Employer of record vs foreign subsidiary: how to decide
You do not need a foreign subsidiary to make your first international hire, and you probably should not build one until you have at least four to six people you plan to keep in that country for three or more years. Below that line, an employer of record almost always wins on cost and speed. Above it, the math flips.
Most founders treat this as a legal question and ask a lawyer. It is actually a headcount forecasting question, and you can answer it yourself with three inputs: how many people, how long, and how much local control you need.
What an EOR and a subsidiary actually do differently
An employer of record is a third party that becomes the legal employer of your international hire on paper, while you keep managing their work day to day. It handles local payroll, tax withholding, statutory benefits, and termination compliance for a monthly per-employee fee, usually $500 to $800.
A foreign subsidiary is your own legal entity in that country. You are the direct employer. You set up your own payroll, register with local tax authorities, and carry the compliance burden yourself, usually with a local accountant or law firm on retainer.
The functional output is similar: a legally compliant employee. The structural difference is who is on the hook if something goes wrong, and how much it costs to maintain that structure over time.
The real decision variable is not cost. It is headcount times time horizon
Founders default to comparing monthly EOR fees against subsidiary setup costs, but that comparison only matters once you fix the other two variables.
Subsidiary setup typically runs $20,000 to $60,000 depending on jurisdiction, plus $1,500 to $5,000 a month in ongoing accounting and compliance. That fixed cost does not change whether you have 2 employees or 20 in that country. EOR fees scale linearly with headcount instead.
Run the crossover math with your own numbers before you commit to either path:
Employees in-country: 1-3 | 3-year horizon: uncertain or testing the market | Better option: EOR
Employees in-country: 4-8 | 3-year horizon: confirmed, growing team | Better option: crossover zone, run your own numbers
Employees in-country: 9+ | 3-year horizon: confirmed hub or office | Better option: subsidiary
If you are hiring one engineer in Germany to see if a remote-first European team works at all, a subsidiary is the wrong tool. If you are opening a 15-person Berlin engineering hub over 18 months, the subsidiary pays for itself inside year one.
The signal most founders miss: who is actually asking for the entity
Cost and headcount are the obvious inputs. The variable that catches founders off guard is procurement. If you are selling into European banks, healthcare systems, or public sector accounts, RFP requirements sometimes explicitly require a local legal entity, not a foreign EOR arrangement. An EOR structure can read as "not really operating here" to a procurement team, even when your product is fully compliant.
Check this before you hire, not after you lose a deal to a legal technicality. Pull the vendor requirements section of your two or three largest target accounts in that market and search for "legal entity" or "registered office."
The hybrid path most fast-growing startups actually take
Few companies pick one option and stay there. The common pattern: use an EOR for your first three to five hires in a new country to de-risk the market bet, start incorporation in parallel once you have signal the team is staying, then transfer those employees onto your own payroll once the entity is operational.
This costs a few months of running both structures at once, but it avoids the two worst outcomes: paying subsidiary overhead for a market you abandon in six months, or discovering at 12 employees that your EOR fees now cost more than an entity would have.
What to do first this week
Before you talk to an EOR vendor or a law firm, write down two numbers: how many people you expect to have in this country in 24 months, and whether any of your top five target accounts in that market require a local entity in their procurement process. Those two answers decide 80% of this question before you spend a dollar on either path.
Frequently asked questions
How fast can I hire through an EOR versus a subsidiary? An EOR can typically onboard someone in five to ten business days once the offer is signed. Entity setup ranges from a few weeks in fast jurisdictions like the UK or Singapore to six months or longer in slower ones like Brazil or India.
Can I switch from an EOR to my own subsidiary later? Yes. This is the hybrid path most companies actually use. You keep the employee's role and compensation the same and transfer the legal employment relationship once your entity is registered and operational.
Does using an EOR limit what benefits I can offer employees? Not usually. Reputable EOR providers administer local statutory benefits and can often layer in supplemental benefits you choose, though the menu is narrower than what you could design directly through your own entity.
Is an EOR more expensive than a subsidiary long term? Only past the headcount crossover point for your specific country and cost structure, generally four to eight employees on a three-year horizon. Below that, EOR fees are almost always cheaper once you account for setup and ongoing entity maintenance costs.
Do I need a subsidiary to close enterprise deals in another country? Sometimes. Regulated buyers like banks and healthcare systems occasionally require a local legal entity in their vendor requirements. Check the specific procurement language for your target accounts before assuming an EOR will be sufficient.