think that there is another new organization that needs to exist - companies that could sustain themselves through revenue, but that no existing capital structure will fund into existence.

Some of these companies are early stage deep tech companies, Raylene Yung with others investigated the shape of some of the capital shortfalls. Not all of the companies I’m talking about require R&D or are even really deep tech — some are primarily capital deployment or coordination problems. What unifies them is economics that no existing capital structure is meant to fund.

They serve an impact need. These companies would exist to solve problems that could affect millions of people. Common goods effects and market dynamics keep these from being addresses. Antibiotic resistance kills over a million people a year, and the development pipeline is dangerously thin - because the economics of bringing a new antibiotic to market are fundamentally broken. Lead exposure causes irreversible neurological damage in children - but lead poisoning is addressable - but the diagnostic tools to catch it early are inadequate in most of the world. Indoor air is a major vector for respiratory disease, and one example of the technology to fix it - far-UVC - exists but has no company building the distribution infrastructure at scale. Climate targets require decarbonizing industrial heat, which accounts for roughly 20% of global emissions, and nobody is doing it one sector at a time with the patience it requires. These problems are well-characterized, urgent, and solvable.

They’re real businesses. A clinical trial infrastructure company serving neglected diseases has paying customers - drug sponsors, foundations, government agencies. An industrial heat decarbonization company sells a service to manufacturers. A company distributing and installing far-UVC disinfection hardware in commercial buildings would sell their devices - though they will have to convince the market to exist. A diagnostics company, once it clears FDA and navigates reimbursement, has a durable revenue stream. None of these are money losing charities.

At scale, they could be big. Like many companies, they might start small. But eventually they could be big enough to deliver impact that matters. The impact of far-UVC comes not from just getting lamps into cruise ships, but from identifying the most important places for built-environment pathogen protection and going there.

They won’t return 100x. Venture capital needs a single investment to return the entire portfolio. That math strongly favors enormous markets, explosive growth, and winner-take-all dynamics. Consider Biobot Analytics, the company that built wastewater pathogen surveillance and became the CDC’s primary partner for national wastewater monitoring. The technology works and provides a value to public health. The company has raised over $40 million, landed government contracts, and expanded from COVID to opioids, influenza, RSV. It’s a real company with real revenue - but, by venture standards, a modest outcome after nearly a decade. Biobot works, just not with portfolio returning results. Diagnostics companies face the same structural challenge. Margins are small, few VCs want to fund diagnostic companies at seed or Series A, and exits are typically only 4–8x revenue for high-growth businesses and 2–4x for low-growth ones. The economics look even worse in global health, where the patients who need the product most are in countries that can’t pay developed-world prices - so the market shaping, volume guarantees, and cost optimization have to be designed into the product from the start.

They require upfront capital. These aren’t companies you bootstrap from a garage. Distributing and installing far-UVC hardware requires manufacturing capacity, building-by-building sales, and regulatory navigation before the first installation generates recurring revenue. Standing up a community-embedded clinical trial site network means credentialing dozens of sites before a single sponsor pays to enroll patients. Antibiotic development costs roughly $1.5 billion per drug, while annual sales for a new antibiotic average only about $46 million.

They often need market-shaping work alongside the company itself. Most capital would see this and get scared. We need investors who understand that there are parallel efforts to make these markets work. Advance market commitments - like Stripe’s $925 million Frontier initiative for carbon removal or the Gavi pneumococcal vaccine AMC that immunized over 760 million children - have showed that we can jumpstart markets.

They may also need regulatory groundwork and coalition-building. These companies only make sense when you account for the ecosystem being built around them. Nan Ransohoff has written about “general managers” to own problems end-to-end. These aren’t often problems VCs or founders decide to overcome on their own.

Who might fund this?

Usually not venture capital. Typical VC funds have 10-year lifespans and need power-law returns. The industry tolerates ~75% failure rates because a small number of investments return 20x or more. For diagnostics, some manufacturing, and many hardware businesses the upside is capped at 3–5x and the math for venture style portfolio returns doesn’t pencil.

It’s worth naming one thing that VC does well, it glamorizes companies and founders. The VC industry has built a cultural apparatus around the companies it funds that makes certain kinds of company-building feel inevitable and exciting. That apparatus is genuinely valuable and could be replicated here.

Private equity buys existing cash flows. The core PE model acquires majority stakes in businesses with existing cash flows, using leverage that those cash flows service. Some PE firms do fix-it-up or buy-and-build strategies, but even those start with operational businesses generating revenue.

Growth equity sits between venture and private equity. Growth equity investors take minority stakes in companies already generating meaningful revenue - typically $10M or more - and scaling toward profitability. The return profile (3–5x) is closer to what we’re talking about. But growth equity is looking for… well the growth stage, not seed stage.

Infrastructure finance funds large, predictable assets - roads, power plants, data centers - where revenue is contracted before construction begins. A pre-revenue pathogen surveillance utility doesn’t have those cash flows. That being said - this is another area ripe for impact innovation.

Loans and debt require either collateral or predictable revenue. There are SBIR and other federal loans - but they are few and far between and often too small.

Bootstrapping fails because the upfront capital requirements are too large, the domain expertise too specialized, and the market-shaping work too extensive for a solo founder on savings.

These companies are also too commercially viable for philanthropic capital — most foundations don’t fund for-profit entities.

To summarize - Too slow for venture. Too early for PE or growth equity. Too capital-intensive for bootstrapping. Too profitable for charity. These companies sit in a structural gap between capital categories, and so they simply don’t get built.

I’m not willing to accept that because these companies don’t exist today they can’t exist. Let’s create them. We need the right capital and a kick start.

We might need a name for these for-profit businesses built to solve problems that are consequential, revenue-generating, and structurally ignored by every existing capital category.

A natural question: if these are real businesses with real returns, why hasn’t a smart family office or impact investor already funded them? FROs didn’t exist before we built them at Convergent, but I think that they are impactful and needed. The problems were real, the solutions were tractable, the people who could do the work were out there - but nobody had named the category, defined what the organization should look like, or made it their job to build it. These companies don’t have a cultural zeitgeist. There’s no accelerator, no conference circuit, no pattern for a founder to follow. The opportunity has to be identified, packaged, and presented to the right capital with the right framing. That’s coordination work, and it doesn’t happen on its own.

I still have lots of open questions. Who identifies the companies and does the pre-investment diligence? Who recruits and backs the founders? Who manages the relationship with market-shaping partners — the foundations, the government agencies, the advance market commitment designers? Is this a holding company, a fund, a hybrid?

The right capital has a specific profile though. It’s patient, aligned, and might expect a return - just not a venture-scale one. Based on analogous hardware and infrastructure investments, I think that the target return profile at the fund level is something like 2.5–4x multiples at 15–25% IRR, with the willingness to hold for ten to twenty years or longer. That’s above public markets’ historical ~10% annualized — but it’s a fundamentally different game than venture.

The math that gets you there is the inverse of venture’s: with capped upside at 3–5x per company, you cannot afford 75% failure rates. A fund of fifteen companies where eleven return nothing and four return 3x yields roughly 0.8x. To clear a 15–25% IRR at the fund level, you need ~80% of companies to return meaningful multiples — which means the selection logic has to be fundamentally different. You’re not betting on “could this be a $10B company?” You’re asking: is this a real business with de-risked technology, a credible path to revenue in three to five years, and an operator who understands both the problem and the market?

The Right Capital Structures

Family offices. Unlike institutional funds constrained by fund cycles and LP expectations, a family office invests on behalf of a single family, often with a multi-generational time horizon. They can hold an investment for fifteen or twenty years without needing to exit. They can structure deals flexibly - equity, debt, convertible instruments, milestone-based tranches. A growing number are explicitly pursuing impact-aligned investments: Builders Vision has committed over $260 million to ocean-related ventures; Ceniarth has deployed more than $500 million in underserved markets.

Evergreen funds have no fixed liquidation date. They can hold investments indefinitely, recycle capital as returns come in, and avoid the forced exits that 10-year fund structures impose. Sequoia structured an evergreen fund in 2021, but Sequoia got away with that because they’re Sequoia - LPs trusted the brand. For the companies I’m describing, the LP base would look different: investors who are aligned because they care about the thesis and the impact.

Management companies and holding structures offer a different path. Instead of funding each company as a standalone startup, a single entity could incubate, launch, and operate a portfolio of these businesses under shared infrastructure - shared back-office, shared regulatory expertise, shared recruiting, shared institutional relationships. IAC has built and spun off more than ten publicly traded companies this way. Berkshire Hathaway holds companies permanently, benefiting from compounding value across a portfolio of durable businesses.

Blended capital stacks combine philanthropic and commercial capital in a single vehicle. Prime Coalition and Azolla Ventures have pioneered a model where roughly one-third catalytic (philanthropic) capital and two-thirds market-rate capital co-invest. The catalytic sleeve absorbs early-stage downside risk while the commercial capital brings discipline and follow-on capacity.

Program-Related Investments from foundations are another lever. PRIs allow foundations to deploy capital into for-profit entities while counting it toward their 5% annual payout requirement. They’re legally well-established but practically underused - most foundations have never deployed a PRI into early-stage hard tech. Building the intermediary infrastructure to make this comfortable for foundations is itself a coordination problem worth solving.

The Portfolio

If the capital problem is solved, the potential impact portfolio is large and varied.

In biosecurity: far-UVC disinfection hardware for buildings. Pathogen monitoring networks. Response coordination companies.

In diagnostics and global health: New lead diagnostics. Diagnostic platforms for neglected diseases that can survive the long march through FDA clearance and reimbursement. Real-world evidence infrastructure that enables regulatory approvals in low- and middle-income countries.

In clinical trials: a longitudinal participant registry for aging research that eliminates the cohort bottleneck. A digital biomarker validation company that gives the FDA accepted endpoints for geroprotector trials. Clinical trial site networks and shared infrastructure purpose-built for speed and scale.

In climate and energy: an industrial heat decarbonization company - about 20% of global emissions, unglamorous, one vertical at a time.

In science infrastructure: a shared equipment network that turns underutilized university instruments into a national resource. A geological data commons that digitizes a century of subsurface data and unlocks critical mineral exploration.

In manufacturing and defense: nuclear decommissioning robotics. A regulatory co-development firm for hard-tech startups. Onshoring critical production capacity.

Every one of these is a real business. Every one addresses a genuine bottleneck. Every one would generate revenue and, at maturity, cash flow. And yet almost none of them are being started. They seem pretty unglamourous in today’s venture ecosystem. Should Stripe have been glamorous? It processed payments. VCs decided it was, built a narrative around it, and a generation of founders followed. The same thing could happen here. The companies in this portfolio could be made to feel like the obvious place to build — the frontier where consequential founders go because the problems are real and the structural support is finally in place. VC won’t do that, because glamorizing a 3x return doesn’t serve their fund math.

With patient, aligned capital and operators who understand both the problem and the business, we can build companies that the current financial market has decided not to fund. We can do it without losing money - in fact, we can generate real, compounding returns.

---

Thanks to Ben Kuhn, Sasha Chapin, Zander Farkas, Tom Kalil, Adam Marblestone, Pamela Mishkin, Nan Ransohoff, Cate Hall