Entity type selection examples: five real company patterns and what each choice turns on

Entity type selection examples repeat across real companies. Five patterns show what each choice turns on. Dell started as a sole proprietorship. Microsoft began as a partnership. Goldman Sachs converted from a partnership to a C corporation in 1999.

Patagonia restructured its ownership in 2022. Franchisees of Two Men and a Truck and CertaPro Painters commonly form LLCs or S corporations. Each pattern asks the same three questions: how many owners, what is the realistic loss mechanism, and how does money leave the business.

Your own facts belong in the same frame, then in front of an accountant and a lawyer.

What to take away

Situation A: sole proprietorship first (Dell, Spanx)

Michael Dell founded PC's Limited, later Dell Technologies, as a sole proprietorship in 1984 from his University of Texas dorm room. He incorporated soon after. Sara Blakely started Spanx in 1998 with $5,000 and operated as a sole proprietorship before incorporating. Both began without an entity. The pattern: start simple, incorporate when scale and risk demand it.

What it turns on: whether there is exposure worth building a structure around, and whether the maintenance would actually get done.

Toward staying a sole proprietor: exposure is genuinely low; income is simple; an entity dissolved in eight months cost money and paperwork to prove a point. Being a sole proprietor is the default, not a failure to decide, sole proprietorships sets out when it remains reasonable.

Toward forming an entity anyway: clients who will not contract with an individual; work that could cause a client a financial loss; a wish to keep a home address off public records, though note that the entity's registered agent address is itself public, so that has to be solved separately.

What an advisor would want to know: what a bad outcome in this work actually looks like, what professional coverage costs, and whether any client has asked for an entity or a certificate of insurance.

Situation B: partnership without paper (Microsoft, Facebook)

Microsoft was founded in 1975 by Bill Gates and Paul Allen as a partnership. Facebook began in 2004 with an oral agreement between Mark Zuckerberg and Eduardo Saverin. The Facebook disputes over ownership became lawsuits.

In most places, carrying on a business together for profit already makes them partners by default, with each able to bind the other and each personally exposed to obligations the other took on.

What it turns on: not the entity at all. The default rules are already running.

Toward doing the agreement first: the urgent gap is the governing document, not the form. Contribution, decision rights, profit split, exit, valuation method, and deadlock all need answers while both people are still reasonable. Partnership formation lists the questions.

Toward a form with liability separation: the exposure they have already accumulated jointly, and the fact that a structure formed now does not retroactively clean up what came before.

What an advisor would want to know: what has already been agreed in writing anywhere, emails and invoices count, what either of them has signed personally, and whether the split of profit is meant to match the split of ownership.

Situation C: physical work and liability (Two Men and a Truck, CertaPro Painters, Servpro)

Franchisees of Two Men and a Truck, CertaPro Painters, and Servpro often form LLCs or S corporations to separate personal assets from business liabilities. A moving crew can damage property or injure someone. A painting crew can cause a fire. A restoration crew can face mold claims. The loss size is not bounded by the job size.

What it turns on: the loss mechanism. This is work where a person can be hurt and property can be damaged, and where the size of a claim is not bounded by the size of the job.

Toward liability separation mattering a great deal: a plausible claim here could exceed what an individual could absorb. That is the threshold at which structure stops being administrative.

Toward insurance being the larger question: an entity decides who a claim lands on; it does not pay anything. General liability, vehicle, and, once there is help, employment-related coverage do work no entity can do. A business that formed an entity and economized on coverage has usually made itself worse off.

Toward getting the employment question right early: taking on help introduces registrations, withholding, reporting, and insurance obligations, and some of those are conditions of operating rather than optional extras.

What an advisor would want to know: the largest single job expected, what the insurance market quotes for this trade, whether the help will be employees or contractors, and that last question has a real answer determined by the facts, not by what the paperwork calls it.

Situation D: outside investment (Goldman Sachs, 1999)

Goldman Sachs was a private partnership for 130 years. In 1999, it converted to a C corporation and sold shares to the public. The partners wanted capital and a stock currency for employees. Investors and their lawyers work from documents built around stock, share classes, and a board.

What it turns on: what the counterparties expect, which here matters as much as what suits the founders.

Toward the corporate form: investors and their lawyers work from documents built around stock, share classes, and a board. Employee equity plans are well-trodden ground in a corporation and more bespoke elsewhere, harder to explain to a new hire and more expensive to paper. Corporation formation covers what that structure commits you to.

Toward pausing: if the raise is speculative rather than imminent, the corporate machinery is real overhead to carry for a plan that may not happen. Ask what conversion later would cost, and get the answer before deciding, not after.

What an advisor would want to know: whether any investor has expressed a preference, the intended timing of a raise, how many people will hold equity, and where those people will be.

Situation E: crossing two thresholds (Patagonia, New Belgium Brewing)

Patagonia restructured its ownership in 2022. Founder Yvon Chouinard transferred the company to a trust and a nonprofit to direct profits to environmental causes. New Belgium Brewing became employee-owned through an ESOP in 2013. Both faced succession and scale changes.

What it turns on: two thresholds crossed at once. Both employment and contract size change the exposure picture, and neither triggers a review by itself.

Toward restructuring now: the new contract may be large enough that failure could not be absorbed personally; an employee's conduct is something an employer can answer for.

Toward care in the transition: forming an entity does not move an existing business into it. Contracts, accounts, licenses, registrations, and insurance all have to be transferred deliberately, and anything left behind remains personal. Post-formation changes covers the handover.

Toward professional input specifically: the counterparty on the large contract may need to consent to a change of contracting party, and anything already personally guaranteed does not move.

What an advisor would want to know: the timing of the contract against the timing of the hire, what has already been signed personally, and whether the new contract is with a counterparty who will want to see the entity's formation documents.

What the five have in common

Read together, the same three facts do most of the work every time: how many owners there are, what the realistic loss mechanism is, and how money is going to come out of the business. Everything else adjusts the answer at the margin.

Two further observations worth carrying away. The situation that most urgently needed something was the Microsoft and Facebook pattern, where a written agreement mattered more than an entity. In the franchise examples, insurance was the larger question.

The order to work through your own facts is in the entity type selection guide; the facts to assemble first are in the checklist.

The five side by side

Read down the third column. In four of the five, the first thing worth pricing is not a formation.

The general federal orientation for any of these is the IRS's material on starting a business, and the SBA's summary of choosing a business structure is a fair map of the same ground. Both will send you to your own state for anything specific, which is the correct answer.

Common questions

Why do none of the five end with a recommendation for your business?

Because a recommendation would need facts that are not in the example, and inventing them would make the illustration worse rather than more useful. Each one shows what the argument turns on, not what you should do.

Are these real companies and public facts?

Yes. Dell, Microsoft, Goldman Sachs, Patagonia, New Belgium Brewing, Two Men and a Truck, CertaPro Painters, and Servpro are real. The facts here are publicly known. No company, person, price or outcome is invented.

Which situation is most people?

The sole proprietorship first pattern, like Dell and Blakely, by a wide margin. Most people working for themselves have exposures that a contract term and a policy address more cheaply than a formation would.

Situation B says the agreement matters more than the form. Is that generally true?

Where there is more than one owner and nothing written down, yes. The default rules are deciding the exit, the profit split and who can bind whom right now, and no choice of entity changes that on its own.

What should I do with these?

Write your own situation in the same shape: what it turns on, what pushes each way, and what an adviser would need to know. That document is what makes the paid conversation short.