When the team is just you, the shape of the company changes. No payroll to run. No org chart to keep. The thing you're building might live for ninety days, or three years, or roll quietly into something bigger. The entity should match — small, sharp, easy to start, easy to fold away.
For a long time, "startup" has meant a small team — two or three founders, a few early hires, a board, an option pool, a roadmap to look like a bigger company as fast as possible. Most of the tooling underneath was built for that.
But a lot of the most interesting work happening right now is being done by one person. Maor Shlomo built Base44 alone in four months — a platform that turns plain English into working apps — and sold it to Wix for $80M last June. Dana Snyder shipped Positive Equation, a software consultant for small nonprofits, with no engineering background and Replit doing most of the work behind the keyboard. Pieter Levels took fly.pieter.com — a browser flight sim — from idea to $1M ARR in seventeen days, still by himself. Marc Lou runs ShipFast, CodeFast, and DataFast in parallel, three products with zero employees between them.
None of that is a "small team." It's not just a startup with fewer chairs at the table. It's a different object. The questions are different, the costs are different, the things that matter are different, and the things you can skip entirely are different. Treating a one-person company like a five-person company is how you end up paying for software, paperwork, and obligations that have nothing to do with the work you're actually trying to do.
The reason this matters now is that the operating reach of one person has gone up sharply. A single person can run more, ship more, sell more, and finish more than even a few years ago. The bottleneck used to be capacity. Increasingly, it's the entity itself — the friction of being a real thing in the world, registered and signing contracts and paying tax. That's the part we want to make small.
A one-person company isn't a small team. It's its own shape, and most of the tooling around it was built for someone else.
The first thing to notice about a one-person startup is how much of the standard corporate machine isn't doing anything yet. A lot of the cost of a small company comes from carrying tools, processes, and obligations that exist to coordinate humans — even when there are no humans to coordinate.
| № | In a small team | In a one-person startup |
|---|---|---|
| 01 | Payroll | Doesn't exist. You're a contractor to your own company, or you're not paying yourself yet at all. No W-2s, no benefits stack, no quarterly 941s to chase. |
| 02 | HR | Nothing to manage. No handbook, no PTO policy, no I-9s on file, no harassment training, no manager layer. You'll add it the day before your first hire — not the day you incorporate. |
| 03 | Founder equity split | There is no split. One person owns the whole company. The conversation that takes weeks among cofounders takes zero seconds with yourself. |
| 04 | Vesting between founders | Not needed. Vesting exists so people who leave don't walk off with the company. If there's nobody else, there's nothing to protect against on day one. |
| 05 | Board | A board of one. Quarterly meetings become a signed consent. Most of the governance overhead drops out until somebody else has a real stake. |
| 06 | Option pool | No pool yet. You don't have anyone to grant options to and no reason to dilute yourself preemptively. |
| 07 | Stock plan + 409A | A long way off. The whole apparatus of priced rounds, valuations, and grants is something you turn on when you need it — not because the playbook said so. |
| 08 | Investor reporting | Nothing to report. You don't have investors, so the recurring updates, data rooms, and metrics decks have no audience and don't need to exist. |
| 09 | Org chart | A dot. There is no chart. There's a person and the company they own. |
| 10 | Meeting cadence | Whatever you want. No standups, no all-hands, no review rituals. The work is the meeting. |
Almost everyone selling tools to a "startup" is selling tools designed around at least one of these. Payroll software, equity platforms, HR systems, board portals, OKR trackers. They're not bad. They were built for a different object. A one-person startup that buys them ends up running the rituals of a company that doesn't exist yet — paying, in time and money, for the simulation of scale.
What's left after you take all of that out is short, and it matters. A one-person company still has to be a real thing in the world. It has to be able to sign a contract, hold money, pay tax, and prove who owns what. That part doesn't shrink.
The substrate is the same as any other company. A legal entity. A registered agent. A federal EIN. A bank account. A small handful of tax obligations — federal, state, sometimes sales tax — that recur whether you want them to or not. A clean record of ownership. A way to wind it down or exit when you're finished.
The differences are in the shape, not the surface area. The questions a one-person founder needs answered are fewer, and the answers should default further toward sensible. You shouldn't have to think about a four-year vest with a one-year cliff when there is no second person to vest against. You shouldn't have to authorize ten million shares to issue stock to yourself. You shouldn't have to model an option pool you're not going to grant from.
One of the unspoken assumptions about a startup is that it's supposed to last. The whole apparatus around it — vesting, pools, multi-year roadmaps, ten-year LP horizons — is shaped by the assumption that the thing exists for a long time. A one-person startup often doesn't, and shouldn't have to.
The shapes a one-person company can take are wider than the venture playbook allows. A ninety-day experiment to test whether something has demand. A six-month consulting vehicle for a single client. A holding entity for a piece of IP you'll license out and otherwise leave alone. A small product you intend to run for two years and then sell. A pop-up that exists only to scope, build, and hand off a single project. None of these need a ten-year company. They need an entity that matches their lifespan.
In most stacks today, closing a company is harder than starting one. The filings are messier, the timelines are longer, the surprises — final returns, franchise tax owed, registered-agent renewals you forgot to cancel — come months after you thought you were done. For a short-lived company, the wind-down is part of the design. It should be one move, on the same surface that opened it, with the same kind of completeness — final returns filed, accounts closed, agents released, history sealed. The folder doesn't grow forever.
A company that exists for ninety days shouldn't carry the running cost of a company built to exist for ten years. Most of what makes a small company expensive today is the recurring stack — agents, filings, software, accounting — priced for indefinite life. A one-person entity with a known end date should look more like a project budget than a payroll.
The other interesting thing about one-person startups is that a single person can credibly run several of them at once. A consulting vehicle. A small product. A side project that hasn't decided what it wants to be. An entity that exists for a single client. Each one is its own thing in the world — its own contracts, its own taxes, its own bank account — even when they all share an owner.
The right shape for this is something most people associate with much larger companies: a parent and a set of children. A small holding entity at the top, owned by the founder. Underneath it, a handful of operating entities — one per venture — that can be opened, paused, sold, or wound down independently. Each is light. The parent keeps the books.
| № | Inside the portfolio | Why it lives on its own |
|---|---|---|
| 01 | Holding entity | The parent. Owns the children, collects what they distribute, files one consolidated return where possible. Outlives any single project. |
| 02 | Operating co — main product | The thing you've been running for two years. Takes payment from customers, pays the agent, holds the IP. |
| 03 | Project entity — ninety days | Stood up to take a single client engagement. Lives for the length of the work, then dissolves cleanly. |
| 04 | Experiment entity | An idea you're testing. Probably folds in six months. If it works, you roll it up into the parent. If it doesn't, you put it away. |
| 05 | IP-holding entity | Owns a piece of work you license out. Doesn't do anything else. Costs almost nothing to keep alive. |
Today, this is mostly impractical. Forming the second entity is as expensive as forming the first. Rolling one up into another involves lawyers, intercompany agreements, and a quarter of someone's time. It shouldn't. The cost of an additional entity, in time and money, should be small enough that you can spin one up for a single experiment without thinking about it, and roll a successful one into the parent in a single move.
The shape that fits is a small parent and a quiet set of children — each one cheap to open, cheap to keep, cheap to fold away.
We're starting from a simple version of this. A single-member LLC, formed in a few minutes, with a single class of units, no vesting, no pool, a registered agent attached, an EIN issued, a bank account opened, a small compliance calendar set up. A company that fits one person.
From there, the work is making the rest of the lifecycle just as small. Spinning up a second entity should be one sentence. Folding it into a parent should be one sentence. Closing it down should be one sentence. The cost of optionality — of trying something, of running several quiet ventures at once, of being honest that some things will only live for a season — should be close to zero.
That's what we mean when we say we want to make the entity small. Small enough to match the work you're actually doing, and no larger.