Payroll software sold in India is built for an office, and the market gives it every reason to be. The majority of Indian companies buying HR software have salaried employees who work fixed hours, take leave from a calendar, and produce a payroll input that barely changes month to month.
A factory does not work like that. The input changes every month, and it changes for reasons the payroll system has to understand, not just absorb. Somebody worked a double shift on the twelfth. Two hundred contract workers came through a contractor whose invoice you cannot pay until you know their attendance matched their wages. A state revised its minimum wages in April and the revised rate is higher than what you are paying an unskilled worker on the second line. Overtime was worked, and the rate it has to be paid at is not the rate anybody in the plant assumes it is.
This piece is about what a payroll system has to do to survive that, and how to tell during a sales conversation whether it does. It is written for the person who owns payroll at a manufacturing company: a plant HR head, a group HR manager with three sites in two states, or a finance controller who has worked out that the payroll spreadsheet is now the largest operational risk on their desk.
It is not a features list. Every vendor selling into manufacturing will tell you they handle shifts, overtime and contract labour. The question is how, and the difference sits in the arithmetic, not the interface.
Positions are current at September 2026 and reflect the four Labour Codes as brought into force on 21 November 2025, the Employees' Provident Funds Scheme, 2026 and its companion schemes effective 29 June 2026, and the Income-tax Act, 2025 with the Income-tax Rules, 2026 governing salary paid from 1 April 2026. State minimum wages and labour welfare fund rates change independently of all of that. Nothing here is legal advice, and nothing here describes any particular vendor's product. Check every rate against the notification that applies to your establishment.
What Makes Factory Payroll Different
Four things, and they compound instead of simply adding up. Attendance decides what a large part of the workforce is paid, where an office keeps it as a compliance artefact. The wage base is not one number, because a plant pays monthly salaried, monthly rated, daily rated, sometimes piece-rated and contract populations side by side. The statutory surface is wider, adding state minimum wages, labour welfare fund, bonus, statutory overtime and prescribed registers on top of the four heads an office deals with. And the difficulty scales with the number of distinct rules rather than with headcount.
The first is that attendance is an input rather than a record. In an office, attendance is largely a compliance artefact: you keep it, you rarely compute from it. In a factory, attendance decides what a substantial share of your workforce is paid. Every hour of clock data is money, and a capture failure at a turnstile on the last day of the month is a payroll error, not an inconvenience. That is why our attendance and payroll sit on the same record rather than on an integration, and why we say so in almost every manufacturing conversation.
The second is that the wage base is not one number. A plant typically pays a monthly salaried population, a monthly rated worker population, a daily rated population, sometimes a piece-rated population, and contract labour through one or more contractors. Each has a different basis for statutory computation and a different treatment when a month is short. A system that models "employee has a monthly CTC" and derives everything from it will do the first population correctly and fudge the rest.
The third is that the statutory surface is wider. On top of provident fund, employees' state insurance, professional tax and tax deducted at source, a factory deals with state minimum wages, labour welfare fund, bonus under the Payment of Bonus Act, overtime at a statutory multiple, and registers that have to be producible on demand. Multiply that by the number of states you operate in.
The fourth is scale in the wrong direction. Manufacturing payroll is not hard because there are many employees. It is hard because a five-hundred-person plant can generate more distinct payroll rules than a five-thousand-person software company. Rules, not headcount, are what break systems.
The Wage Base Problem, and Why It Is Worse in a Factory
Since the Code on Wages, 2019 came into force with the rest of the Labour Codes on 21 November 2025, the definition of wages in section 2(y) has done something specific: where the allowances excluded from the definition exceed one-half of the total remuneration, the excess is added back into wages. Provident fund, gratuity, bonus and overtime all compute off that base.
Every Indian employer is affected by this. Factories are affected worse, and the reason is structural.
Office salary structures are usually built out of a small number of large components: basic, house rent allowance, a special allowance that absorbs the remainder. Factory wage structures accumulate. Over years of settlements, a plant ends up with a basic, a dearness allowance, an attendance incentive, a production incentive, a shift allowance, a canteen allowance, a washing allowance, a conveyance figure, and two or three heads that exist because a specific union agreement created them in a specific year. Each one has to be classified as inside or outside the wage definition, and the one-half test then has to be applied to whatever is left outside.
This is not a computation a system does once at implementation. It has to be applied per employee, per month, because the components vary per employee and per month. An employee who worked heavy overtime in a month has a different total remuneration, which can move them across the one-half boundary in one direction and back in the next.
Ask any vendor to show you this on a real employee of yours with a heavy allowance structure, in a month with variable pay, and to show you the same employee in a month without it. If the wage base is identical in both months, the system is not applying the test. The detailed mechanics are in our piece on how payroll is calculated in India, and the component-by-component view is in the salary breakup explainer.
The consequence of getting this wrong is not a rounding difference. It is an under-contribution to provident fund across a whole workforce, discovered at inspection, payable with interest and damages, for every month since the Codes came into force.
Overtime: The Calculation Most Systems Get Structurally Wrong
Section 14 of the Code on Wages requires overtime to be paid at not less than twice the ordinary rate of wages. That much most people know. Two things follow from it that most systems handle badly.
| What has to be decided | The requirement | Where systems go wrong |
|---|---|---|
| The ordinary rate of wages | Derived from wages as defined in section 2(y), so the same add-back applies | Computed on basic alone, understated the same way every month for the same people |
| The divisor | A stated basis for turning a monthly wage into an hourly rate | The choice between 26 days and calendar days, and between eight hours and the actual shift length, is not exposed |
| Overtime on a weekly off | Treated differently from a working day in many establishments | One rate applied to every hour of overtime |
| Overtime on a national or festival holiday | Treated differently again, and the holidays are state-specific | A single national holiday list applied to every plant |
| Compensatory off granted instead of overtime pay | Its own rules and its own expiry | Comp-off left on the books indefinitely as a liability nobody has quantified |
| Piece rate and production incentive | The minimum wage floor still applies underneath, and the incentive has to be classified for provident fund | Mapped wrongly at implementation because it looks like a bonus and behaves like an allowance |
The first is what "ordinary rate of wages" means. It is a rate derived from wages as defined in section 2(y), which means the same add-back from section 2 applies here. An overtime rate computed on basic alone is understated wherever the one-half test bites. This is the single most common structural error we see in factory payroll, and it is systematic rather than occasional: it is wrong the same way every month, for the same people, which is exactly the pattern an inspection finds.
The second is the divisor. Converting a monthly wage into an hourly rate requires a decision about how many hours are in a month, and that decision is not arbitrary. Whether you divide by 26 days or by the calendar days, and whether the daily figure is then divided by eight hours or by the actual shift length, produces materially different hourly rates. Systems differ here and most do not expose the choice. Ask to see the divisor, in writing, and ask whether it changes for a daily rated worker.
Then there are the cases underneath. Overtime on a weekly off is not the same as overtime on a working day in many establishments. Overtime on a national or festival holiday is treated differently again, and the holidays themselves are state-specific. Compensatory off granted instead of overtime pay has its own rules and its own expiry, and a system that lets a comp-off sit on the books indefinitely is accruing a liability nobody has quantified. Our overtime piece works through the arithmetic.
Piece rate and production incentive sit alongside all of this and are handled worse than anything else in factory payroll. A piece-rated worker's earnings vary with output, which means the wage base for statutory computation varies with output too, and the minimum wage floor still applies underneath: earnings below the applicable floor for the period have to be topped up regardless of what the pieces came to. Production incentives paid on a plant or line basis raise a different question, which is whether the incentive falls inside the definition of wages at all. That is a classification decision with consequences for provident fund, and it is one of the components that most often gets mapped wrongly at implementation because it looks like a bonus and behaves like an allowance. Whichever way it is classified, the classification has to be consistent, documented and applied by the system rather than remembered by a person.
A practical test for a demo: give the vendor an employee who worked ten hours of overtime, four of them on a weekly off, in a month where the employee also had two days of unpaid absence. Ask for the overtime amount and the working. Most systems will produce a number. Getting a working out of them that you can check is much harder.
Contract Labour and the Principal Employer's Exposure
This is the part of factory payroll that sits outside most payroll software entirely, and it is the part with the sharpest downside. The workers are engaged through a contractor and never appear on your payroll, but where the contractor fails to pay wages or to make statutory contributions for people deployed at your premises, the liability comes back to the principal employer. What follows is what a system has to hold for those workers, and what the monthly reconciliation has to compare.
Contract labour is engaged through a contractor, paid by the contractor, and does not appear on your payroll. What does appear is a contractor invoice. The exposure is that the principal employer is not insulated from the contractor's obligations: where a contractor fails to pay wages or fails to make statutory contributions for workers deployed at your premises, the liability comes back to the principal employer. Under the Occupational Safety, Health and Working Conditions Code, 2020, which subsumes the Contract Labour (Regulation and Abolition) Act, 1970 along with the Factories Act, 1948, that structure is preserved.
What this means operationally is that you cannot treat a contractor invoice as an accounts payable document. You have to be able to reconcile it: these workers, at this site, on these days, at these rates, with these contributions actually remitted against these numbers. The distinction between on-roll and off-roll payroll is a legal one, not an operational one, and the operational work does not go away.
Most payroll systems handle this by ignoring it, which pushes the reconciliation into a spreadsheet maintained by whoever at the plant is least able to refuse. What a system built for manufacturing should do instead is capture contract worker attendance in the same place as everybody else's, hold each contractor's rate card, produce the reconciliation against the invoice, and hold the contractor's compliance documents against a date so that an expired licence or a missing challan is visible before the invoice is passed rather than after an inspection.
Ask the vendor directly: does contract labour attendance come into the same system, and can you produce a wage reconciliation against a contractor invoice. The answer is frequently no. That is manageable if you know it going in, and expensive if you find out in month three.
There is a threshold question underneath this that is worth knowing. Contractor licensing obligations under the OSH Code attach at establishments engaging fifty or more contract workers, which is higher than the old threshold. A plant sitting just under that line should still do the reconciliation, because the wage liability does not depend on the licensing threshold.
Multi-State, Multi-Plant: Minimum Wages and Labour Welfare Fund
Two plants in two states are not one payroll run twice. They are two payrolls that happen to share a chart of accounts, and almost everything below that line differs.
Minimum wages are fixed by state, by scheduled employment, by skill category and often by geographic zone within the state. They are revised on variable dearness allowance, typically twice a year, and the revision is retrospective often enough that you should assume it will be. A payroll system that holds a minimum wage figure at all has to hold it per state, per category, with an effective date, and has to be able to tell you which of your employees are now below it.
That last capability is the one that matters and the one most systems do not have. Nobody deliberately pays below minimum wage. It happens because a rate was correct when it was set and a notification moved underneath it. The control is a report that compares every employee's applicable wage against the current floor for their state and category, run before the payroll rather than after it. If a vendor cannot show you that report, they do not handle minimum wages, whatever the brochure says. Our state minimum wage pages publish revision dates and flag where a revision is contested, which is the same discipline applied to published data.
Labour welfare fund is smaller in rupee terms and equally capable of generating an inspection finding. It applies in some states and not others, at different rates, with different contribution periods, some half-yearly rather than monthly. The employee who transfers from a state where it applies to one where it does not, mid-period, is the case that breaks configurations. Our labour welfare fund pages set out the state-by-state position.
Professional tax has the same shape, with an additional wrinkle: the levy itself is not permanent. Odisha repealed professional tax entirely with effect from 1 April 2026. A system still deducting it for an Odisha employee is deducting something that no longer exists, and unlike an underpayment, the employee notices this one immediately.
Inter-state migrant workers are the group most often missed in a multi-plant setup. Where workers are recruited in one state and deployed in another, obligations attach that do not attach to local hires, and those obligations survived the consolidation into the Occupational Safety, Health and Working Conditions Code, 2020. The payroll consequence is that home state and work state are two different fields with two different uses, and a system that stores one address per employee cannot tell you which population you are looking at. Ask how the system distinguishes them, and ask before you need the answer for a return.
Then there are holidays. National and festival holidays are declared by state, and a plant that runs on a holiday pays differently for it. If your two plants are in different states, they have different holiday calendars, and the payroll system needs both.
Shifts, Muster Rolls and the Attendance Input
Everything in sections 2 to 5 depends on attendance being right, which makes attendance capture a payroll question and not an HR convenience. Three things decide whether it is right in a factory: shift patterns that rotate and cross midnight, a muster roll that can be produced in the prescribed form on demand, and a capture method that survives a turnstile failure on the last day of the month. Each of them changes what payroll computes, not just what HR reports.
Shift handling in a factory means rotating patterns, night shifts that cross midnight, shift allowances that attach to a specific shift rather than to a person, and shift changes agreed on the floor at short notice. The midnight crossing is a better test of a system than it sounds: ask which day a shift starting at 22:00 on the thirty-first belongs to, for attendance, for overtime and for the wage period. Systems answer that question differently and some answer it inconsistently between the three.
Capture has to survive the environment. Fingerprint readers struggle with hands that have been working, which is why face recognition and card-based capture are more common on shop floors than in offices. Biometric devices at plant gates need to keep working when the network does not, and the sync afterwards has to land without creating duplicate punches. A worker who punches at a gate device in the morning and on a mobile at a warehouse in the afternoon should produce one attendance record, not two. Our attendance system is built around that assumption, and the buyer's guide to biometric attendance covers what to look for in the hardware.
Deductions are the other thing a factory payroll does that an office payroll rarely does. Canteen recoveries, uniform or shoe recoveries, transport, loans and advances, damage or loss recoveries, and union subscriptions all appear on shop floor payslips, and they are not unlimited. The Payment of Wages Act framework, now carried into the Code on Wages, caps total deductions as a proportion of wages for the wage period and restricts what may be deducted at all. A system that applies a deduction because somebody entered it, without checking it against the cap or the permitted categories, will eventually produce a payslip that is indefensible. Ask what the system does when the deductions entered for an employee exceed the permitted proportion: whether it stops, warns, or carries the excess forward, and whether the carry-forward is visible.
The muster roll is the compliance output of all this. Registers of wages, attendance and overtime have to be maintained and producible, and the Codes have consolidated the forms rather than removed the obligation. A system that computes payroll beautifully but cannot produce a register in the prescribed form has moved your problem rather than solved it. Our piece on attendance under the Labour Codes sets out what has to be kept.
One more thing that only shows up in factories: attendance regularisation at volume. On a shop floor, missed punches are routine, not exceptional, and the correction workflow has to be usable by a supervisor on a phone at the end of a shift, with a record of who approved what. If regularisation runs through HR by email, HR becomes the bottleneck on payroll closing, every month, permanently.
What the Labour Codes Changed for Factories
The four Codes came into force on 21 November 2025 and consolidated twenty-nine central labour laws. For a factory the practical effects fall into four groups.
The wage definition, covered in section 2, is the largest. It moves the base for provident fund, gratuity, bonus and overtime at once, and it moves it by an amount that depends on your own allowance structure. Gratuity in particular is worth recomputing: the fifteen days of wages for every completed year is now fifteen days of a base that may be higher than the one you have been provisioning against. Our gratuity piece works through the current definition.
The consolidation of establishment law. The Occupational Safety, Health and Working Conditions Code, 2020 subsumes the Factories Act, 1948, the Contract Labour (Regulation and Abolition) Act, 1970 and the Inter-State Migrant Workmen Act among others. Registrations, licences and registers that were separate are being consolidated, and the transition is where administrative errors happen. If your plant holds a licence issued under a repealed act, confirm what it converts to and when.
The social security architecture. The Employees' Provident Funds Scheme, 2026 and its companion schemes took effect on 29 June 2026, superseding the 1952, 1995 and 1976 schemes. The provident fund wage ceiling of fifteen thousand rupees was re-notified on 29 May 2026 under Chapter III of the Code on Social Security, 2020. The ceiling figure did not change. Its legal basis did, and a system whose configuration still cites the 1952 scheme has not been revisited.
The tax forms. From 1 April 2026, under the Income-tax Rules, 2026, the quarterly salary statement moved from Form 24Q to Form 138 and the salary tax certificate from Form 16 to Form 130. This one is trivially checkable: look at what your last quarterly statement actually went out on.
Buying: What to Ask a Vendor Who Says They Do Manufacturing
Every vendor selling to you will say they handle manufacturing. Six demonstrations separate the ones that do, and each is something to watch on a screen instead of something to be told: the wage base moving between two months, the overtime working including the divisor, a contract labour reconciliation against a contractor invoice, the report of employees below the applicable state minimum wage, a night shift that crosses month end, and the wage and attendance registers in the prescribed form. A vendor who can do four of the six may still be the right purchase, provided you know which two are missing before you sign.
Show me the wage base on my own employee, in two months, one with heavy overtime. If the base does not move, the section 2(y) test is not being applied.
Show me the overtime working, including the divisor. Ask for it in writing. Ask what changes for a daily rated worker.
Show me a contract labour reconciliation against a contractor invoice. Not a report of contract workers. A reconciliation.
Show me the report that lists every employee currently below the applicable state minimum wage. Run before payroll, not after.
Show me a night shift that crosses month end, and tell me which wage period it lands in for attendance, for overtime and for the shift allowance.
Show me the wage register and the attendance register in the prescribed form, generated from the system, not exported to Excel and formatted by hand.
If a vendor can do all six on a demo with your data, they have built for factories. If they can do four, they have built for offices and sold into factories, which is a different thing and occasionally still the right purchase if the other two are handled elsewhere. Our comparison of payroll platforms covers who is scoped for what, and our manufacturing HRMS page sets out how we handle the six.
| What it is | Office payroll | Factory payroll |
|---|---|---|
| Attendance | A record kept for compliance | An input that decides pay for most of the workforce |
| Wage structure | A few large components, uniform across the population | Many components accumulated through settlements, varying by population |
| Populations | Monthly salaried | Monthly salaried, monthly rated, daily rated, piece rated, contract |
| Overtime | Occasional, often unpaid or absorbed | Routine, statutory multiple, computed on a defined wage base |
| Statutory heads | PF, ESI, PT, TDS | The same, plus minimum wages, LWF, bonus computation, overtime, registers |
| State variation | Usually one state | Often several, each with its own wages, LWF and holidays |
| Third parties | None | Contractors, whose wage liability can return to you |
| Registers | Rarely inspected | Producible on demand in the prescribed form |
The Implementation, and Where It Stalls
Manufacturing payroll implementations do not fail. They stall, in the same three places, and knowing where means you can put time against them at the start rather than discovering them in week six.
The first is the component mapping. Somebody has to decide, for each of your wage components, whether it sits inside or outside the definition of wages, and that person has to be senior enough for the decision to stand. This is a two-day exercise that gets scheduled as a two-hour call, and it is the single most common cause of an implementation running long. Do it first, do it with whoever owns the union agreements, and write the reasoning down rather than just the answer.
The second is historical data. Opening balances for provident fund and tax, year-to-date figures, leave and comp-off balances, and enough salary history to produce a correct annual tax certificate. For a plant with a long-serving workforce, the leave and comp-off balances are usually the mess: they exist in a register, they have never been reconciled, and the migration is the first time anybody has added them up. Expect the total to surprise finance.
The third is the attendance devices. Existing readers have to be integrated or replaced, network at plant gates is frequently worse than anybody in the head office believes, and the offline behaviour has to be tested rather than assumed. Test it by unplugging the network, not by asking.
One thing to decide before any of that: whether you are moving one plant or all of them. Group HR usually wants a single rollout and plant HR usually wants to go last. The compromise that works is to take the most complicated plant first, not the easiest one. A pilot at the simplest site proves nothing, because the configuration that survives a straightforward single-state salaried population is not the configuration that will survive contract labour, three shifts and a state revision. Whatever breaks is going to break somewhere, and it is cheaper to find it while the implementation team is still assembled.
Run a parallel month before you switch. Take a month you have already paid and closed, load it with the same inputs, and compare employee by employee: gross, each statutory deduction, net, employer contributions, overtime. Then look only at the mismatches. Pay particular attention to anybody with unpaid days, because the treatment of non-contributory days in the provident fund return is where systems that look identical stop being identical, and to anybody near the employees' state insurance wage limit, because crossing it mid-year has consequences the new system may handle differently from the one that recorded it.
Switch in April if you can. The tax year starts there and no opening balances have to be carried.
Questions People Ask
What is the best payroll software for a manufacturing company in India?
The honest answer is that the category is narrower than the marketing suggests, and the test is not features but the six demonstrations in section 8. What separates products here is whether contract labour, shift-based wage bases, state minimum wage monitoring and prescribed registers are in the product or in a spreadsheet beside it. Platforms genuinely built for Indian manufacturing are a smaller set than the platforms sold into it. We are one of them, which is why this article exists, and the comparison of the wider category linked in section 8 sets out who is scoped for what without pretending we are the only answer.
How should payroll software handle contract labour?
At minimum it should capture contract worker attendance in the same system as everyone else, hold each contractor's rate card, and produce a wage reconciliation you can put against the contractor's invoice before you pay it. This is about exposure, not efficiency: where a contractor fails to pay wages or make statutory contributions for workers at your premises, the liability returns to the principal employer, and that structure is preserved under the Occupational Safety, Health and Working Conditions Code, 2020, which subsumed the Contract Labour Act. Holding contractor licences and challans against expiry dates in the same place is the second half of the control, because an expired document discovered at inspection is worse than one discovered before an invoice is passed.
Does payroll software calculate overtime correctly under the Labour Codes?
Some do. Section 14 of the Code on Wages requires not less than twice the ordinary rate of wages, and the failure is almost never the multiple, which every system knows. It is the base and the divisor. The ordinary rate derives from wages as defined in section 2(y), so the one-half add-back applies to it, and a system computing overtime on basic alone will understate it systematically. The divisor question, whether a monthly wage converts to an hourly rate through 26 days or calendar days and through eight hours or the actual shift length, is rarely exposed in the interface at all. Ask for both in writing during the demo and check them against one of your own employees.
How does the new wage definition affect factory payroll specifically?
More than it affects office payroll, because factory wage structures have more components. Section 2(y) of the Code on Wages adds back into wages the amount by which excluded allowances exceed one-half of total remuneration, and a plant that has accumulated a dozen allowance heads through successive settlements will usually find that test biting. Because it applies per employee per month, an employee with heavy overtime in one month can fall on a different side of the line than in the next. The downstream effects reach provident fund, gratuity, bonus and overtime at once, which is why the component mapping exercise in section 9 is worth doing properly rather than quickly.
Can one payroll system handle plants in different states?
It can, but ask specifically how rather than whether. State variation shows up in minimum wages by category and zone, labour welfare fund applicability and rate, professional tax slabs, holiday calendars, and the registers each state expects. The capability that matters most is not storing those rates, it is the report that tells you which employees are now below the applicable floor after a revision, run before the payroll and not after. The employee who transfers between states mid-period is the case worth testing on the demo, because that is where configurations break.
Do we need attendance and payroll from the same vendor?
In a factory, effectively yes. Attendance decides pay for most of the workforce, so a separate attendance system means a monthly reconciliation between two sets of numbers before every payroll run, performed under time pressure by whoever is available. That recurring cost is almost always larger than the licence difference between buying together and buying separately. In an office population where attendance is a record rather than an input, buying separately is defensible. On a shop floor it is a decision to do a manual reconciliation forever.
What registers does a factory have to produce, and should software generate them?
Registers of wages, attendance, overtime and, where contract labour is engaged, the associated records, have to be maintained and produced on demand. The Codes consolidated the forms; they did not remove the obligation. Software should generate them in the prescribed form directly, because a register assembled in Excel from an export is correct on the day it is built and drifts quietly from then on. On a demo, ask to see the register generated, not exported, and check the form against what your state expects.
How long does a manufacturing payroll implementation take?
Four to six weeks for a single plant in one state with clean data. Two to three months where there are multiple states, multiple entities, or contract labour to bring in. The variable is almost never the software. It is the component mapping decision described in section 9, the state of your leave and comp-off registers, and how quickly the attendance devices at your gates can be integrated and tested. Implementations stall in the middle, not at the start, and the two-day component mapping exercise is where to spend the time you would otherwise lose.
Is minimum wage compliance really a software problem?
The payment is a management decision, but the detection is a software problem, and detection is where companies fail. Nobody sets out to pay below the floor. It happens because a wage was correct when it was set and a notification moved underneath it, usually retrospectively, in one of the states you operate in. Without a report that compares each employee's applicable wage against the current floor for their state and category, you will find out when somebody complains or when an inspector asks. With one, you find out before the payroll runs. That is the entire difference and it is worth asking for by name.
Where This Leaves You
Do these in order.
Map your wage components against the definition in section 2(y) first, before you look at software at all. You will need the answer for any system you buy, the exercise takes two days, and doing it early turns a stalled implementation into a short one. Get it signed off by whoever owns the union agreements.
Then count your populations. Monthly salaried, monthly rated, daily rated, piece rated, contract. Write down how many of each and which states they sit in. That list is your requirement, and it will eliminate half of any shortlist on its own.
Then take the six demonstrations in section 8 to each vendor, with your own data. Not a demo employee. One of yours, in a month with overtime and unpaid days.
Then run a parallel month before you sign anything you cannot exit, and look only at the mismatches.
If you want to run that against us, the manufacturing HRMS page sets out how we handle the six, payroll and attendance share one record instead of an integration, and the pricing is published rather than quoted. Bring an employee with a messy allowance structure and a month with a night shift crossing month end. That is the demo worth having.
Related reading: how to calculate payroll in India, overtime pay calculation, attendance compliance under the Labour Codes, and the payroll compliance checklist.
Sources
- Code on Wages, 2019, enacted text on India Code. Section 2(y) on the definition of wages and the one-half proportion rule, and section 14 on overtime at not less than twice the ordinary rate of wages.
- Occupational Safety, Health and Working Conditions Code, 2020, enacted text on India Code. Consolidation of the Factories Act, 1948, the Contract Labour (Regulation and Abolition) Act, 1970 and the Inter-State Migrant Workmen Act, and the provisions on contract labour and principal employer liability.
- Ministry of Labour and Employment, labour.gov.in. Commencement of the four Labour Codes on 21 November 2025, the Employees' Provident Funds Scheme, 2026 and its companion schemes effective 29 June 2026, and the ₹15,000 provident fund wage ceiling notified on 29 May 2026 under Chapter III of the Code on Social Security, 2020.
- Income Tax Department, incometax.gov.in. The Income-tax Rules, 2026 renumbering the quarterly salary statement and the salary tax certificate with effect from 1 April 2026.
- Employees' State Insurance Corporation, esic.gov.in. Contribution rates and the wage limit for coverage.
- Payment of Bonus Act, 1965, enacted text on India Code. Eligibility, the minimum and maximum percentages, and the computation base.
- Finance Department, Government of Odisha. Repeal of professional tax with effect from 1 April 2026.
- State labour department notifications for minimum wages and labour welfare fund. Rates and revision dates vary by state and are published on our own minimum wages and labour welfare fund pages with the notification each is drawn from.
Nothing above describes any particular vendor's product. The six demonstrations in section 8 are written to be asked of every vendor equally, including us.
