Decision tools

I built these tools because the same questions kept coming back. Every founder conversation about the US raised them. Answering them one at a time on calls didn't scale. Work through them honestly; the useful outcome is often "not yet", and that's a good answer, not a failure. Going too early is how deeptech companies run out of money.

Free simulator

Funding in France first?

Before the US question, check what the French public side could give you. Five questions, a first map of grants, tax credits and loans.

Run the simulator →

Where this sits: the Playbook is for understanding the terrain and the method. These tools are for testing your own readiness against it. And if you want to work through your actual case, that is what I do with founders.

Tree 1

Should I go to the US at all?

Question 1

Is there a real, sizeable US market for what you build, or are you assuming one because the US is big?

Question 2

Can you name three American companies that would buy this, and say why they'd switch from what they use today?

Question 3

Is your differentiation real in the US context, where the competitive set is different from Europe?

Question 4

Do you have at least 12 months of runway that doesn't depend on the US working?

Tree 2

Do I need a US entity?

Question 1

What are you actually trying to do right now?

Question 2

Has a customer or investor explicitly told you a US entity is required?

Question 3

Where does your IP sit, and would moving it jeopardize CIR, JEI or existing grants?

Tree 3

Where should I land?

Question 1

Who is your buyer?

Question 2

What matters more in the next 18 months?

Budget frame

Four scenarios, from testing the market without leaving France to a full local presence. The line items are the ones that actually appear; the numbers are yours to fill in. Nothing here is pre-filled, because no two deeptech companies have the same cost structure.

1 · Validate from France

2 · Commercial launch

3 · Entity + first hire

4 · Full presence

No entity, no relocation. You test whether anyone in the US actually wants this. Twelve months.

Line	Low	Base	High
Travel, 3 to 4 trips
Conferences and events
Market research and data
Localized collateral and site
Advisory or fractional support
Total	—	—	—

Still no entity. You're selling: discovery calls, pilots, a first paid contract invoiced from France.

Line	Low	Base	High
Travel, 6 to 8 trips
Conferences, booth or speaking
Sales tooling and data
Pilot delivery cost
Contracts reviewed by US counsel
Insurance and compliance
Total	—	—	—

A US entity exists and someone is on the ground. First full year including setup.

Line	Low	Base	High
Incorporation and legal setup
Registered agent and state fees
US accounting and tax filings
First hire, fully loaded
Immigration, legal and filing fees
Payroll, benefits, insurance
Workspace
Travel and founder time
Total	—	—	—

A real local team. Add the transatlantic structure cost most founders forget to price.

Line	Low	Base	High
Team, 3 to 5 people loaded
Office and operations
Legal, corporate and IP
Transfer pricing and cross-border tax
Marketing and demand generation
Immigration, ongoing
Insurance, D&O, compliance
Total	—	—	—

Fill in your own figures. The calculation updates as you type.

Visa routes

The main routes French founders actually use, what each one allows, and what it costs to file.

Before anything else

Visa, status and work authorization are three different things. A valid visa lets you request admission. Status is the set of conditions you were admitted under. Work authorization is separate again. People conflate them constantly, and that's where most avoidable problems start.

Route	Who it fits	Core condition	Can you work?	Government fees	Green card path
ESTA	Short business trips, conferences, meetings	French passport, no visa needed, stays up to 90 days	No. Meetings and negotiation only, no productive work	Low, application fee only	No
B-1	Business visitors needing more than ESTA allows	Consular interview, demonstrated intent to return	No. Same restriction as ESTA	Consular application fee	No
O-1	Founders and specialists with a documented record	Extraordinary ability, evidenced. No lottery, no annual cap	Yes, for the petitioning employer or agent	I-129 base filing plus category fees. Premium processing $2,805 optional	Yes, commonly via EB-1A
E-2	Founders investing in a US business	Treaty investor. France qualifies. Substantial investment, at-risk capital	Yes, in the invested business	Consular fees, plus legal and structuring cost	No direct path
L-1	Transferring from your French company to a US entity	One year of qualifying employment abroad in the prior three. Requires a related US entity	Yes, for the petitioning employer	Government fees generally $1,195 to $2,485 for an initial individual petition, depending on employer profile	Yes, L-1A commonly via EB-1C
H-1B	Employees joining a US company in a specialty occupation	Annual lottery. Sponsoring employer required. Registration window in March	Yes, for the petitioning employer	$215 registration, plus I-129 and statutory surcharges. Roughly $2,880 to $6,880 in USCIS fees for a mid-size employer before premium processing	Yes, employer-sponsored

Fee figures last verified September 2026 against USCIS published guidance and current practitioner sources. Fees, quotas and rules change; verify before filing.

Start with the trigger, not the visa

The right question isn't "which visa can I get" but "what am I actually trying to do". Selling remotely, attending a conference, hiring a US employee and relocating yourself are four different problems with four different answers.

O-1 rewards preparation

The evidentiary bar is high: publications, press, awards, judging, measurable impact. Building that case takes eighteen months or more. Founders who look at O-1 only when they need it have already lost the window.

L-1 needs a real French company first

You need a year of qualifying employment abroad and a genuine corporate relationship between the two entities. It can't be assembled retroactively.

H-1B is a lottery, so it's not a plan

Registration happens in March for an October start. Selection isn't guaranteed. Building a US launch timeline that depends on winning it's a planning error.

This page provides general information based on public sources and personal experience. It's not legal, immigration or tax advice, and it doesn't create an attorney-client relationship. Immigration outcomes depend on facts specific to each case. Consult a qualified US immigration attorney before acting.

Government fees and filing rules on this page were last checked in September 2026. If a figure looks wrong, it probably moved: tell me and I'll fix it.

Three paths

Three ways the crossing tends to unfold, depending on what you build and who buys it.

These are composite scenarios. They're built from patterns observed across the 57 interviews behind the research and from advisory work, not from a single identifiable company. No figures are attributed to any specific firm, and none should be read as a promise of outcome.

Scenario A

Early US interest

Starting point: Two technical founders out of a French lab. A working prototype, one French pilot customer, funded by JEI status and a Bpifrance innovation grant. No US presence, no US revenue.
The trigger: An inbound inquiry from an American company, which the founders initially read as validation. It was interest, not demand. The useful move was treating it as one data point and running proper discovery rather than booking flights.
Discovery: Roughly forty conversations with US prospects over two quarters, run from France. The product-market fit that worked in Europe didn't survive contact: the American buyer had a different job to be done and a different competitive set.
First paid pilot: A small paid engagement, invoiced from the French entity. No US company needed at this stage. The paid pilot is the evidence that changes every subsequent conversation, with customers and investors alike.
The entity question: Only when a US seed fund expressed lead interest did the structure question become real. That's the correct order: the entity follows the deal, not the ambition.
The flip: A Delaware holding above the French operating company, structured with franco-American counsel to preserve CIR and JEI eligibility. Expensive, slow, and much harder if it had been improvised earlier.
What mattered: Selling before structuring. The non-dilutive French stack funded the discovery period, which meant the US raise happened from a position of evidence rather than need.

Scenario B

Industrial expansion

Starting point: A materials or process technology with a pilot line in France. Several years of R&D behind it, funded by a mix of CIR, regional aid and a France 2030 instrument. Long sales cycles, industrial buyers.
Why the US at all: The addressable industrial base is larger and the procurement cycles, once started, move faster. But the supply chain and certification constraints are heavier than in software, and they don't travel.
The constraint most founders miss: Industrial customers buy from suppliers they can audit. Geographic distance is a procurement risk, not just an inconvenience. Certification requirements can add twelve to eighteen months that no amount of capital compresses.
Sequencing: Keep R&D and the pilot line in France, where the grant stack and the engineering talent already are. Sell into the US from there. Establish a subsidiary once a customer commits to volume, not before.
Manufacturing: Contract manufacturing in the US ahead of any owned facility. A joint venture, if it happens, comes later and usually at the customer's initiative rather than the founder's.
What mattered: Resisting the pressure to relocate the industrial base early. The French public stack is genuinely hard to replicate. Moving it to chase a market you haven't yet won is how industrial deeptech runs out of money.

Scenario C

Regulated market entry

Starting point: A device or diagnostic with French clinical validation underway, CE marking in progress, hospital partnerships in France. Capital-intensive, regulator-gated, long.
The regulatory reality: European approval doesn't transfer. The FDA pathway is separate, and which pathway applies depends on classification and predicate. This is determined early or it derails the timeline later.
The part founders underestimate: Regulatory clearance isn't commercial access. Reimbursement is a distinct and often longer process. A cleared device with no reimbursement pathway has a small addressable market regardless of clinical merit.
Hospital adoption: American hospital procurement runs through committees, existing group purchasing agreements and clinical champions. A single enthusiastic clinician is necessary and nowhere near sufficient.
Financing: The timeline usually outruns what French public funding covers alone. US health-specialist investors understand the regulatory clock better than generalists, which is often the real reason to raise there.
What mattered: Mapping the regulatory and reimbursement path before committing capital to US market entry. Where regulation shapes the market, sequencing errors are measured in years, not quarters.

If one of these resembles your situation, the details are where it gets useful. Book a meeting

Decision tools

Start with the question you actually have

Should I go to the US at all?

Do I need a US entity?

Where should I land?

Budget frame

Visa routes

Start with the trigger, not the visa

O-1 rewards preparation

L-1 needs a real French company first

H-1B is a lottery, so it's not a plan

Three paths

Early US interest

Industrial expansion

Regulated market entry