Non-Profit Startup

Most venture-funded startups are non-profits. Not legally, but materially — they take in money and do not produce a profit. Everyone involved knows this. The money gets spent, the company shuts down, and everyone moves on to the next one. This is described, without irony, as free-market capitalism. It is essentially a private subsidy system.

But here's what actually happens. They ruin markets. They have to be hyper-competitive, winner-takes-all. So about three quarters of VC-backed startups never return a cent to their investors. Of the rest, most returns come from a tiny handful of winners. And that kind of works. For investors. It works the way a casino works: the house wins because it plays enough hands, not because any individual hand is a good bet.

But it seems extremely misaligned for entrepreneurs and consumers. The game is stacked so that only outcomes of a very specific shape (exponential, monopolistic, winner-takes-all) are considered successes. Everything else is failure, regardless of how much value it created.

Total US venture capital investment runs around $170 billion a year. Federal R&D spending is roughly $200 billion. The scale of money is not the barrier. The question is what the money demands in return.

The early phase of a startup is never really profit-driven anyway. It's driven by ideas. By someone thinking "this should exist." The earliest users are not customers in any meaningful sense. They are collaborators, people who showed up because the idea interested them. This phase, which everyone involved will later describe as the best part, is the phase where no one is trying to make money. It operates, in miniature, like socialism: people building things, freely, because they can.

But useful to whom? The things that get built in the startup ecosystem are shaped by who can afford to spend two years without income, overwhelmingly people with access to capital, elite education, and the professional networks of a narrow class. A founder in San Francisco building a meal delivery app is solving a problem that exists primarily for people like themselves. The entire category of what counts as worth building is filtered through the class position of the builders. This is not a problem of funding. It is the class structure of the entire ecosystem reproducing itself in what gets built and for whom.

The money enters later, and with it the distortion. The investor holds equity. Equity is only liquid through an exit event. Exit events require scale. Scale requires extraction. So the investor who funded the idea now needs to see it monetised, not because they are greedy as individuals, but because the structure demands it. Features that users love but do not generate revenue are cut. Features that users hate but generate revenue are added. The subscription that was free becomes paid. The data that was private becomes a product. The phase where usefulness governed production is over. Capital has reasserted its logic.

What if that phase never had to end? Not through a different funding model. Grant models have their own problems, and the state that would distribute grants is not a neutral actor; it serves the same class interests that shape the market. But in a different society entirely. One where people's basic needs are already met, housing, healthcare, education, so no one has to bend to the will of whoever controls the capital. Engineers, scientists, artists, and organisers produce without the profit motive every day. The constraint has never been willingness. The constraint is that eventually someone has to eat, and under capitalism that means submitting to whoever holds the money.

Even then, removing the profit motive does not automatically make technology serve people. Who controls it, who benefits from it, whose labor it replaces: these questions persist regardless of how the work is funded. Technology built under different economic conditions would still need to be governed, and governed by the people it affects, not just the people who build it.

A project that covers its costs and serves ten thousand people well would not be a failure. It would be useful to the public. It could last as long as it remained so. The workers building it would not be disposable inputs on a burn rate spreadsheet. The users would not be engagement metrics. There'd be no investor demanding an exit, so no incentive to sell it off to Google or Meta. Which means a more open, more competitive landscape. More projects sticking around. Less consolidation.

This is especially true for startups that require very little capital. The ones that mainly need to pay employees and rent an office. If there's no need for 10x or 100x profit, the math changes completely.

Right now, a VC fund backs a portfolio of startups knowing that most of the returns will come from one or two massive winners. Instead of expecting that most of them are useful. Those one or two winners that have to deliver outsized returns? They have to do weird things to get there. They classify workers as contractors to avoid paying benefits. They harvest user data and sell it because that is the fastest path to revenue. They make their products addictive rather than useful because engagement metrics are what the next funding round requires. They lobby against regulation that would protect consumers because regulation slows growth. These are not failures of individual morality. They are the rational responses to an incentive structure that rewards only one shape of success and punishes everything else as failure.

What if they didn't need to do that? They could be more open source. Pay their people better. Charge less to consumers. Build things that last.

The current model chases outsized returns from a handful of startups, while the rest have to shut down much earlier than they should. A startup that is useful, sustainable, and loved by its users gets killed because it will never deliver those returns. Not because it is failing. Because it is not failing in the right direction. So it gets cut. The resources that went into building it, the years of work, the accumulated knowledge, the user relationships, are discarded. The team disperses. The codebase rots. The users find a worse alternative or go without.

Thousands of useful products destroyed every year, not by the market, but by the terms on which capital enters the market. And for what? The outsized returns do not flow to the users who made the product valuable by showing up and using it. They do not flow to the employees who built it. Their equity, typically a fraction of a percent, is diluted by multiple rounds until it amounts to almost nothing. The gains are private. The losses are socialised: dead companies, abandoned users, wasted labor absorbed by everyone except the investors who produced them.

The current model occasionally produces something useful to the public despite itself. The question is which model makes usefulness the default rather than the exception. One produces billionaires. The other would produce public goods. The investor class has decided, for reasons that have nothing to do with efficiency and everything to do with preserving their position, that the first option is the only serious one.

But maybe the deeper question is why we are talking about startups at all. The startup, a small private firm founded by one or a few individuals scaling toward an exit, is itself an artifact of capitalist social relations. It presupposes private ownership of collectively produced value. It centers the founder as protagonist when the work is done by many. In a society organised differently, the things people want to build might not take this form at all. They might be cooperatives, public services, commons. The startup itself is a shape that only exists because of the system that produces it. Change the system, and people will still build. They will just build differently.