Piece-rate costing: why your machine-hour rate is lying to you

The mismatch nobody notices

Almost every ERP on the market assumes labour is paid by time. You define a work centre, give it an hourly rate, put a standard time on each operation, and the system multiplies. That model is not wrong — it is just built for factories where it is true.

In a large share of Indian engineering shops, it is not true. The drilling operator is paid per hole. The deburring is paid per piece. The polishing line is paid per kilogram. The operator's earnings for the day are a function of output, not of elapsed time — and the shop knows this perfectly well, because it is exactly how the wage slip is calculated.

The mismatch survives because two different people own the two numbers. Whoever runs payroll knows the piece rate. Whoever prepares quotes uses a rate per hour that was arrived at years ago, adjusted upward occasionally by feel. Nobody sits both numbers next to each other, because nothing in the process forces them to.

A worked example in rupees

Take a flange that needs 24 holes drilled. The operator is paid ₹4.50 per hole. Actual labour cost per piece is therefore straightforward:

24 holes × ₹4.50 = ₹108.00 per piece

Now cost the same operation the way the ERP does it. The drilling work centre carries a rate of ₹420 per hour. The standard time on the operation says 12 minutes:

12 min ÷ 60 × ₹420 = ₹84.00 per piece

A gap of ₹24 per piece — the hourly model is understating the real cost by roughly 22 per cent. On a batch of 500 flanges that is ₹12,000 of labour cost that exists on the wage slip but not in the quote. Repeat that across a year of orders and you have a plant that is busy, well utilised, and quietly less profitable than its own reports claim.

Which direction the error runs

The obvious worry is under-costing — quoting too low and eroding margin. That is real, but it is the recoverable failure. You lose money, you notice eventually, you correct.

The more damaging case is over-costing. Suppose the standard time was set generously, or the operator has become faster than the standard assumes. The hourly model now says the operation costs ₹140 when it truly costs ₹108. You quote high. You lose the order to a competitor. And crucially, you never find out why — a lost tender produces no data. You simply conclude the market is difficult, and carry on with the same inflated number.

Under-costing produces a symptom. Over-costing produces silence. That asymmetry is why this is worth fixing even in shops that are currently profitable.

Four ways to cost an operation

The fix is not to replace hourly costing with piece-rate costing. Both are correct, for different operations, and a single routing usually contains both. What you need is the ability to say per operation how that operation is paid.

In practice four modes cover almost everything in a discrete manufacturing shop:

1. Rate × quantity

A rate per unit of output, multiplied by actual output. ₹4.50 per hole, ₹12 per piece, ₹80 per kilogram. This is the mode most ERPs lack and most Indian shops need.

2. Calculated hourly

Setup time plus run time per unit, multiplied by a work-centre or machine hourly rate. Correct for genuinely time-based operations — a long CNC cycle where the machine, not the operator, determines the cost.

3. Fixed rate

A flat charge for the operation regardless of quantity or duration. Final inspection, documentation, a one-off fixture setup. Trying to express these as a rate per hour produces a fictional number with false precision.

4. Subcontracted

A per-unit rate paid to an outside job worker. Distinct from the others because it is a purchase, not internal labour — which matters enormously for overhead, as below.

Where overhead belongs (and where it does not)

Overhead is where the second common error lives. Most shops apply a blanket percentage to total cost. If that percentage is meant to recover factory overhead — rent, power, supervision, depreciation — then applying it to subcontracted work is straightforwardly wrong.

When a job worker plates your components at ₹180 a unit, your factory consumed no power, no floor space and no supervision for that step. Adding 45 per cent overhead to it invents ₹81 of cost that does not exist, and makes outsourcing look more expensive than it is. Shops make genuinely bad make-versus-buy decisions on the back of this.

Overhead should apply to the in-house cost base and exclude subcontract charges:

Apply the 45 per cent to everything instead, and the same part costs ₹1,012.10 — an invented ₹81 that will lose you work you should have won.

Fixing it without boiling the ocean

You do not need to re-cost your entire item master to benefit from this. The distribution is almost always heavily skewed: a small number of operations account for most of your labour spend. Start there.

1. List the operations you actually pay per piece. Payroll knows this. It is usually a shorter list than people expect — five to fifteen operations in a typical shop.
2. Get the current piece rates. Not last year's; what was paid last month.
3. Compare against what your costing assumes for those same operations. Do this for your ten highest-volume parts first.
4. Look at the direction of each gap. Parts where you are over-costed are your immediate commercial opportunity — you can quote more sharply tomorrow without losing a rupee of margin.
5. Fix overhead on subcontracted steps, which is usually a single configuration change and often the largest single correction.

Whether you do this in a spreadsheet or in a system is secondary. What matters is that the rate used to cost an operation is the rate you actually pay for it, and that it stays current when the piece rate changes.

Where software helps is durability. A spreadsheet reconciliation is accurate on the day it is done and decays from then on. In NextGenManager the costing mode is a property of the operation itself, so the rate is applied every time that operation runs, on every part that uses it — and updating a piece rate updates every quote that depends on it.

---

Related reading

• GST job work under Section 143: the return deadlines explained
• How to choose an ERP for an Indian manufacturing MSME
• The four costing modes in NextGenManager