Payroll in Excel — until it breaks
Excel is where almost every Indian payroll starts, and it is genuinely good at the first ten employees. This page maps exactly where it stops being good, with the arithmetic of why, no scare copy required.
What Excel does well, said honestly
A competent spreadsheet runs a small, stable payroll correctly: fixed salaries, few changes, one person who understands the file. Excel is free, universal and infinitely flexible, which is why it is the incumbent in most businesses reading this page, and why 'just use software' is lazy advice. The honest claim is narrower: every capability that makes Excel powerful (free-form editing, manual formulas, copies) becomes a specific liability as headcount, churn and statutory complexity grow.
Where the break actually happens
The break is not size alone; it is events per month. Ten joiners-leavers-revisions a month means ten hand-derived pro-ratas, arrears and settlements, each a fresh chance to diverge from statute. Add contribution-period ESI, a second state's PT, an inspector's request for registers that reconcile, and the file's single-author fragility (one person, one laptop, one version among many) becomes the operational risk. Most teams recognise the moment: payday works, but only because someone heroic checks everything twice.
The honest ledger: engine vs spreadsheet
Concrete claims below are about PeopleDeck only. For the other column, we state category facts and what to verify. Products change, so check current details on their own pages before you decide.
| Dimension | PeopleDeck | Excel / spreadsheets |
|---|---|---|
| Computation | Engine applies structures, pro-ratas, arrears and statutory rates, same rules every month, workings printed on every payslip | Manual formulas per cell — correct until edited; every edge case is hand-derived, and the working lives in the author's head |
| PF, ESI, PT & TDS | Applied at current rates with effect-dated updates; contribution-period ESI and state-wise PT mapped automatically | Rates hard-coded until someone updates them; period rules and multi-state PT tracked by memory: the classic silent-error zone |
| Statutory files | ECR, challan and return files generated from the approved run, reconciling to payslips by construction | Built by hand in the portal formats each deadline, a second derivation that can disagree with the payslips it summarises |
| Versions & authority | One live system, role-scoped access, approved runs lock, every change attributed | payroll_final_v3_REAL.xlsx — copies multiply, edits are anonymous, history is whoever remembers |
| Audit & inspection | Registers, contribution histories and settlements export per period, contemporaneous | Reconstruction from files and email, and reconstructed records read as exactly that to an inspector |
| Employee experience | Payslips, balances, declarations and Form 16 data on self-service | PDF payslips if someone generates them; every question is an email to the file's owner |
| Cost | Per employee, per month, priced in rupees, visible at sign-up | Free software; the real cost is verification hours, error exposure and the single person the file depends on |
When the switch actually earns its keep
-
1
Count events, not employees
If your month has more than a handful of joins, exits, revisions or variable-pay changes, each one is a manual derivation in Excel. Event volume is the honest trigger for switching, not headcount vanity.
-
2
Price the checking, not the licence
Excel's cost is the hours spent verifying and the one person who cannot take leave in the last week of the month. Put an hourly rate on both and the comparison prices itself.
-
3
Keep Excel for what it's good at
Nobody is taking your spreadsheets — analysis, modelling and one-off workings stay in Excel. What leaves is the system of record for pay, which was never a job spreadsheets were designed to hold.
PeopleDeck in depth: the payroll engine · statutory compliance · per-employee pricing
Migrating without a bad payday
Moving off Excel is the easiest migration in payroll: the data is already yours: the master sheet imports, structures get declared once (usually their first written-down version), and policies become enforced rules.
Run one month in parallel: the system against the sheet. Expect the reconciliation to find the sheet's quiet drift: a stale rate, an ESI edge, an allowance outside the wage base. That finding, whichever way it cuts, is the decision made for you.
The boundary that decides it
PeopleDeck is a standalone HRMS any business signs up for, priced per employee, per month, with PF, ESI, professional tax and TDS applied to every payslip and upload-ready return files generated, never filed behind your back. If you are an education institution comparing ERP-bundled payroll instead, that is a different product shape: see SchoolDeck staff payroll for schools or CampusAlly payroll for colleges, licensed per institution.
Frequently asked questions
Is Excel payroll actually non-compliant?
Excel itself is neutral — plenty of compliant payrolls have run on careful spreadsheets. The compliance risk is operational: manual rate updates, hand-built files and reconstructable-only records raise the probability of the errors inspections find.
At what size should we switch?
Measure events, not heads: when monthly joins, exits, revisions and variable-pay cases stop being rare, each is a manual derivation and an error chance. For many businesses that is well under fifty employees.
What happens to our years of Excel payroll history?
It imports as opening balances and reference records (gratuity tenure, leave balances, TDS year-to-date) and the old files stay archived for the audit trail. Nothing is lost; it just stops being live.
Our accountant runs the sheet and likes it. What then?
Give them the generated files to review instead of building them; most consultants prefer auditing computed workings to deriving them. Scoped access for the accountant is a standard setup.
Does the switch cost more than our current setup?
Per-employee pricing against your verification hours and error exposure — do the multiplication for your team. The parallel month is free either way, and it produces the reconciliation that answers this question with your own data.
Retire the heroics, keep the spreadsheets.
Run a parallel month on your own data — the reconciliation decides, not the brochure.
Start free