Starting a company is one of the most exciting things a person can do. Closing one is one of the most painful. Months of paperwork on top of whatever grief or relief came with the decision in the first place. That asymmetry is the reason the world is full of entities nobody is running, and the reason founders avoid the structures they actually need. We're building the infrastructure to make stopping feel as light as starting — so a company can be a tool you use, not a thing you carry forever.
DE
Dominic Esposito
Founder · matter
The hard part is stopping.
Starting a company is exciting. It is the moment a founder picks a name and commits to it, opens a bank account, signs the first lease, hires the first person, ships the first thing. The on-ramp has been rebuilt so many times by so many people that it is now a genuinely good experience — Stripe Atlas, Mercury, Doola, Firstbase and a dozen others have flattened it into a checkout flow that feels, for a few minutes, like nothing is in the way of the future you are about to build. The off-ramp has not been touched.
Closing a single-state LLC — cleanly, in good standing, the way a lawyer would do it — is a sequence of roughly fifteen to twenty discrete steps spread over two to six months. Members have to adopt a written plan of dissolution. Final state and federal tax returns have to be filed, with the "final return" box checked, in the right order. In states like California, New York, New Jersey and Massachusetts you cannot file articles of dissolution at all until the tax authority issues a tax clearance certificate, which can take six to twelve weeks on its own. The IRS gets a separate letter requesting EIN closure. The bank account has to be closed, but only after the final tax payments clear. Business licenses, sales-tax permits, foreign qualifications in other states, DBA registrations — each one has its own cancellation form, fee, and waiting period. Creditors have to be noticed. A claim window has to run, often 120 days or more, before remaining assets can be distributed without personal liability to the members. Then the registered agent is released. Then the records have to be retained for seven years, because the state can still come back.
None of that is on a screen anywhere. None of it has an API. None of it is one button. It is paper, PDFs, certified mail, hold music, and a quote from a paralegal that starts at $1,200 and ends wherever the complications end. And it lands on the founder at the exact moment they have the least energy for it — the project is over, the partnership ended, the experiment didn't work, the brand has run its course. Whatever the reason, closing a company is something a person has to do while they are already grieving the company. So most founders do the rational thing: they stop using the entity and hope it goes away.
It does not go away. In California alone, the franchise tax board lists roughly 800,000 entities in suspended or forfeited status — companies that were never properly closed, still accruing the $800 annual minimum tax, still on the hook for penalties and interest, still legally exposed in the names of their members. Across the country the count is in the millions. Every one of them is a person who started something with real excitement, didn't quite need it, and is now paying rent on a structure they cannot easily turn off.
This is what we mean when we say the asymmetry distorts the upstream. If starting is exciting and stopping is painful, people start fewer of the things they should. Founders avoid spinning up a subsidiary to test an idea because they don't want to inherit the wind-down if it doesn't pan out. They avoid the cleaner holding-company structure because the cost is denominated in future grief nobody has budgeted for. They avoid the right tax election because changing it later will mean closing and re-forming. The cost of an entity is not the formation fee — it is the eventual dissolution, discounted to today, and tax-loaded with whatever emotional weight comes with the ending.
As easy to stop as to start.
The reason dissolution is hard is that it touches everything an entity touches. To close a company you need to know what it owes, what it owns, who its registered agent is in each state, which licenses it holds, what filings it has open, what its bank balance is, whether payroll is wound down, whether the final return is filed, whether there are outstanding contracts, whether anyone has served it papers in the last sixty days. No single party knows all of that today. The accountant has one slice. The lawyer has another. The registered agent has another. The bank has a fourth. The founder, increasingly, has none of them.
Matter is building the substrate where all of those slices live in one place. We are the system of record for the entity itself — its members, its capitalization, its operating documents, its tax elections, its filings, its compliance calendar. We are the registered agent in every state it operates in, so the mail comes to us and the state writes to us. We are the counterparty on its bank account and the issuer of its EIN-bound infrastructure. When matter knows everything an entity has open, dissolution stops being a project and starts being a function call.
The same intent that made forming a company feel like momentum is what makes closing one feel like momentum, too. The work didn't disappear; it moved into software that already had all the inputs in front of it. We are not asking the founder to assemble fifteen pieces of paper from six different counterparties at the exact moment they have the least patience for it. We are doing the assembly because we are six of the counterparties.
The dissolvable entity.
Once stopping is cheap, the shape of how people use companies changes. The entity stops being a monument and starts being a tool. You reach for one when you need one, and you put it back when you're done.
Agents forming their own companies, for micro reasons.
We think a large fraction of the entities formed five years from now will not be formed by people. A procurement agent buying GPU capacity on behalf of a research team will sometimes need to be a legal counterparty in its own right — to sign an NDA, to take on a credit line, to indemnify against a vendor's terms. An ad-buying agent running a single campaign for a single client may want a thin LLC to hold that campaign's contracts and ad-account spend, so the books are clean when the campaign ends. A marketplace agent posting on behalf of a long-tail seller may need a transient entity to sign a single bill of sale.
None of these are companies in the cultural sense. They have no employees, no roadmap, no aspirations. They exist for an afternoon, a week, a quarter — then they're done. They only make sense if the cost of forming and dissolving rounds to zero. An agent that can stand up its own counterparty for one transaction, sign one paper, and fold the entity an hour later is doing something that wasn't possible before. It is treating a legal entity the way a backend treats a database row.
Subsidiaries that spin up to test an idea, and fold when the answer is no.
Inside a single founder or a single holding co, the same shape unlocks something more familiar. Today, if you have an idea — a new brand, a new product line, a new market — you face a binary. You can launch it under your existing entity and inherit the brand confusion, contract entanglement, and tax cross-contamination. Or you can stand up a real subsidiary and pay the formation tax twice: once on the way in, once on the way out, plus the months of dissolution work in between if it doesn't work.
Most ideas are tested under the parent because the second option is too expensive. The cost is not the formation fee; it is the option on closing. Strip out that option, and the calculus changes. You can open six subsidiaries to test six product concepts in parallel, give each one its own books, its own bank account, its own contracts, and its own brand exposure. Five of them you close in ninety days when the data isn't there. One of them you roll up into the parent. Or you let the one that's working keep running on its own, and you re-form the holding structure around it.
This is how venture studios, accelerators, and indie holding cos already want to operate. It is how Tiny, IAC, and Berkshire-style holding cos operate at a much larger scale and a much higher cost per move. Bringing that operating model down to a single founder with a laptop is a different category of leverage.
Holding structures that breathe.
The end state is a holding structure that breathes — entities forming, merging, and dissolving on the timescale of the experiments running inside them, not on the timescale of legal calendars. A founder running a portfolio of five products is, structurally, running a portfolio of five companies; the legal substrate should match the operating reality, not lag two years behind it. A creator with three distinct revenue streams should be able to put each one in its own entity for tax and liability reasons without taking on the operational burden of running three companies.
The disposability is what makes the structure honest. Today, the structures founders end up with are the ones they happened to set up in year one, frozen in place by the cost of changing them. When closing is cheap, the structure can track the truth.
What this means in practice.
None of this changes what a company is. The state's statute is the same. The IRS's forms are the same. The duties members owe each other and the duties an entity owes the public are the same. We are not arguing the law should be different. We are arguing that the work of meeting the law should not be a barrier to using it.
If forming a company is the moment a founder commits to an idea — the exciting one, the one we celebrate — dissolution is the moment they admit something is finished. A contract is over, an experiment didn't work, a brand has run its course, a holding structure has outgrown a child. The endings deserve the same care as the beginnings. They deserve to be one motion. One sentence in. A sealed entity out. The papers signed, the taxes settled, the agent released, the history archived, the costs ended — so the founder is free to start the next thing.
Companies should be as easy to put down as they are to pick up. That is what we are building.