Manufacturing Scrap Rate Benchmarks 2026

Manufacturing Scrap and Rework Cost Benchmarks

The table below is meant for owners, plant managers, quality managers, and operators who need a fast sanity check. It combines published cost of poor quality ranges with quality-performance gaps reported by McKinsey. Use it as a benchmark screen, not as an excuse to ignore your own process data.

The spread matters. McKinsey found that a best-performing medical device producer had waste and rework costs only one-fifth of the median producer, while the top pharmaceutical producer's costs were one-fourteenth of the median level. That is not a small quality tweak. That is a structural cost advantage.

Scrap Rate, Rework Rate, First Pass Yield, and COPQ Formulas

If your team only reports total rejects, you are missing the story. Scrap rate, rework rate, first pass yield, OEE quality, and cost of poor quality answer different questions. A plant can have low scrap and still burn margin through rework. Another can have decent first pass yield and still suffer warranty claims because inspection catches the wrong failures.

ASQ groups quality cost into prevention, appraisal, internal failure, and external failure categories. Scrap and rework usually sit inside internal failure cost, but a bad part that reaches the customer moves into external failure cost. That shift is where quality problems get expensive fast.

EPA's lean waste guidance lists defects as scrap, rework, replacement production, and inspection. That definition is useful because it keeps teams from treating inspection as free. Inspection is necessary in many processes, but inspection does not make the first bad part profitable.

OEE Quality Example: Turning Rejects Into a Scrap Rate

OEE is useful because it forces scrap and rework into the same conversation as uptime and speed. OEE.com defines quality as good count divided by total count, with good parts passing through the process the first time without needing rework. Here is the math using the published OEE.com example.

The example quality rate is 97.80%, which sounds strong until you translate it into money. If those 423 rejected units are low-cost widgets, the scrap may be tolerable. If they are machined castings, molded medical parts, electronics assemblies, or fabricated stainless components, that same 2.20% reject rate can wreck the job.

The second lesson is more uncomfortable. OEE.com's two-week example shows OEE rising from 85.1% to 85.7% even as quality falls from 99.5% to 95.0%. If the production meeting only celebrates the OEE gain, the team misses the defect problem hiding underneath.

When OEE Goes Up But Scrap Gets Worse

This is one of the more common traps in manufacturing metrics. A line runs longer, the throughput board looks better, and the top-line OEE number improves. Meanwhile, the quality department sees more nonconforming material reports, more reinspection, more retesting, and more rework queues.

The fix is simple, but not always easy. Put quality rate, scrap cost, rework hours, and first pass yield on the same board as OEE. Do not let a longer run time hide a weaker process.

Why Scrap and Rework Hurt More in 2026

A bad part is always waste. It hurts more when materials cost more, labor cost rises, and capacity is already tight. The public indicators below are not scrap rate benchmarks by themselves, but they explain why quality loss has become a board-level issue for more manufacturers.

The Federal Reserve reported May 2026 manufacturing capacity utilization at 75.7%, 2.5 percentage points below the 1972 to 2025 average. Even when national utilization looks moderate, individual plants can be capacity constrained. Rework consumes the same skilled people and machines that would otherwise make sellable parts.

BLS reported producer prices for goods rose 2.5% in 2025. When material inputs rise, scrap reduction moves from a quality initiative to a gross margin initiative. The plant may not need a new sales target first. It may need to stop remaking the same order twice.

How to Use These Benchmarks Without Fooling Yourself

Scrap data gets political. Operators may think the number will be used against them. Quality may count defects one way. Finance may count cost another way. Production may exclude startup loss because it has always been part of the process. That is how a plant ends up with five different versions of the truth.

Start with a definition that everyone can live with. Then keep it stable long enough to learn something. If you change the definition every month, the trend is useless.

NIST's engineering statistics handbook is a good reminder that process data needs discipline. Control charts are built to distinguish routine variation from special-cause variation. That matters because not every scrap spike deserves the same fix. A one-time supplier lot issue, a worn tool, a bad setup sheet, and an operator training gap are different problems.

What to Include in a Scrap and Rework Dashboard

A useful dashboard should make the next action obvious. If it only reports the monthly scrap dollar total, the team will argue after the money is already gone. A better version shows where the loss came from, what changed, and which defect families are worth attacking first.

Do not bury rework in overhead. Rework should have a clock, a route, a cost, and a reason code. If the part comes back to the same station three times, that is not labor efficiency. It is a process asking for help.