Where the money leaks
Most factories can tell you what shipped. Far fewer can tell you, on the day it
happened, which line ran below target and why, how much fabric a style actually
consumed against what was booked, or which operator a recurring defect traces
back to. Those three gaps are where margin quietly disappears, and they are all
data-capture problems before they are analysis problems.
Capture at the point the work happens
Hourly output entered by the line supervisor on a tablet at the end of the shed.
Defects logged at the quality table against the operator and the operation.
Fabric issued and returned recorded at the store window. None of it takes more
than a few taps, because anything that takes longer will not be done
consistently, and inconsistent data is worse than none.
What the buyer pack needs
| Requirement |
What the system must hold |
| Traceability |
Order to style to line to shipment, with dates |
| Working hours |
Attendance and overtime reconciled with payroll |
| Subcontracting |
Declared units, what was sent, what returned |
| Capacity |
Booked versus actual, by line and month |
| Quality |
Defect rates by operation, with corrective action |
Rolling out without stopping production
Go-live happens unit by unit and line by line, starting with one shed for a full
production week while the old paper process runs alongside. Supervisors are
trained on their own line during their own shift. Nothing is switched off until
the numbers from both methods agree.