Cost per minute, not cost per hour

The number that is not the wage

Everyone in the trade knows the wage in their own origin and roughly believes they know it in three others. Almost nobody can state their cost per standard minute, and it is the number that decides the comparison.

It is an ordinary division:

cost per standard minute = total cost of running the line for an hour
                           ------------------------------------------
                           standard minutes the line produced in that hour

and the minutes produced are

standard minutes produced = operators x 60 x efficiency

The wage sits in the numerator. Efficiency sits in the denominator. A factory that cuts the numerator and damages the denominator has raised its cost. That single sentence explains most of what looks mysterious about origin competitiveness, and the rest of this lesson is the arithmetic behind it.

The four lines, measured

Each of the four factories from lesson 1 ran the same exercise on the line that would take ALD-2210. The figures are each factory's own, for a single working hour on a single line.

Two readings fall straight out of that table and both are counter-intuitive.

Wages are about half the cost of a minute, even in the dearest factory. Direct wages are 50.1% of Halbrent's hour, 49.3% of Tolvane's, 50.7% of Draymill's and 39.4% of Sennary's. The rest is people who are not sewing and things that are not people: machines, buildings, power, financing, management. A low-wage origin does not get those cheaper in proportion — a needle machine, a fusing press and a kilowatt hour are priced much closer to a world price than an operator hour is.

Overhead per minute is nearly identical at the cheapest and the dearest. Sennary carries USD 0.0422 of overhead in every standard minute and Draymill carries USD 0.0441 — Draymill's is 4.6% dearer — while Draymill's operators cost 2.63 times Sennary's. The overhead is spread over minutes, and Sennary produces so few minutes an hour that its cheap building becomes an expensive minute.

From a minute to a garment

The minute is not what the buyer buys. The garment is, and the bridge is the garment's own standard minute value at that factory.

sewing cost per garment = cost per standard minute x standard minutes for this garment

The four do not agree on the minutes either, because a standard minute value is a fact about a method, a machine layout and a trained operator — not a fact about a garment.

Here is the sentence this lesson exists for.

Sennary pays the lowest wage in the comparison — 61.9% below Draymill's — and produces the second-dearest sewn jacket. Draymill's operators cost 2.63 times Sennary's, and Draymill's sewing cost per jacket is only 8.0% higher. Sennary's minute is 28.3% cheaper than Draymill's and it needs 46.5 of them against Draymill's 36.0.

That is not an accident of the numbers chosen. It is the mechanism. An origin's wage advantage arrives divided by its efficiency and multiplied by its method, and both of those are things the factory built rather than things the origin gave it.

The mistake nobody made

Two years ago Halbrent opened a second unit. Every decision in the chain was correct and defensible.

The main plant was full and turning work away. The second site is where labour is available and the wage is 26.8% lower — USD 1.50 an hour against USD 2.05. Halbrent moved the same machines, the same style range, two of its own supervisors and its own quality standard. The board approved it on a costing that used the main plant's efficiency, because the machines and the method were identical, and nobody could name a reason they would not be.

In month three the new unit ran at 41% efficiency.

The new unit's minute costs 45.4% more than the main plant's. Nothing went wrong. A new unit has no line balance history, no trained mechanics on those machines, no operators who have made this style four hundred times, and an absence rate that has not settled. Efficiency is accumulated, and it was not in the costing because it does not appear on a machine list.

The honest half of the story is that this is temporary, and the useful question is how temporary. Hold the new unit's hourly cost still and solve for the efficiency at which its minute matches the main plant's:

minutes needed  = 129.24 / 0.1004 = 1,287.6
efficiency then = 1,287.6 / (36 x 60) = 59.6%

59.6%. That is the number the board should have been given, and it is a completely different conversation from the one it had. It converts an open-ended hope into a target with a date against it, and it prices the gap in the meantime: every month the unit spends below 59.6% costs the difference between two known cost-per-minute figures on a known volume.

What to measure, this month

You can build your own cost per standard minute from figures your accounts department already produces, and it will take an afternoon.

1. Take one line and one month. Add every cost attributable to it: direct wages and benefits; the helpers, supervisors, mechanics and quality staff who serve it; and the share of factory overhead your accounts already allocate to it.
2. Take the standard minutes that line actually produced in the month — pieces despatched, multiplied by their standard minute values. Despatched, not cut.
3. Divide. That is your cost per standard minute, and it is the honest one, because it carries your real efficiency and your real absence rate rather than a target.
4. Do it again for each of the last twelve months. The trend matters more than the number, and the twelve points tell you whether the thing you did in March worked.