Margin on one line and product fell two points, and the yield and defects behind it moved in the same week
Contribution margin on Line 50 running SKU-4417 fell to in the middle week of March. Yield on the same line and product fell to in that week and the defect rate rose to . Process temperature at the line's heating stage ran at , and that reading sits above the upper limit this stage is held to. A week earlier the same product was already carrying it across every line. Yield was at , temperature at and margin at . Across the month the book finished at .
So what? A margin number tells the commercial team something is wrong and nothing about where. The same week's yield, defects and temperature on the same line and product name the stage. That turns a month-end variance into a maintenance job someone can be sent to do.
21.44 vs 23.43%
two points under its own
90.12 vs 95.49%
the same line, product and week
4.98 vs 0.89%
rejected at inspection, not lost to speed
168.4 vs 162.0 °C
above the limit the stage is held to
94.54 vs 95.98%
already slipping one level up
23.62 vs 23.33%
the month shows none of it
One line, one product, one week, read through four systems
Each reading against the level that line and product hold together.
| Reading | The middle week, against its own level | What it contributes |
|---|---|---|
| Process temperature · Line 50 and SKU-4417 | 168.4 °C against its own 162.0 | the parameter end, above the limit the stage holds |
| Defect rate · Line 50 and SKU-4417 | 4.98% against its own 0.89% | units made and then rejected |
| Yield % · Line 50 and SKU-4417 | 90.12% against its own 95.49% | what reached good stock |
| CM % · Line 50 and SKU-4417 | 21.44% against its own 23.43% | what it was worth, per unit sold |
| Process temperature · SKU-4417 across every line | 165.2 °C against its own 162.0 | already running hot one level up |
| Yield % · SKU-4417 across every line | 94.54% against its own 95.98% | already slipping one level up |
| What is left | the heating stage on Line 50 | a setpoint, a sensor, or the controller behind it |
On Line 50 running SKU-4417 margin came in at , yield at and the defect rate at , all in the same week. Process temperature at the heating stage ran at , and that reading sits above the upper limit this stage is held to. Four readings, one line, one product, one week. That makes this a chain to troubleshoot rather than four separate numbers.
Process temperature week by week · every line and product
The book-wide reading, against the level it holds. It drifts a fortnight before anything is joined to a line.
The book-wide reading was already at a fortnight before anything was joined to a line, and at the week after. Supervisory monitoring could have caught the drift there. What it could not say is which line and which product to go to, and that is the whole difference the join makes.
CM % · the book against the pair underneath it
The same week, read at two levels, each against the level it holds.
Every line and product
Line 50 · SKU-4417
Across the month the book finished at , above the level it was expected to hold, while the pair underneath it read . The same product across every line sat at . A reading that never dips is the one that cannot find a pocket two levels down.
21.44 vs 23.43%
on everything that line ran
90.12 vs 95.49%
the units that did not come good
4.98 vs 0.89%
and these were made before being rejected
Four readings agree on one line and one product
Margin came in at and yield at . The defect rate reached and process temperature ran at , all in the same week.
Why it matters: One reading out of band is a number to watch. Four moving together on the same slice is a stage to go and look at.
The defects say it is quality, not throughput
The defect rate rose to in the same week that yield fell to , which is four of the five points the line lost.
Why it matters: Yield can fall because the line ran slow or because what it made was rejected. The defect rate carrying most of it says which, and they have different owners.
The process moved before the margin did
Yield on the product across every line was already at and the stage at before the line-and-product margin fell to .
Why it matters: It makes this a process problem with a commercial consequence, which is a different piece of work from a commercial problem with a process excuse.
What was the setpoint at that stage through the week?
Separates a changed setpoint from a failing sensor or controller.
Which shift on Line 50 carried the worst of the temperature?
Turns a line and a week into something that can be scheduled against.
What kind of defect were the rejected units failing on?
Says whether the heat explains the failure mode or only accompanies it.
Did the margin drag come through price, mix or conversion cost?
Says how much of the two points the yield loss actually explains.
Which line and product pairs are closest to their own limit now?
The same read, run before the next chain forms.