← Back to Field Notes

September 15, 2025 · The Bloomfield Team

10 Simple Ways to Reduce Scrap and Rework

A 70-person contract manufacturer tracked their scrap and rework costs for a full quarter. The number came to 6.8% of revenue, roughly $680,000 annualized. When they broke down the root causes, 41% of the rework events traced to information errors: wrong revision on the floor, incorrect material pulled from stock, setup notes missing from the traveler, tolerance callouts transcribed incorrectly from the drawing. The machines ran fine. The information feeding the machines was the problem.

Here are ten changes that reduce scrap and rework. Most cost nothing to implement. All of them address information flow rather than machine capability.

1. Verify Drawing Revisions at the Point of Setup

Revision control failures cause more rework than tool wear. A single outdated drawing running on the floor can scrap an entire batch. The fix is simple: require the setup operator to confirm the revision level against the current customer record before starting the job. A 30-second check that prevents a $5,000 scrap event.

2. Standardize Job Travelers

If every job traveler looks different depending on who created it, the operators filling in the data will fill it in differently too. Standardize the format. Every traveler should include the same fields in the same order: material, revision, operation sequence, critical dimensions, inspection requirements, and any special notes from previous runs. The instructions that people actually follow are the ones that look the same every time.

3. Tag Jobs With Historical Quality Notes

If a part caused a quality issue on a previous run, that information should appear on the traveler for the next run. "Previous run: 0.0003" oversize on bore diameter, adjusted offset by -0.0002 on second setup." That note takes 15 seconds to write and saves two hours of rework on the repeat order. Your ERP data holds most of these notes already, buried in NCR records and job cost entries. Making them visible at the point of work is the leverage.

4. First-Piece Inspection on Every Setup

This sounds basic because it is. Shops that enforce first-piece inspection on every setup see scrap rates 30 to 40% lower than shops that skip it on repeat jobs. The assumption that a repeat job will run the same as last time is the assumption that creates the most expensive scrap events, because an entire run completes before anyone discovers the problem.

5. Track Scrap by Cause Category

Most shops track total scrap cost. Fewer track scrap by root cause category. Without cause data, you are managing an aggregate number that tells you the problem is big but not where to fix it. Break scrap into five categories: material defect, machine error, setup error, information error, and operator error. When you see that 40% of scrap comes from information errors, the improvement strategy changes entirely.

6. Keep Setup Sheets Attached to the Job

A setup machinist who ran a job eight months ago recorded the offsets, the fixture location, the tool stick-out lengths, and the cycle time on a setup sheet. That sheet is in a binder on a shelf. The machinist running the same job today does not know it exists, so they start from scratch and spend 90 minutes figuring out what the first machinist already documented.

Attach setup sheets to the job record. When the job comes back, the setup data comes with it. This is one of the simplest paths to cutting rework in any shop.

7. Build Material Verification Into the Process

Wrong material pulled from stock causes the most expensive scrap events because the parts often pass initial inspection and the error surfaces at the customer or during final testing. A material verification step at raw stock issue, with the operator confirming the heat number and material cert against the job requirements, eliminates this category of scrap almost entirely.

8. Use In-Process Gauging at Critical Operations

Relying on final inspection to catch problems means the rework happens after every operation is complete. Moving inspection to the critical operations, the ones where tolerance is tightest or the cost of error is highest, catches problems before downstream operations add more labor and material to a part that is already out of spec.

9. Document and Share Lessons From Every NCR

Non-conformance reports get filed. They rarely get shared with the people who need them. When an NCR reveals that a specific feature on a specific material tends to drift on a specific machine, that finding should flow into the setup notes for every future job with similar characteristics. The complaint-to-improvement loop only works if the improvement reaches the floor.

10. Review Scrap Data Monthly With the Floor Team

The people closest to the problem have the best ideas for fixing it. A 30-minute monthly review where the quality manager presents scrap data by category and the floor team discusses root causes and potential fixes produces more actionable improvement than any top-down quality initiative. The operators see things the data misses. The data shows patterns the operators cannot see from one machine. Together, they cover the full picture.

The Underlying Pattern

Seven of these ten items address information flow. The right drawing reaching the right machine at the right time. Historical quality data appearing where it can prevent a repeat failure. Setup knowledge transferring from one run to the next. The machines in most shops are capable of holding tolerance. The information systems feeding those machines are where scrap and rework originate, and that is where the highest-leverage improvements live.