The in-house versus outsourced payroll question is almost always presented as a cost comparison, and almost always decided on something else — usually who carries the compliance risk when a return is rejected at 11pm on the fifteenth. Both framings are legitimate. The problem is that the published cost evidence is close to worthless, and the risk framing is rarely made explicit.
This piece separates the two. It sets out what market pricing genuinely supports, gives you a model for the side that nobody publishes honestly, identifies the headcount at which the arithmetic flips, and is specific about what does not transfer to a provider no matter what the contract says.
What the pricing evidence does support
Quoted outsourced pricing is more reliable than claimed in-house cost, because providers are publishing their own rate cards and several independent sources converge. Per employee per month, in Indian rupees:
| Service tier | Typical PEPM | What it covers |
|---|---|---|
| Basic processing | ₹100 – ₹200 | Salary computation and payslips. Statutory filings usually excluded |
| Compliance-only / hybrid | ₹150 – ₹250 | You process; the provider handles statutory filing |
| Fully managed | ₹300 – ₹500 | Processing, compliance filings, TDS returns, employee queries, year-end certificates |
| Mid-market full-compliance band | ₹300 – ₹800 | Reported sweet spot for 100–500 employees |
| Payroll plus broader HR outsourcing | ₹800 – ₹2,500+ | Adds recruitment, benefits administration, employee relations |
Compiled from published 2026 rate guidance across several Indian providers. These are quoted prices, not audited averages, and they exclude one-time onboarding or data-migration fees, which multiple providers confirm are charged separately.
Three drivers move a quote more than headcount does: how many states you operate in, whether statutory filing is in scope, and who owns corrections. Which brings us to the question worth asking before any price discussion.
Modelling the in-house side yourself
Nobody publishes a credible in-house figure because it is almost entirely determined by decisions specific to you. Six components, all of which you can source from your own records:
| Component | How to source it | Behaviour |
|---|---|---|
| Payroll staff cost | Fully loaded cost of the FTE fraction actually spent on payroll | Largely fixed |
| Payroll software | Annual licence, usually per employee | Variable |
| External review | CA or consultant retainer for statutory sign-off | Fixed |
| Training and rule-change absorption | Time spent staying current — unusually high in 2026 | Fixed |
| Correction cost | Corrections per run multiplied by loaded cost to investigate and fix | Semi-variable |
| Penalty exposure | Probability-weighted, not a certainty — model it as a range | Risk, not cost |
The distinction in the third column is the whole analysis. In-house payroll is mostly fixed cost; outsourced payroll is almost entirely variable. That single structural fact determines the answer, and it means the answer changes as you grow.
The crossover
Take an illustrative company: one payroll executive at a fully loaded ₹12,00,000 a year, software at ₹100 per employee per month, against a fully managed outsourced quote of ₹500 per employee per month. Every figure is an assumption; replace them.
Below the crossover, outsourcing is cheaper and the gap widens as you get smaller — at 50 people the in-house fixed cost is being spread over almost nothing. Above it, in-house is cheaper and the gap widens as you grow, because you are no longer paying a per-head margin on work whose cost barely changes.
Two cautions. The crossover moves sharply with the fraction of an FTE payroll genuinely consumes — if it is half a person rather than one, the crossover roughly halves. And it moves the other way with the number of states you operate in, because multi-state compliance raises the in-house workload and the outsourced quote simultaneously.
What does not transfer
This is the part missing from vendor comparisons, and it is not a cost item — it is the reason the cheaper option is sometimes the wrong one.
Statutory liability stays with you. You remain the employer and the deductor. A provider files on your behalf; the obligation, and the consequence of its failure, remains yours. Contractual indemnities may recover money afterwards. They do not prevent the default, the interest, or the enforcement action.
Data accuracy stays with you. Providers process what you send. Almost every payroll error originates upstream — a joiner not registered, a revision not communicated, an exit date wrong. Outsourcing moves the processing and leaves the data-quality problem exactly where it was.
Response time changes hands. An in-house team can rerun payroll on a Friday evening. A provider works to a cut-off, typically receiving your data between the 20th and 25th, and a change after that becomes an off-cycle correction with a fee attached. Companies with volatile monthly inputs feel this immediately.
Institutional knowledge leaves. After three years outsourced, nobody internally knows how your payroll works. That is fine until you want to switch, or until the provider's account manager changes.
The hybrid model
The option most often ignored: process internally, outsource statutory filing only, at a reported ₹150 to ₹250 per employee per month. It keeps control of the salary calculation and the timing, and buys specialist coverage of the part that changes most and carries the statutory consequence. For a multi-state mid-market company in 2026 — with the Labour Codes' state rules still arriving — this is frequently the best-value structure, and almost nobody's rate card leads with it.
A decision framework
| Choose in-house when | Choose outsourced when |
|---|---|
| Above roughly 200–250 employees on stable headcount | Below roughly 100 employees |
| Monthly inputs are volatile and cut-offs would hurt | Inputs are stable and predictable |
| You have or can hire genuine payroll expertise | You have no payroll expertise and cannot justify hiring it |
| Payroll data feeds other processes you control | You need coverage across many states quickly |
| Confidentiality of compensation data is a board-level concern | You are entering India as a foreign parent |
The input problem underneath both options
Neither model fixes the actual cause of most payroll pain. Both depend on a monthly handover — either to a provider or to your own payroll module — of joiners, exits, revisions, attendance and reimbursements, assembled from wherever those facts live. Where hiring, HR and payroll are separate systems, that handover is a reconciliation, and the corrections you pay for downstream were created upstream in the days before it.
Helion holds those records on one database, so the handover is a state rather than a file. That does not decide the outsourcing question for you — but it does mean the in-house option needs less of the FTE fraction that sets the crossover point, which shifts the arithmetic in the chart above.
Frequently asked questions
Is outsourcing payroll cheaper?
Below roughly 100 employees, almost always. Above roughly 250, usually not. The reason is structural rather than competitive: in-house cost is mostly fixed and outsourced cost is mostly variable, so the answer depends on the size of your denominator.
Does outsourcing transfer compliance liability?
No. You remain the employer and deductor. A provider performs the work; the statutory obligation stays with you, and an indemnity is a recovery mechanism rather than a shield.
What should I check in a quote before comparing prices?
Which of PF, ESI, PT and TDS are in scope; whether corrections to rejected returns are billable; the one-time onboarding fee; the data cut-off date; and the charge for off-cycle runs. Two quotes at the same PEPM can differ by a factor of two once these are settled.
Should a company with 400 people ever outsource?
Yes — if it operates across many states, has no payroll expertise, or is a foreign parent without an established local function. Cost is one input and usually not the deciding one at that size.
What is the most under-used option?
The hybrid: internal processing with outsourced statutory filing. It targets the part that changes most while keeping control of timing.
Outsourced PEPM ranges are compiled from published 2026 rate guidance across several Indian payroll providers; these are quoted prices rather than audited market averages, and every source is a seller. The widely repeated in-house figure of ₹870 to ₹1,800 PEPM appears in near-identical form across multiple unrelated provider sites and we have not used it. All figures in the crossover example are assumptions chosen for illustration. Helion sells payroll software and therefore has a commercial interest in the in-house side of this comparison. Statements about statutory liability are general and not legal advice — confirm your position with counsel. Current as at July 2026.