An Australian SaaS founder closes her seed round and immediately faces the question every modern startup faces: she needs a senior engineer in Berlin, a customer success lead in Singapore, and a part-time growth marketer in São Paulo. None of them are willing to relocate, and her runway will not stretch to flying out and incorporating in three jurisdictions. Her accountant has mentioned “EOR” in passing. Her cousin who runs a Delaware-flipped company says she should “just use contractors.” Her board observer wants to know what her “permanent establishment exposure” looks like.
The honest answer is that there are three credible structuring options for hiring across borders — engaging the worker as a contractor, using an employer of record (EOR), or standing up a foreign subsidiary — and each one prices a different bundle of risks. Most Australian startups end up using all three at different points in their growth, often in the same year. Knowing which to reach for at which moment, and which red lines to never cross, is what separates the founders who scale cleanly from the ones who inherit a tax or employment problem at Series B.
The Three Models in One Paragraph Each
Contractor. The startup engages the worker on an independent contractor agreement, usually with the worker invoicing through their local entity or ABN-equivalent. Cheapest and fastest. No payroll, no statutory benefits, no local registration. But the relationship has to be genuinely a contractor relationship in substance — and not just in the country where the worker sits, but in Australia too.
Employer of Record (EOR). A third-party provider (Deel, Remote, Oyster, Velocity Global, Multiplier and others) employs the worker locally as their statutory employer, runs local payroll, handles statutory benefits and tax withholding, and on-charges the startup for the cost plus a margin. The startup directs the work day to day, the EOR carries the local employer obligations. Typical cost is $400–$700 per worker per month on top of salary.
Foreign subsidiary. The startup incorporates a wholly-owned subsidiary in the target country, registers as an employer locally, opens a bank account, files local tax returns, and employs the worker directly. Highest cost and complexity. But it is the only model that scales, and it is the only one that gives you genuine operational control of a local business.
What Actually Drives the Choice
Founders often choose between these three on cost alone. The more useful frame is to look at the four legal risks they price differently.
Permanent establishment (PE) risk. A foreign worker can, in the wrong configuration, create a “permanent establishment” of the Australian parent in their country — exposing the parent to local corporate income tax on attributable profits. The trigger is usually one of two things: a fixed place of business (the worker’s home office, if it functions as the company’s effective office in that country) or a “dependent agent” (a worker with authority to negotiate and conclude contracts in the company’s name). The OECD’s November 2025 update to the Model Tax Convention sharpened both tests, introducing a 50% working-time benchmark and a “commercial reason” test for home-office PE. Sales hires with authority to close deals are the highest-risk category. EORs reduce — but do not eliminate — PE risk; foreign subsidiaries create their own PE deliberately and ringfence it.
Misclassification risk. Treating an effective employee as a contractor is illegal in both directions. Locally — in the worker’s jurisdiction — most countries now have aggressive misclassification regimes (the US ABC test, EU “presumption of employment” reforms, Singapore’s Tripartite Standard). And in Australia, the Fair Work Commission’s 2023 decision in Doessel Group v Pascua held that a Philippines-based “contractor” was in substance an employee covered by the Fair Work Act 2009 (Cth). The post-Closing Loopholes sham-contracting regime applies whether the worker sits in Sydney or São Paulo.
Payroll and superannuation exposure in Australia. Section 12(3) of the Superannuation Guarantee (Administration) Act 1992 extends the SG net to anyone working under a contract “wholly or principally for their labour.” Where an Australian company engages an overseas individual on a labour-hire basis, that section can reach across borders. State payroll tax can also bite under the harmonised “relevant contracts” provisions in NSW, Victoria and elsewhere.
Co-employment and EOR-specific risk. EORs are not a legal forcefield. The arrangement only works if the EOR’s employment of the worker is genuine and the startup respects the boundaries — no direct contracting with the worker, no parallel “consulting agreement,” no granting of authority to bind the parent in the local market. Where the substance breaks down, regulators in the worker’s country can collapse the structure and treat the Australian parent as the true employer.
A Stage-Based Heuristic
The pattern that works for most Australian startups looks like this.
Pre-Series A, one or two workers per country. Use EOR for anyone who looks like an employee — full-time, exclusive, directed. Use a clean contractor agreement only for workers who are genuinely operating an independent business (multiple clients, project-based, own tools, fixed deliverables). Never use a contractor agreement to “save costs” on someone who is operationally an employee. The cost of an EOR is trivial relative to the cost of a misclassification finding plus retrospective payroll, super and tax.
Series A, 3–10 workers in a single country. Continue with EOR until the headcount or the strategic importance of the market justifies a subsidiary. The crossover point is typically around 8–12 employees per country, where the EOR margin starts to exceed the cost of running a local payroll. Sales leadership in a target market is often the trigger for incorporation, because the PE risk of an EOR-employed sales head with binding authority is meaningful.
Series B and beyond. Move to subsidiaries in countries that are core to revenue or product. Keep EOR for “long tail” countries where one or two specialists sit. Maintain a clean contractor population only for genuinely independent vendors.
Five Things to Get Right Regardless of Model
Where IP sits. Every contract — contractor, EOR sub-employment, subsidiary employment — must assign all IP created in the course of the engagement to the Australian parent. EOR templates do this, but the assignment chain from EOR to startup needs to be checked. Investors will diligence this.
Restrictive covenants. Confidentiality, non-solicit and (where lawful) non-compete clauses must be drafted to the worker’s local law, not Australian law. Australian boilerplate non-competes are unenforceable in much of Europe and parts of the US.
Equity. Granting options or shares to overseas workers triggers local securities, tax and exchange-control issues. Plan the ESOP rollout country by country; do not promise equity until you have checked the granting mechanism in each jurisdiction.
Termination. “At-will” employment exists in essentially one country (the US). Everywhere else, termination requires statutory cause, notice, and often severance. EORs handle the mechanics; subsidiaries do not get to ignore them; contractors should not be terminated like employees.
Data flows. A worker overseas almost always means cross-border personal-information transfer, which engages the Privacy Act 1988 (Cth) and the worker’s local data law. Vendor due diligence on the EOR’s data practices is now a standard ask.
What Founders Should Do
Start with the risk map, not the price list. Identify which of the four risks above applies to the role and the country. A senior salesperson with binding authority needs a different structure to a backend engineer working asynchronously.
Default to EOR for first hires. It is the cleanest balance of speed, compliance and cost for the first one or two roles in any country.
Resist the contractor shortcut. The cost difference between EOR and contractor is small. The legal difference, if the relationship is misclassified, is enormous — particularly in light of Doessel Group v Pascua and the post-Closing Loopholes enforcement environment.
Plan the subsidiary trigger early. Know the headcount or revenue point at which you will incorporate locally, and start the lead-time on registration before you hit it. Subsidiaries take months in some jurisdictions.
Document everything for the data room. Investors performing legal diligence in 2026 expect to see EOR agreements, contractor classifications, IP assignments, and a written rationale for the choice of model in each country. Build the artefacts as you go.
The Bottom Line
The choice between EOR, foreign subsidiary and contractor is rarely about which is “best.” It is about which is right for this role, in this country, at this stage of the company. Founders who think in terms of risk allocation — permanent establishment, misclassification, super, co-employment — make cleaner decisions than founders who think in terms of monthly cost. The startups that scale internationally without legal scars are almost always the ones that started with EOR, knew when to incorporate, and never used a contractor agreement to dress up an employee.
Viridian Lawyers advises Australian startups on international expansion, cross-border employment, and the structuring choices that come with hiring across jurisdictions. If you are about to make your first overseas hire and want to think through the right model, get in touch.
