The Situation
A CPG plant was running acceptable OEE numbers but bleeding money on scrap. The problem: scrap was tracked in aggregate monthly totals, not by line, shift, or defect type. Nobody knew where the waste was actually coming from, so nobody could fix it.
What We Did
Built the defect Pareto. Installed shift-level scrap tracking by defect category. Within two weeks, the data showed that 60% of scrap came from three defect types — two related to setup errors and one to a sensor calibration drift.
Poka-yoke implementation. Designed and installed mistake-proofing at the three root causes:
- First-piece verification fixture for the setup-related defects
- Visual alignment guides at the problem stations
- Automated sensor check added to startup sequence
Operator coaching. Ran targeted training with the crews that had the highest scrap rates. Not punitive — just skill gaps that had never been addressed because nobody had the data to see them.
Results (Verified)
| Metric | Before | After | Impact |
|---|---|---|---|
| Scrap rate | Baseline | −19% | Significant yield improvement |
| Annualized savings | ~$260K | ||
| Defect categories | Untracked | Pareto by shift/line | Visibility enabled targeting |
Key Takeaway
Scrap is the quiet cost that PE firms miss during diligence. It doesn’t show up as a line item — it’s buried in COGS. Once you make it visible at the shift level, the fixes are often simple and cheap. Poka-yoke doesn’t require capital. It requires understanding what’s actually going wrong.
Company details anonymized. Engagement completed under contract leadership.