TL;DR
- The trap: Per-employee EOR fees look expensive next to incorporation. That comparison prices only the entry, not the exit.
- When doing nothing is right: Headcount is small, the country is exploratory, or the next 18 months are uncertain.
- What has to be true to switch: You can answer yes to all five diagnostic questions in this piece without flinching.
- How the options split: Five structures exist. Only one of them is your own local entity, and only some readers should be using it.
- Decision rule: You're not buying an entity. You're buying an obligation that's harder to leave than to enter.
- Outcome to expect: Either a clean switch that protects employment continuity, or a permanent cost that grows quietly for years.
The Quiet Conversation With Finance
A chief operating officer opens a shared spreadsheet at half past nine on a Tuesday. The line items are familiar by now. One row for each country. A monthly figure for every international hire, charged by an employer of record per employee. Another row, much smaller, marked "entity cost consideration," that someone on the finance team has been updating each quarter. The math on the screen is simple. The EOR is bleeding the budget, and the local entity wouldn't. The COO knows this. The finance director knows this. The board asked about it last week.
So the meeting is on the calendar. The question on the agenda is written in the language of cost, because that's the language the board speaks. The COO has a different question in her head, and it doesn't fit on a slide. She has been told the comparison is fees against incorporation. But the comparison she has lived through, three times at previous companies, is fees against the permanent thing an entity becomes once it exists. The local entity doesn't show up as a one-time cost. It shows up as a payroll provider, a local accountant, a registered office, a director whose name she can't pronounce and whose signature she depends on, a statutory audit that happens whether the year went well or badly, and a wind-down bill the day someone proposes leaving the market. None of that's in the spreadsheet. The real issue isn't which option is cheaper per head. The real issue is whether she is willing to take on something that's much harder to exit than to enter, in a country where she isn't certain the company will still be operating in three years.
When You Genuinely Do Not Need to Act Yet
There's a version of this decision where the right answer is to do nothing for another year, and it's worth naming first, because the rest of this article is wasted on you if that's your situation.
Best tools for Global HR
The genuinely fine setup. You have between one and four people in a country. You hired them this year or last. You don't know if the headcount will be ten or zero two years from now. The role itself is sales, business development, or a regional commercial lead whose job is to find out whether the market is real. The EOR is doing its job, which is to let you learn about a country without committing to it. The fees feel large because the headcount is small. That ratio inverts as headcount grows, but the inversion isn't yet large enough to overcome the optionality the EOR gives you. Stay where you're. Spend the next two quarters learning whether the country is going to be a market or a lesson.
The setup with friction. You're between five and fifteen people in a country, the EOR is performing fine, and the work is steady. The friction isn't strategic. It's operational. Benefits administration has become awkward, onboarding for new joiners is slow, and your finance partner has started asking why the per-employee fee is what it's. None of this is a reason to incorporate. Each of them is a reason to push the EOR for better terms, or to bring in a benefits broker who can run a parallel local policy. The EOR isn't the problem here. The EOR is the surface the problem is showing up on.
The setup with real risk. You're over twenty people in a country and the role mix has shifted. You have engineers, customer success managers, and at least one person whose contract is being negotiated by a lawyer on the other side. The EOR is still functioning, but you've started to feel the ceiling of what an EOR can do. Senior hires are asking equity questions the EOR can't answer cleanly. The sales lead is closing a deal that requires a locally incorporated counterparty. Your tax advisor has mentioned the words "permanent establishment" in a meeting and you wrote them down on a napkin. The friction is no longer operational. It has become a constraint on the business, and ignoring it has a cost you can feel.
The edge case worth naming. You're a small team of two or three, but one of them is a country manager who reports into a regional head, has signing authority for a local supplier, and was hired to build a market that has not yet produced revenue. The structure works, technically. The risk lives in the gap between what the country manager is authorised to do and what the structure is set up to handle. That gap is where a permanent establishment can be born. This isn't yet a reason to incorporate. It's a reason to have a conversation with a tax adviser about what the country manager is actually doing, and whether the activities need a different container.
Five Questions This Reader Asks Themselves at 11pm
How long will I actually be in this country? The honest answer for most readers is "I don't know yet," and that's the wrong answer for an entity decision. An entity is a multi-year commitment with a non-trivial wind-down. If you can't give the country three more years with conviction, the math isn't the math you think it's. If you can, the question changes shape.
Is the EOR fee really the cost, or is it the visible part of the cost? The EOR fee is what shows up on the invoice. The cost of an entity shows up across payroll, accounting, statutory filings, local director fees, registered office, and a wind-down you'll pay the day you leave. None of those line items is large. Together they're a permanent overhead. What happens if you get this wrong is that the entity quietly costs more than the EOR for years, and no one notices until someone leaves and the wind-down bill lands.
What would the business look like at three times the current headcount in this country? If the answer is "very different," and the difference includes a sales team, a finance lead, or anyone signing contracts locally, you're not choosing between EOR and entity. You're choosing between EOR now and a forced conversion in eighteen months, under time pressure, with senior hires watching. What happens if you get this wrong is that you convert in a panic, and conversions done in a panic lose people.
What is the work, and is it the kind of work that needs a local counterparty? Some work is fine on an EOR forever. Some work isn't. The signal is whether your customer, your regulator, or your bank is asking who the local entity is. If the answer is no, the EOR continues to do its job. If the answer is yes, and the question keeps coming up, the constraint is being set by the market, not by you.
What would I have to believe to make this decision worth reversing in two years? This is the question that catches most switches. A COO incorporates in a country because the headcount justified it. Eighteen months later, the country is deprioritised, the headcount is half what it was, and the entity is still there. Switching back to an EOR isn't a thing. You can do it, but the people on the ground carry the cost. The question isn't whether the entity was the right call. The question is whether you would make the same call again, knowing the headcount is now half.
Three Honest Categories the Approaches Split Into
The light footprint. An EOR, contractors, or a small hybrid where most of the headcount sits on one structure and a few sit on another. This category is right when the country is exploratory, when the work is mostly commercial, and when the headcount is small enough that the per-employee fee isn't yet the loudest line in the budget. It fails when the work starts to require a local counterparty, when senior hires want equity arrangements the EOR can't administer cleanly, or when the headcount has crossed a level where the per-employee fees have become the line finance keeps asking about. A regional sales lead, two account executives, and a customer success manager hired in the last nine months in a market the company entered this year. That's the picture. It works until one of three things changes.
The middle path. A branch or representative office, or a professional employer organisation arrangement that gives you more local presence without a full entity. This category is right when you need a local address and a local signatory, but you don't yet need a full local balance sheet. It fails when the regulator or the customer requires a full local company, not a representative office. When the tax authority asks for a local entity, a branch stops being a creative answer and starts being a problem. A company that has outgrown the EOR but isn't ready to commit to a full local structure. The middle path buys time. It doesn't buy a permanent solution.
The full commitment. Your own local entity, run by a local director, with its own payroll, its own accounting, and its own statutory obligations. This category is right when the headcount is large, the work is operational rather than exploratory, and the commitment to the country is multi-year. It fails when any of those conditions aren't met, and it fails expensively. An entity isn't a cost-saving measure. It's an admission that the country is part of the company. A company that has decided the country is permanent, that the work there's operational, and that the headcount is going to grow.
Five Diagnostic Questions to Self-Assess Against
Is the country on a three-year plan, not a one-year hope? Pull the strategy document. Find the country. Read what it actually says, not what was said in the board meeting. If the country is named as a core market with specific revenue or product commitments attached, the answer is probably yes. If the country is one of fourteen in a slide titled "explore," the answer is no, and an entity is the wrong tool for an exploration.
Has the work shifted from selling to operating? Selling in a country can be done by an EOR for a long time. Operating in a country, with engineers, support staff, finance, and people who manage other people locally, changes the calculus. Look at the job descriptions of the people already hired. If the work is mostly finding customers, the EOR is fine. If the work is serving customers that have already been found, the conversation is different.
Can you name the local director, or have you thought about who that person is? A local entity needs a local director. That director is a person, with a name, a CV, and a set of obligations to the local regulator that are independent of your company. If you don't have a candidate in mind, and you've not budgeted the cost of finding one, you're not ready to incorporate. If you've, you're further along than most.
Is finance asking the question because the fees are large, or because the fees are unexplained? Fees that are large but well-understood are a different problem from fees that are unexplained. The first is a cost question. The second is a vendor question. Address the second before you do anything structural. An entity doesn't become a better answer just because the EOR relationship has drifted.
What is the wind-down plan if you leave the market in three years? This is the question most readers can't answer, and it's the most important one. A wind-down has a cost, a timeline, and a set of obligations to local staff. If you can't sketch the shape of that wind-down now, before you incorporate, you're committing to something you don't yet understand. Sit with the question until you can answer it. If the answer is "I don't know," that's the answer. Respect it.
Five Structures, Reviewed
Employer of Record
An EOR is the local employer on paper, running payroll, statutory filings, and the employee-level obligations a local entity would otherwise carry. It exists so the customer doesn't have to.
It earns a place in the structure mix when the country is new, the headcount is small, or the commitment is short. It's the right answer for exploration. It's the right answer for sales-led entry. It's the right answer for the first three to five hires in a market.
It falls short in specific ways. It doesn't handle senior equity cleanly. It doesn't let you sign local contracts as the local counterparty. It doesn't answer corporate tax, and the activities of the EOR, the part where it employs your staff, isn't the only activity that can create a permanent establishment. If your team is closing deals locally, holding inventory, or working from a fixed office, an EOR reduces the exposure but doesn't remove it.
Your Own Local Entity
A locally incorporated company, with its own payroll, its own statutory accounts, its own tax filings, and its own director. The customer becomes the employer of record for its own people.
It earns a place when the headcount has crossed a level where the per-employee fee is larger than the entity's running cost, when the work in the country is operational rather than exploratory, and when the commitment to the market is multi-year.
It falls short in ways that compound. It's expensive to wind down. It needs a local director, a local accountant, and a registered office. The obligations to local staff continue whether the business is doing well or badly. The cost of an entity isn't the incorporation. The cost of an entity is the next ten years of running it, and the day you decide you no longer need it.
A Branch or Representative Office
A presence that's not a full local company, often used to give the business a local address and a local signatory without the obligations of a full incorporation. The legal form varies by country, which is part of the problem.
It earns a place when the work in the country doesn't yet justify a full entity, but a local presence of some kind is required. A regulator, a bank, or a customer may require a local address. A representative office can be the answer.
It falls short when the work crosses from "we've a local address" to "we've a local team doing operational work." A representative office is a presence, not an employer, and the line between the two is drawn by the local regulator. What you can do through a representative office is narrower than most readers expect, and the consequences of crossing the line fall on the parent company.
Contractors Engaged Directly
Hiring people as independent contractors, paid against invoices, with no local employer of record and no local entity. Common in the early days of a market entry, particularly for sales and consulting work.
It earns a place when the work is project-shaped, the person is genuinely running their own business, and the relationship is closer to a supplier than to a role. A consultant brought in to set up a channel partnership, for example.
It falls short, and it fails badly, when the relationship is actually employment wearing a contractor's clothes. The IRS uses a common-law test weighing all factors together, with no single factor decisive and no presumption either way. California applies the ABC test, which presumes the worker is an employee unless the hiring entity proves all three prongs: freedom from control, work outside the usual course of the hiring entity's business, and an independently established trade. A company can satisfy the IRS test and still fail the ABC test. The misclassification exposure is real, and it sits with the hiring company, not with the contractor. Treat this structure as a narrow tool, not a general one.
A Professional Employer Organisation Arrangement
A PEO is similar to an EOR in shape but different in intent. It's a co-employment model in which the PEO shares employer responsibilities with the customer, often used when the customer wants a more involved local partner than an EOR provides.
It earns a place when the work is operational, the headcount is meaningful, and the customer wants a closer relationship with the local employment function than a pure EOR offers. In some markets, this is the dominant structure for mid-sized international teams.
It falls short when the customer actually needs a local entity for non-employment reasons, such as signing local contracts, holding a local bank account in the company's own name, or being the counterparty a regulator recognises. A PEO doesn't solve those problems. It's an employment solution, and the limits of employment solutions are the same in this structure as in any other.
A Hybrid: Entity in Core Markets, EOR Elsewhere
A structure in which the company has incorporated in the countries it has committed to, and uses an EOR in the countries it's still learning about. The two structures coexist, and the boundary between them moves as the business commits or withdraws.
It earns a place when the company has a tiered view of its international footprint, with some markets treated as core and others as exploratory. It's the most common shape for companies that have been operating internationally for more than three years and have learned, the hard way, that a single structure across all markets is the wrong answer.
It falls short when the boundary between the two structures isn't actively managed. An entity in a country that no longer deserves one is a cost without a benefit. An EOR in a country that has outgrown the EOR is a constraint. The hybrid works only if someone owns the decision about which countries sit on which side, and reviews it at least annually.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| New market, first 1-4 hires, work is commercial | Small | EOR | Per-employee fees, no local counterparty | Stay on EOR, push for better terms |
| Country is 12-18 months old, 5-15 hires, work is operational | Mid | EOR with friction | Benefits administration, onboarding speed, equity questions | Stay on EOR, address operational friction, revisit in 12 months |
| Country is 24+ months, 20+ hires, mix of functions | Large | EOR is the ceiling | Equity, local contracts, customer asks for local entity | Begin incorporation planning, expect 6-12 month transition |
| Country is core, multi-year commitment, 50+ hires | Core | Your own entity | None structurally, but permanent overhead is real | Confirm the wind-down plan you would not need |
| Work is project-based, person is genuinely independent | Variable | Contractors | None if structure is right | Keep as is, monitor for misclassification risk |
| Multiple countries, mixed commitment, no consistent review | Portfolio | Hybrid at risk | Boundary between EOR and entity has drifted | Audit each country against the five diagnostic questions |
| Country manager with signing authority, no local entity | Small but exposed | EOR, but the activities are wider | Permanent establishment risk from manager's actions | Take tax advice, do not assume the EOR covers the gap |
| Workforce split between employment and equity-grant senior hires | Mid to large | EOR with workaround | Equity administration across borders | Move to entity or accept the friction as the cost of optionality |
The Costs Finance Usually Leaves Out
| Obligation | What it looks like in practice | When it bites |
|---|---|---|
| Local director | A person whose name is on filings, who carries personal liability, who must be replaced if they leave | From day one of incorporation |
| Statutory accounts | Annual or more frequent filings, often requiring a local auditor | Year one, every year |
| Tax filings | Corporate tax, payroll tax, social contributions, VAT or its local equivalent | Monthly, quarterly, annually, depending on country |
| Registered office | A physical address the regulator can write to | From day one, until wind-down is complete |
| Local payroll | Either a local provider or in-house capability | Every pay cycle, every employee |
| Local benefits | Statutory and market-practice benefits, often running parallel to the EOR's offering | Every employee, every year |
| Wind-down cost | The cost of closing the entity, settling obligations to staff, deregistering with the regulator | Only at the end, but inevitable at some point |
| Equity administration | Local sub-plans or shadow arrangements for senior hires | When the first senior hire joins |
Each of these is small. None of them appears on a single line of the kind of spreadsheet the board reviews. Together, they're a permanent overhead that runs for as long as the entity exists, which is usually longer than the original business case assumed.
A Switching Sequence That Does Not Break Employment
| Step | What happens | Why it matters |
|---|---|---|
| Confirm the decision | The country passes all five diagnostic questions, and the wind-down plan is sketched | Without this, the sequence is a guess |
| Incorporate the local entity | The new local company is registered, the local director is appointed | This is the slow step, and it sets the calendar |
| Run parallel payroll | For one or two cycles, the EOR and the new entity both run, with reconciliation | Catches errors before they become employment problems |
| Transfer employment | Employees move from the EOR's local entity to the customer's new local entity | The legal transfer, not a rehire, with continuity of service preserved |
| Communicate to staff | In writing, in the local language, explaining what changes and what does not | Most staff questions are about benefits and tenure, not structure |
| Wind down the EOR relationship | The country is moved off the EOR's books, the contract for that country is closed | EOR fees stop only when the relationship is formally closed |
| Audit at 12 months | A review of the new structure against the original business case | Most switches look good in month three. The real test is month twelve. |
The transfer is the moment that matters. If the new entity is treated as a rehire, service resets, vacation accruals reset, and the people who matter most find out about it. If the transfer is structured as a continuity event under local law, service carries over, and the people stay. Take local advice on the structure of the transfer. The shape of the question is the same in most countries, but the answer isn't.
The Cost of Getting This Wrong
The second-order costs never appear on an invoice. They show up in the form of a senior hire who left because the equity administration was wrong, in the form of a regulator who asked why a local director had not been replaced, in the form of a tax authority that reassessed the prior three years because the structure didn't match the activities, in the form of a wind-down bill that arrived eighteen months after the decision to leave the market. None of these is the kind of cost that shows up on a per-employee comparison. All of them are real, and all of them arrive after the spreadsheet has been closed.
So the question the COO is left with, in the meeting that started all of this, isn't "how do I save money on per-employee fees." The question is whether she is buying a saving, or buying an obligation. The answer depends on whether the country is going to be part of the company for the next decade, or for the next eighteen months. If she can answer that question with conviction, the rest is arithmetic. If she can't, the EOR isn't the expensive option. It's the option that lets her answer the question later.
When You Are Ready to Go Further
If the diagnostic questions in this article have given you a clearer picture of where each of your countries sits, and the decision table has given you a starting point, the next step is to pressure-test that picture against the structures other operators in your position have actually used. HROpsLab maintains an independent library of side-by-side comparisons of employer of record arrangements, professional employer organisation models, and the entity structures companies run themselves, with the trade-offs each one carries in practice.
The comparison work is paid for by the publication. It isn't sponsored by vendors, and the providers themselves don't see the editorial conclusions before publication. HROpsLab doesn't sell software, doesn't provide consulting, and doesn't give legal advice. If you want help interpreting the comparison in the context of a specific country or a specific headcount, the team can point you toward advisers who do that work, and away from the ones who don't.
Frequently Asked Questions
At what headcount does an entity make sense?
There's no universal number, because the answer depends on the per-employee fee, the country's running costs, and the wind-down exposure you're willing to carry. A more useful question is whether you can pass the five diagnostic questions in this article, in particular the three-year plan and the wind-down plan. If you can, the headcount is large enough. If you can't, more headcount doesn't change the answer.
Does service carry over when employees transfer from an EOR to my own entity?
In most countries, a structured transfer of employment preserves continuity of service, which means tenure, vacation accruals, and any entitlements tied to length of service carry over. The legal mechanism isn't a rehire, it's a transfer, and the difference matters. The specifics are local, so take advice on the form the transfer takes in the country in question.
Who is the employer during a transfer?
There's a window, usually one or two pay cycles, where the EOR and the new entity are both in play. The cleanest approach is to run parallel payroll, with the EOR paying through the cutover date and the new entity paying from the cutover date forward. The employee is never without a paycheque, and the legal employer is unambiguous at every moment.
Is an EOR a permanent arrangement?
It can be, for some companies in some countries. For most readers, the EOR is a phase, and the question is how long the phase lasts. If the country is permanent and the headcount is large, the entity is usually the better answer. If the country is exploratory, the EOR is usually the better answer. The middle is where most readers live, and the middle is where this article earns its keep.
What happens to benefits when we switch?
In most countries, statutory benefits transfer with the employment. Market-practice benefits, such as supplemental health cover or pension contributions, may need to be re-established under the new entity. The EOR's benefits package isn't portable in the way the employment is. Plan for a short window where benefits administration is in transition, and communicate that window clearly to staff.
How long does incorporation take?
It varies by country, and the shape of the variation matters more than any specific number. Some countries can incorporate in weeks, others take many months, and the bottleneck is often the local director or the registered address rather than the filing itself. Confirm the timeline locally, and build the plan around the slowest moving part.
Can we run both an EOR and an entity in the same country?
Yes, and some readers do, particularly during a transition or where the work splits between employment and contract. The cost of running both is the obvious friction. The less obvious friction is the governance overhead of two structures in one country, and the need to be clear about which employees sit on which structure and why. Most readers find that one structure per country is simpler, even if the choice between them is hard.
A company exists to take the friction out of cross-border employment decisions, one comparison at a time. The library is open, the work is independent, and the conclusions aren't for sale.
