In-house or outsourced payroll: from how many employees?

Quick answer. There is no legal threshold: the decision compares the full cost per payslip in-house (loaded salary of the manager, software, training, supervision time) with the price of outsourcing, weighted by complexity (multiple agreements, variables, turnover) and risk. In practice, hiring an in-house payroll manager is often justified only from several dozen payslips a month; below that, outsourcing usually remains cheaper and safer.

Payroll has become technical: monthly DSN, withholding tax, collective agreements, Urssaf audits. For a director, the question is not "who types the payslips" but "which organisation minimises total cost and risk". The trade-off between in-house and outsourcing therefore rests on two axes: full cost per payslip and the accepted level of risk.

This article offers a calculation method and decision criteria. For the detailed pros and cons of delegation, see our analyses on the pros and cons of outsourcing and on payroll outsourcing and its watch points.

The in-house cost is not limited to software. It includes:

Per payslip, this full cost falls as volume rises: a full-time manager produces many payslips, but their fixed cost weighs heavily on a small headcount.

Outsourcing to a firm is generally billed per payslip (and by scope: payslips only, DSN, advice, joiner/leaver handling). The cost is variable: it follows headcount, without the fixed cost of a salary or software, and includes expertise and legal updates.

To compare quotes, reduce everything to cost per payslip, same scope. See also our benchmark on payroll pricing by an accountant.

The switching point is not a universal number: it depends on the volume of payslips, complexity (several agreements, many variables, high turnover) and risk appetite. A few markers:

• Small headcount (up to about ten employees): outsourcing is almost always cheaper and safer; a manager's fixed cost does not pay off.
• Mid headcount: the decision turns on complexity and the need for responsiveness; a mixed solution (variables managed in-house, production and DSN by the firm) is often optimal.
• Large headcount: hiring a manager (or a team) becomes relevant, provided monitoring and continuity (absences, leave) are ensured.

Multiple collective agreements. They raise the in-house cost (setup and monitoring) and favour a firm's expertise.

Strong seasonality or turnover. Frequent joiners/leavers add to the workload; outsourcing absorbs variations better.

Fast growth. Start outsourced, internalise when volume and stability justify it; a mixed solution eases the transition.

Take a company hesitating to hire a payroll manager. In-house, the full cost adds the manager's loaded salary, the software subscription, training, monitoring and the supervision time spent on collection. This cost is largely fixed: it barely varies whether you produce ten or a hundred payslips. Per payslip, it therefore falls as headcount grows. With outsourcing, the cost is variable: it follows the number of payslips, without the fixed cost of a salary or software. As long as volume stays modest, the variable cost of outsourcing is below the fixed in-house cost per payslip; beyond a certain volume, the two logics converge then reverse, and internalising becomes competitive. The switching point depends on the price negotiated per payslip, the manager's salary and complexity (multiple agreements, many variables). It is this reasoning — not a ready-made threshold — that should guide the decision, adding the cost of risk and the need for service continuity.

1. Full cost, not just software: include loaded salary, training and supervision time.
2. Continuity: a single in-house manager creates a risk during absences.
3. Legal monitoring: DSN, withholding tax and agreements evolve; monitoring has a cost.
4. Liability: in-house, errors fall on the employer; check the split in an engagement letter if you outsource.
5. Reversibility: keep your data and settings so you can switch solutions.

The question "from how many employees" calls for a full-cost-and-risk answer, not a magic number. We see companies internalise too early — an under-used manager is expensive and bears the risk alone — and others outsource without scoping the perimeter, then be surprised by extras. The mixed solution (variables managed in-house, production and DSN delegated) is frequently the best compromise at mid headcount.

As a chartered accountant registered with the French Ordre des experts-comptables, I advise deciding over three years: project headcount, complexity and payslip volume, then compare the full cost per payslip of both scenarios. Link this choice with your monthly payroll cycle to clarify who does what.

Another criterion, sometimes decisive, goes beyond cost: confidentiality and continuity. Internalising payroll means giving an employee access to the pay of the whole company, directors included; some organisations prefer outsourcing for this reason, as it keeps payroll outside the internal perimeter. Conversely, outsourcing means accepting dependence on a provider and organising the timely transmission of variables. No model is inherently superior: it all depends on how sensitive the company is to these issues. We invite directors to set out their priorities — cost, confidentiality, responsiveness, continuity — before comparing options, because the right trade-off is the one that respects those priorities, not merely the cheapest per payslip.

Hayot Expertise advice. Before deciding, set out the full cost per payslip of both options, same scope, over a three-year horizon. Include risk (cost of an error) and continuity. For many small and mid-sized firms, a mixed formula avoids both the fixed cost of an under-used role and the loss of control of full delegation.