When should you actually incorporate?

Two failure modes, and first-time founders manage to hit both.

The first: incorporating on day one because it feels like what real founders do. You now have an entity, an annual filing obligation, a registered-agent bill, possibly a franchise tax, and — if you picked the structure everyone recommends without checking whether it fits you — a cross-border tax question you'll pay an accountant to explain.

The second: waiting so long that a contractor has been shipping code for six months, with no entity to assign the IP to and no agreement saying it should be.

The right answer is a trigger, not a date.

The four triggers

Money is about to move. Someone is paying you, or you're paying someone meaningful. Invoicing as a person works until it doesn't — for taxes, for a customer's procurement process, for anything that needs a VAT number or a W-9.

You're signing something with real liability. A customer contract with an SLA. A lease. Anything where "you personally" and "the company" being the same thing matters.

There's a second person. A cofounder, or an early employee with equity. Equity requires something to hold equity in. This one is also the deadline for the vesting conversation — see Cofounder vesting, explained.

An investor asked. Nobody wires money to an individual. If a raise is genuinely near, the structure question becomes urgent and also constrained — investors have strong opinions about what they'll invest in.

If none of these is true, you're probably fine, and the money is better spent on almost anything else.

Why "just do a Delaware C-corp" is bad advice

It is the right answer for a specific founder: one raising US venture capital, or planning to, with a US presence.

For a solo founder in Europe with European customers and no intention of raising, it is a structure that adds annual US filings, franchise tax, an agent, and a permanent question about where the company is actually tax-resident. The advice is repeated so confidently because the people repeating it are in the situation it fits — American founders raising American money. Silicon Valley advice is not wrong; it's specific, and it travels badly.

The questions that actually determine the answer:

• Where do you live and pay tax?
• Where are your customers?
• Are you raising, honestly, in the next 18 months — or is that a someday?
• Is anyone else getting equity?
• What does the product do, and does it touch a regulated area?

Notice that none of these is "what do successful startups do."

What incorporating does and doesn't get you

Does: separates your personal assets from the company's liabilities. Gives you something to sign contracts as, issue equity in, and receive investment into. Makes you legible to procurement departments.

Doesn't: protect an idea. Protect you from a contract you personally guaranteed. Protect you from your own negligence. Assign IP that was written before the entity existed — that needs a separate assignment, which is exactly the thing people forget.

That last one is worth repeating. An entity does not retroactively own the code your contractor wrote in March. Someone has to sign a document.

The order that avoids both failure modes

1. Get the IP assignments right now, entity or not. Every contractor, every collaborator, every friend who "helped with the frontend." A short written assignment is cheap; tracking someone down in three years is not always possible.
2. Wait for a trigger.
3. When a trigger fires, choose the structure from your situation — where you live, where your customers are, whether you're raising — rather than from folklore.
4. Get a professional to check it if there's a cross-border element. Not to decide it; to check it. Walking in knowing what you're trying to achieve makes that a cheap conversation rather than an open-ended one.

Working it through

Incorporate Now — and As What? is the Groundwork project type for exactly this. It's a Decision project: no document to build, you leave with the decision made and a shareable summary of the reasoning — including what you ruled out and what would make you revisit it, which is the part you'll want when someone asks in six months.

Related: Founder Agreement if there's a second person, and Contractor / Freelancer Agreement for the IP problem above.

Groundwork is a structured thinking tool, not a law firm, and nothing here is legal advice. It helps you arrive at a well-reasoned position and know which questions to bring to a professional — which is usually what makes that conversation short.