How to Choose an EOR

Only 5 to 6 EOR providers have strong direct infrastructure outside Western Europe and North America, which narrows the decision faster than most buyers expect. The four variables that matter: the jurisdictions you are hiring in, headcount per country, your compliance risk tolerance, and your budget. For most companies adding 1 to 10 employees in a new market, picking a provider in five working days under interview pressure is the realistic scenario most teams face.

Picture the moment that pushes most People Ops leaders into EOR territory. The CEO has just promised the board a sales hire in Singapore in six weeks. Legal will not sponsor an entity for one headcount.

The candidate is interviewing elsewhere.

You have a Friday to pick a partner, a Monday to start the contract, and no second chances if the first payroll runs late or the visa documentation comes back wrong.

Choosing an Employer of Record at that pace is less about features and more about which provider will not break the hire.

This guide is written for the buyer who has already decided that an EOR is the right route and now needs to pick one. It assumes you know what an EOR does. It does not waste pages re-explaining the model.

Instead, it walks through the decisions that separate a provider that ships hires cleanly from one that creates a quiet drag on your team for the next two years.

The three decisions to make before you even look at vendors

Most evaluations go wrong before the first demo because the buying team has not agreed on three internal questions. Vendors will happily take your team in any direction once you start talking, and the answers will look different depending on what you actually need.

Lock these three down inside your own organisation first, in writing, before any sales call.

Decision one: is this a bridge or a destination?

A bridge engagement runs nine to twenty-four months while you build a local entity; bridge buyers care about migration tooling and data portability. A destination engagement is the permanent home for the hire; destination buyers care about long-tail country depth and ongoing benefits depth. A provider excellent for a two-year bridge can be the wrong choice for a permanent arrangement.

Decision two: how many countries, and which ones?

Single-country buyers should weight country depth heavily; multi-country buyers should weight consistency across markets and centralised reporting. A shortlist dominated by tier-one markets rewards owned-entity providers. A list that includes Vietnam, Nigeria, or Kazakhstan exposes the limits of owned-entity providers and forces a partner-network conversation.

Decision three: who owns the relationship internally?

TA-led programmes will not tolerate slow contract issuance. People-Ops-led programmes will not tolerate poor HRIS integration. Finance-led programmes will not tolerate opaque pricing.

Decide the internal owner first; the shortlist follows from that.

Payoff: if you cannot answer these three questions in a single sentence each before your first vendor call, the demo is going to lead you, not the other way around.

Owned-entity vs partner-network EORs: why the infrastructure model matters

The single most useful filter in EOR selection is the provider’s infrastructure model. Providers operate either their own legal entities in the countries they serve, a network of local in-country partners, or a hybrid of the two.

Each model has consequences that do not show up in the brochure and that you will live with for the entire engagement.

Owned-entity providers

An owned-entity provider controls compliance interpretation, contract templates, and payroll directly through its own subsidiary. Quality is consistent, escalations stay in one organisation, and data flows through one system. The trade-off is geography: owned-entity EORs cover thirty to sixty countries, not one hundred and sixty.

If your roadmap includes Senegal or Uzbekistan, the provider will quietly route through a partner anyway.

Partner-network providers

A partner-network provider contracts with local specialists and resells their service. The advantage is reach: 150-plus countries by adding contracts, not building entities. The trade-offs are supply-chain risk and inconsistency: the legal employer is a third party you have never spoken to, compliance interpretation varies by partner, and the EOR’s SLA is only as good as the partner’s ability to honour it.

Hybrid providers

Most large EORs are hybrid: owned entities in tier-one markets, partner network in long-tail. Always ask which countries fall into which category: it determines who employs your worker and who is liable. A reputable provider will tell you; a less reputable one will deflect with talk of its global compliance team.

Payoff: ask every shortlisted provider for an entity-by-entity breakdown of every country you plan to hire in. Refuse to accept a marketing list. The answer to that single question separates serious operators from glossy ones.

Country coverage depth: how to test beyond the marketing list

Marketing pages list two hundred countries. The reality is that an EOR can be excellent in fifteen, competent in another twenty, and barely operational in the rest. The risk is concentrated in the countries you care about most.

There are five tests that strip the gloss off a coverage claim, and a sophisticated buyer should run all of them on every shortlisted provider.

Test one: how many active employees in this country, today?

Ask for active worker headcount per target country. Fifty workers in Germany means a working playbook; three means untested edge cases. A provider that refuses to give numbers is telling you something.

You need order of magnitude, not precision.

Test two: who is the in-country payroll lead, and where do they sit?

Ask for the name and location of the in-country payroll specialist. A named partner in-country means investment. A regional manager plus a shared-services queue means they have not invested.

Test three: show me a recent statutory change you implemented

Ask for a specific statutory change in that country in the last twelve months and how it was implemented. France’s BDESE reporting, Germany’s minimum-wage adjustments, and the UK’s Working Time amendments are real recent examples. Providers with depth answer fluently; thin providers give generic statements.

Test four: contract templates by country

Ask to see the country-specific employment contract template. Visibly localised contracts signal depth. Templates clearly adapted from a US original, lacking clauses on probation, severance, and notice, indicate a thin operation.

Test five: customer references in your target country

Ask for two references from clients with employees in the same country. A thin provider will offer references from other countries or industries and hope you do not push back.

Payoff: you are not buying a coverage map; you are buying compliance work in specific countries. Test the specific ones, not the average.

Compliance guarantees and SLAs: what to demand in writing

Every EOR markets compliance. The question is what they commit to in the contract when they fail to deliver. Most buyers accept the standard MSA without negotiating, then discover the SLAs were aspirational.

Five clauses are worth fighting for.

First-payroll SLA

The first payroll is where most EOR engagements break. Ask for a written SLA: the first payroll runs on the country’s standard date if documents land seven business days in advance, with a fee credit if the EOR misses through its own fault. Anything weaker means they are telling you they will try.

Compliance indemnity

A defensible indemnity covers misclassification claims, unpaid statutory contributions discovered after the fact, and regulator penalties arising from the EOR’s own interpretation. Push back on caps equal to or below twelve months of fees.

Support response times

Ask for response-time SLAs by ticket category: payslip query within one business day, compliance question within two, payroll error acknowledged within four hours. If the contract has no response SLA, you have no escalation path when service degrades, and service does degrade.

Statutory change notification

The contract should oblige the EOR to notify you of cost-impacting statutory changes at least thirty days before they take effect. A structured advisory process (email plus dashboard plus quarterly review) is the standard; a newsletter you are expected to read is not.

Termination and offboarding SLA

Negotiate a clause obliging the provider to complete final payroll, statutory filings, and document handover within a defined window after termination. Without this, the provider can slow-walk records you need to migrate the worker.

Payoff: a provider that resists every one of these clauses is telling you they expect to underperform and want the right to do so without consequence. Treat that as the answer.

Exit terms and offboarding: what happens when you no longer need the EOR

Almost no buyer evaluates exit terms during selection, and most regret it. Three exit scenarios deserve attention before you sign.

Scenario one: you open your own entity in the country

The questions are whether the workers’ employment can transfer cleanly under TUPE-equivalent legislation, whether historical payroll data is released in a usable format, and whether there is an exit fee. A provider with clean terms will commit to a structured handover within thirty business days at no additional cost. Negotiate the handover scope before signing.

Scenario two: you switch to a different EOR

The worker’s employment terminates with the original EOR and restarts with the new one, which can break statutory continuity for benefits and severance accrual. A serious provider helps structure the transition to preserve continuity; a weaker one treats it as a clean termination.

Scenario three: you offboard the worker entirely

Termination in France can take three to six months; the Netherlands requires UWV approval or a settlement; California rules differ from New York. Ask the provider to walk through a real termination playbook per target country. Uncertainty about the steps in any given country is a signal.

Data ownership and portability

Across all three scenarios, the contract should state explicitly that the client owns all employment records and that the provider delivers them in machine-readable format on termination. Without this clause, some providers delay or charge for data extraction.

Payoff: read the termination clause out loud before signing. If it makes you feel trapped, the rest of the contract will too.

How to run a meaningful EOR evaluation in four weeks

Four weeks is realistic and produces a defensible choice if structured properly.

Week one: internal alignment and shortlist

Lock the three internal decisions described earlier: bridge or destination, country list, and internal owner. Build a one-page brief that captures hire volume, target countries, HRIS, integration requirements, budget envelope, and timeline. Build the shortlist to four providers.

Three is too few because one will drop out; five is too many to evaluate properly in the time available. Mix at least one owned-entity-leaning provider with one partner-network-leaning provider so the comparison surfaces the trade-off rather than hides it.

Send the brief to all four providers and book ninety-minute working sessions for week two.

Week two: structured demos with prepared questions

Run the demos against a fixed agenda, not the provider’s deck. Cover: country-by-country entity status, in-country payroll lead names, recent statutory change examples, contract template review, integration deep-dive with your HRIS, SLA wording, and pricing model.

Take notes in a shared scoring sheet that all internal stakeholders see, with one row per question and one column per provider. Avoid the temptation to let the provider drive the conversation.

The deck is the same in every demo; your scoring sheet is what makes the comparison real.

Weeks three and four: references, contract review, and decision

Take two references per provider in your target countries. Have legal review the MSA and flag the indemnity, SLA, and termination clauses. Build the twelve-month total-cost model with your real roles and countries.

Narrow to two providers, run a final negotiation on SLA wording, indemnity caps, and exit terms, then sign.

After signature: set a quarterly review cadence from day one covering SLA performance, ticket volumes, and open compliance questions by country. EORs that perform well in year one routinely degrade when account teams are reassigned. The quarterly review catches that early.

Frequently asked questions

How long does a typical EOR engagement last?

Bridge engagements typically run eighteen months to three years before entity setup. Destination engagements last as long as the worker is employed. The bridge-or-destination decision should drive your shortlist before you evaluate providers.

Can the EOR sponsor work visas?

Coverage varies widely. The UK, Singapore, and Canada are widely covered; Germany, France, and the Netherlands are partial. Many countries do not allow EOR entities to sponsor visas at all.

Confirm per country in writing before assuming it is possible.

What happens if the EOR loses its operating licence?

The worker’s employment is technically void from that date. Owned-entity providers carry less risk; partner-network providers more. Ask every shortlisted provider for their continuity plan in this scenario.

Is the EOR liable if my classification turns out to be wrong?

Generally yes. The EOR carries compliance liability as legal employer, but if your instructions caused the misclassification, liability can shift back to you. Read the indemnity clause carefully; reputable EORs accept liability for their own compliance interpretation, not the client’s deliberate instruction.

Can I have the same EOR run my contractors as well as my employees?

Most EORs now offer contractor-of-record on the same platform. One invoice, one integration is the upside. The risk is that consolidating can mask workers who should be employees but are being run as contractors for cost reasons.

Have compliance review the contractor population before consolidating.

How quickly can an EOR onboard a new hire?

Three to five business days in tier-one countries; ten business days is the median. The variable is country, not provider: Brazil takes longer than the UK regardless of EOR. Ask for country-specific timelines, not the headline average.

Do I need legal review of every EOR contract template?

Legal review of the MSA is non-negotiable. You generally do not need to review every country-specific employment contract: that is the EOR’s compliance work and their liability. Reviewing templates for your two or three most critical countries is a useful depth check.