Should a European company use an employer of record, an outsourcing vendor, or its own entity in India?
Use an outsourcing vendor when you want delivered software and do not want to manage people abroad; use an employer of record (EOR) when you want to direct named individuals as if they were employees but do not want to open a company; and open your own entity when the offshore team is large enough, permanent enough or strategic enough that control and retention outweigh the cost of running a subsidiary. The three models differ in who employs the engineers, who directs them, and who carries the risk.
Most European companies get this decision wrong in one of two directions: they open an entity far too early for a team of six, or they run twenty embedded contractors through an EOR for three years and never revisit it. The rest of this guide gives you the decision rule.
How do the three models actually differ?
| Dimension | Outsourcing vendor | Employer of record | Own entity |
|---|---|---|---|
| Legal employer | The vendor | The EOR provider | You |
| Who directs daily work | Vendor delivery lead, against your roadmap | You | You |
| What you buy | Delivered outcomes and service levels | Employment administration for people you choose | Nothing — you operate it |
| Time to first engineer | Weeks | Weeks, once you have found the person | Months, including registration and banking |
| Recruiting effort | Vendor's | Yours, or a separate recruiter | Yours |
| Cost shape | Rate per engineer or per outcome | Salary plus statutory costs plus EOR fee | Salary plus statutory costs plus entity overhead |
| Scaling down | Contract notice period | Local termination rules apply | Local termination rules apply |
| IP ownership | By assignment in the contract | By employment terms and assignment | Direct, subject to local law |
| Tax footprint | Purchase of services | Depends on control and duration | Taxable presence by design |
| Best at | Throughput and speed | Bridging while you decide | Long-term retention and control |
When is an outsourcing vendor the right choice?
Choose a vendor when the constraint is delivery capacity rather than organisational control. It is the fastest route to a working team, because recruiting, employment, workspace, HR and replacement are someone else's problem, and because a good vendor can put a functioning squad — engineers, a lead, QA — in place rather than a collection of individuals you then have to assemble.
It is also the model that keeps your tax and employment exposure smallest, provided you buy outcomes rather than hours. The trade-off is that you do not control who is on the team beyond the contractual continuity terms, and institutional knowledge sits partly with the vendor. Mitigate that with named-role continuity clauses, documentation requirements and a defined exit plan. Choosing well matters more than the model itself — our guide on how to choose a software outsourcing partner covers the evaluation criteria that actually predict delivery.
When does an employer of record make sense?
An EOR is a bridge, not a destination. It is genuinely useful in three cases: you have found one or two specific people you want and cannot employ them yet; you are testing a location before committing to an entity; or you are absorbing an acquired team while the corporate structure catches up.
The costs of using it as a permanent structure are real. You pay a fee per head on top of full local employment cost, so it is rarely cheaper than a vendor at scale. You do your own recruiting, so time-to-hire is yours to own. And because you direct the individuals directly, the arrangement leans toward the higher-risk end of the tax spectrum described in our article on permanent establishment and transfer pricing risk when offshoring development.
A reasonable rule: if you expect to still be using the EOR in eighteen months with more than about ten people, model the entity properly.
When should you open your own entity?
An entity earns its overhead when several of these are true at once:
- The offshore team is heading past roughly 25 to 50 people.
- The work is core product engineering you intend to own permanently.
- You want a local employer brand to compete for senior talent directly.
- Retention matters more than flexibility — you are optimising for five-year tenure, not five-month projects.
- You need IP, security or regulatory control that only direct employment gives you.
The overhead is not just registration. It is statutory filings, payroll compliance, an office lease, local HR, audit, transfer pricing documentation and a management layer that reports into the European organisation. That overhead is fixed, which is exactly why it is punishing for a team of eight and negligible for a team of eighty. The operational detail is in our guide to setting up a tech development centre in India.
Which model fits which stage?
| Your situation | Start with | Revisit when |
|---|---|---|
| First offshore engagement, scope still moving | Vendor, dedicated team | The team has been stable for a year and is core to the product |
| Fixed-scope build with a deadline | Vendor, outcome-based | The build turns into ongoing product work |
| Two specialists you have already identified | EOR | Headcount passes about ten |
| Testing a new location | EOR or vendor | You are confident in the location and the hiring market |
| Committed multi-year platform investment | Own entity, or build-operate-transfer via a vendor | Annually, against retention and cost data |
Can you combine them?
Yes, and mature setups usually do. A common sequence is: start with a vendor team to prove the location and the working model; keep the vendor for surge and specialist capacity; then open an entity for the core platform group once the shape of the team is stable. Build-operate-transfer formalises exactly this — the vendor recruits and runs the team, and you acquire it at an agreed point. The comparison in our post on GCC versus BOT versus outsourcing covers how those transitions are priced.
What does not work well is running all three in parallel with no distinction between them, so that engineers on three different contractual bases sit in the same squad doing the same work. That is confusing for the team and hard to defend to an auditor.
What should drive the decision?
- Control. Do you need to direct individuals, or do you need software delivered?
- Horizon. Under two years favours a vendor or an EOR. Over five years favours an entity.
- Scale. Fixed entity overhead only amortises above a certain headcount — calculate yours rather than assuming.
- Speed. If you need engineers this quarter, an entity is not a realistic answer.
- Risk appetite. Employment, tax and compliance exposure differs sharply across the three, as does who absorbs a bad hire.
Whichever route you pick, the local employment and compliance ground rules still apply — those are set out in our guide to legal and compliance considerations when hiring developers in India.
Next step
If your horizon is under two years, or you need working software before you need an org chart, start with a dedicated vendor team and keep the entity decision open. Our senior engineering talent service covers exactly that: a vetted India-based team with a delivery lead, structured so you can transition to your own entity later if the volume justifies it.
