Own entity vs ICP for Italy: the operations view for EOR platforms

Most comparisons of an own entity and an in-country partner (ICP) are written for the client. They ask which route is cheaper for the employer. This article asks a different question. Which route is better for your platform’s P&L, and for the team that has to run Italy every month?

The decision is about your margin, not the client’s  

An own entity gives you the full gross margin on every Italian employee. It also gives you the full fixed cost of being present. An ICP turns much of that fixed cost into a variable one. You pay for what you use, and you share the margin.

So the real question is volume. Below some headcount, fixed costs of an entity are not absorbed. Above it, the partner’s share of margin costs you more than a local team would. Where that crossover sits is specific to your platform. We do not give a figure here, because it depends on your pricing, your mix of roles and your cost of local staff.

What you can do is model it. Take your expected Italian headcount over the coming planning period. Add the fixed costs of the entity: registrations, local accounting, legal cover, insurance, a responsible person, audit and filing. Compare that with the partner’s fee and revenue share at the same headcount. The line where the two cross is your threshold.

Headcount thresholds in practice  

There are usually three zones.

Low volume. Few employees, uncertain pipeline. An ICP is almost always the better choice. You avoid fixed cost and you can stop without winding anything down. Your risk is concentration in the partner, not in the entity.

Middle volume. This is where most platforms get stuck. The entity looks attractive on paper, but the team you need to run it is under-used. Payroll runs, authority correspondence, onboarding and terminations each need a trained person. If volume is uneven, those people sit idle in some months and overloaded in others.

High volume. The partner’s margin share becomes visible in your reporting. At this point an own entity, or a hybrid, often wins on cost. It also gives you the control that large accounts ask for in their procurement questionnaires.

Do not set the threshold only on current headcount. Use committed pipeline, and apply a realistic conversion rate to it. Platforms often build an entity for a pipeline that never fully closes.

Staffing: who actually does the work  

An own entity means you hire for Italy. You need people who know Italian payroll, collective agreements, social security filings and the habits of local authorities. They need to work in Italian with advisers, unions and employees. Few platforms can cover that from an existing team abroad.

With an ICP, the partner carries that staffing. Your own team shrinks to a governance role. It checks outputs, manages escalations and owns the client relationship. That is a smaller team, and a cheaper one, but it needs different skills. Vendor management and data quality matter more than payroll depth.

There is also a continuity point. Italian expertise is not easy to replace at short notice. If one person holds the knowledge of your Italian setup, you carry a key-person risk. A partner with a bench reduces that risk. An own entity with a thin team increases it.

Control: where each model gives and takes  

Own entity:

• You set the service levels and the payroll calendar.
• You hold the data and the bank flows.
• You decide how to handle exceptions and corrections.
• You also own every error, every late filing and every dispute.

ICP:

• You depend on the partner’s calendar, cut-off times and tooling.
• Your visibility depends on the data you receive and how fast.
• Liability allocation sits in the contract, so the contract needs care.
• You can change partner, but migration of live employees is real work.

Control is not only a risk topic. It affects your product. If you want a single employee experience across countries, you need structured data from Italy in a format your system can read. Under an ICP this is a contract and integration question. Under an own entity it is an internal build.

What to check in the ICP model  

If you lean toward a partner, test these points before you sign.

• How does the partner exchange data with your platform? Files, API, or email?
• How are payroll inputs cut off, and what happens to late changes?
• Who answers employee questions, in which language, and in what time?
• How are errors reported, corrected and costed?
• What are the exit terms, and how is employee data returned?
• Can the commercial model scale down as well as up?

A partner that cannot answer these clearly will cost you more in operations than the fee suggests.

A hybrid path  

Many platforms start with an ICP and keep the option to move to an own entity later. This works if you plan for it from the start. Keep your employee data in your own system. Keep contract terms that allow transfer. Track your Italian cost per employee from day one, so the switch decision rests on your numbers and not on opinion.

The reverse also happens. A platform opens an entity, finds volume lower than planned, and moves part of the load to a partner. That is harder, because fixed costs are already in your P&L.

Summary  

Choose an ICP when volume is uncertain, when you lack Italian staff, or when speed matters. Choose an own entity when volume is proven, when margin share is material, and when you can staff the team properly. In between, use a hybrid and measure.

The decision should come from your own cost model, not from a general rule. The data you exchange with the partner is what makes that model reliable.

Partnership teams are invited to request the data-exchange specification.