How to Reduce Scrap and Rework: A Practical Guide

Q: What is the real cost of scrap and rework?

Far more than the material. Scrap carries the full cost of everything already spent on the part (material, machine time, labour and overhead), plus the lost capacity to make a good one, plus the downstream cost of sorting, rework and sometimes a lost customer. Rework hides the same losses as extra hours that never appear as scrap. The cost of poor quality (COPQ) is the honest total, and it is usually several times the scrap-bin value.

Q: How do I reduce scrap without slowing the line?

By building quality in rather than inspecting it out. Find where defects are actually made (usually startup, changeover or a few unstable process conditions), fix those root causes, and mistake-proof so the defect cannot recur. Stabilising the process reduces scrap and improves flow at the same time, whereas end-of-line sorting adds cost without fixing anything.

Q: Where do most defects come from?

On most lines a few sources dominate: startup and first-part rejects, changeover rejects while the process settles, and running defects from an unstable condition (a drifting setting, worn tooling, inconsistent material). Rank defects by cause and location with real data, then attack the top few. Chasing every defect equally, or only sorting at the end, is why scrap programs stall.

Key takeaways

Scrap costs far more than material: it carries all the work already put into the part plus the lost capacity to make a good one. Price the full cost of poor quality, not the bin.
Most defects come from a few sources, usually startup, changeover and a handful of unstable conditions. Rank by cause and attack the top few.
Build quality in, do not inspect it out. Fixing root causes and mistake-proofing beats end-of-line sorting, which adds cost without fixing anything.
Scrap is the Quality factor of OEE. Cutting it raises OEE directly and frees capacity you were spending to make rejects.

Scrap and rework are the Quality factor of OEE, the third multiplier after availability and performance. Only good, first-pass units count, so every reject hurts twice: it consumes the capacity that made it and it fails the quality gate. Reducing scrap follows the same loop as the rest of OEE improvement: see the true cost, find where defects are actually made, remove the top causes, and prove the rate fell. This guide walks that loop for quality specifically.

Step 1: Price the true cost of poor quality

The scrap bin badly understates the loss. A scrapped part carries the full cost of everything already spent on it (material, machine time, labour and overhead) plus the lost capacity to make a good one instead, plus downstream sorting and, at worst, a lost customer. Rework hides the same losses as extra hours that never show up as scrap at all. That total is the cost of poor quality (COPQ), and it is usually several times the material value. Start by putting a real number on it with the scrap rate and COPQ calculator, and get your baseline first-pass yield so you can tell whether changes actually help.

Step 2: Find where defects are actually made

Defects are not evenly spread; a few sources produce most of them. Collect defect data by type and by where in the process it occurred, then build a Pareto. Almost always the top of the list is dominated by startup and first-part rejects, changeover rejects while the process settles, and running defects from an unstable condition such as a drifting setting, worn tooling or inconsistent incoming material. Attack the top two or three. Sorting good from bad at the end of the line feels like action but changes nothing upstream, which is why quality programs that live at final inspection never move the number.

Step 3: Attack the top defect modes at the source

For the causes at the top of the Pareto, fix the process, not the symptom.

Find the true cause: run root cause analysis (5 Whys, Fishbone) on the recurring defects instead of reworking them again. A defect that returns every run is a root cause that was never removed.
Mistake-proof it (poka-yoke): design the operation so the defect cannot be made, or is caught the instant it is, rather than relying on attention and inspection.
Stabilise the process: control the variation behind the defect, whether that is a setting, tooling wear or material consistency. Statistical process control helps you see drift before it becomes scrap.
Fix startup and changeover rejects: validate the first part properly and tighten setup so the process reaches good output faster, which also links straight to changeover reduction.

Step 4: Size the prize

Because scrap consumes capacity you already pay for, the recovered money is large and fast. The scrap and COPQ calculator turns your reject rate into an annual figure, and the OEE improvement ROI calculator shows what the OEE points you recover are worth per year. A costed target moves the conversation from "we have a scrap problem" to a funded case with a number on it.

Put numbers on it

Price your scrap and cost of poor quality with the free calculator.

Cost your scrap

Step 5: Build quality in, and keep it there

Scrap creeps back when the conditions that cause it drift back: tooling wears, a setting slips, material changes. Permanence comes from building quality into the process (mistake-proofing and source control) and keeping the defect rate visible with a metric the team owns and a fast loop from a defect to a fix. Much scrap is also tied to the same instability that causes stops: startup rejects after a breakdown, drift during a chronic minor-stop condition, changeover rejects while the line settles.

That overlap is why the partner we recommend, Fabrico, helps here too. By reading every stop from the machine's PLC signals and showing the true cause on video, it makes the process instability behind many defects visible, and closes the loop to an auto-routed work order so the root condition gets fixed rather than reworked around. It is EU-built with EU data residency and holds ISO 27001, 20000-1 and 9001 (which supports audit-readiness). The calculators and guides here stay free either way; Fabrico is what we point to when a team wants the process visibility that keeps both stops and the defects they cause down. Book a Fabrico demo to see it on your lines.

FAQ

What is the real cost of scrap and rework?

Far more than material. Scrap carries everything already spent on the part plus the lost capacity to make a good one, plus downstream sorting and sometimes a lost customer. Rework hides the same losses as extra hours. The cost of poor quality is the honest total, usually several times the bin value.

How do I reduce scrap without slowing the line?

Build quality in rather than inspecting it out. Find where defects are actually made (usually startup, changeover or unstable conditions), fix those root causes and mistake-proof them. Stabilising the process cuts scrap and improves flow at once; end-of-line sorting adds cost without fixing anything.

Where do most defects come from?

A few sources dominate: startup and first-part rejects, changeover rejects while the process settles, and running defects from a drifting setting, worn tooling or inconsistent material. Rank by cause and location with real data, then attack the top few rather than every defect equally.

How does scrap relate to OEE?

Scrap and rework are the Quality factor of OEE. Only good first-pass units count, so every reject drags OEE down twice: it consumes capacity and fails the quality gate. Cutting scrap raises Quality directly and frees the capacity spent making parts you throw away.