Rework Rate

Rework rate is the metric that exposes the hidden factory. Every shop has one. The reworked parts eventually ship, the final yield numbers look fine, and the cost of the rework loop is buried in indirect labor and longer lead times. A shop with a 5 percent rework rate is paying for those parts twice: once to make them wrong and again to make them right. The customer never sees it. The operator who runs the rework cell sees it every day. The metric exists to make that work visible.

"A reworked part is a defect the shop solved instead of preventing. Solving the same defect every week is not a system."

How rework rate works

The calculation is units reworked divided by total units produced, multiplied by 100. The discipline lives in two questions: what counts as rework, and how do you capture it without disincentivizing honest reporting?

What counts as rework

A working definition for a small shop:

• Anything that required a return to a process step. A part sent back from inspection to be reworked at the operation counts.
• Anything that required an off-line correction. A part pulled aside for adjustment, deburring, touch-up, or any other corrective work.
• Anything that needed informal at-station fixing beyond setup adjustment. This is the gray zone. Operators routinely make small adjustments; the test is whether the adjustment was for a part that would otherwise have failed.

The metric collapses if operators feel punished for reporting rework. The best practice is to make the rework tag a process step, not a violation. Operators tag, the count rolls up, and the conversation focuses on root causes, not on which operator tagged the most parts. The data has to be honest before it can be useful.

Where rework rate fits on the shop floor

Imagine a 28-person fab shop welding stainless steel assemblies for a food-equipment customer. The shop reports a final yield of 98 percent and a scrap rate of 1.5 percent. Management considers quality strong. A new operations lead spends a week observing and finds something the metrics had been hiding: every shift, one welder spends roughly two hours grinding and re-welding joints that did not pass first inspection. That is informal rework, never logged.

Formal tracking starts. Every reworked part gets a small tag and an entry in a single-page log at each cell. The first month produces a startling number: rework rate is 9 percent. The labor cost of the rework loop is about 35 hours a week, or close to one full-time welder of effort, going into recovering parts that should have passed the first time.

The Pareto on the rework log shows two defect modes responsible for 75 percent of the volume: penetration on a specific joint geometry, and a minor splatter pattern from one of the older welders. A fixture redesign and a wire-feed adjustment on the older machine cuts rework rate from 9 to 3 percent over three months. The freed labor goes to a new product launch the shop had been planning to outsource. The hidden factory shrunk; the visible factory got more work done with the same crew.

Common mistakes with rework rate

• Treating rework as normal background noise. It is not background. It is the operating system absorbing waste rather than removing it.
• Counting only formal rework. Informal at-station fixes are real rework. Without capturing them, the metric understates by half or more in many shops.
• Reporting the headline without mode breakdown. A 6 percent rework rate is unactionable as a single number. The defect-mode Pareto is what shows you where to start.
• Punishing reporting. Operators who feel scolded for tagging rework will stop tagging. Bad data follows. The metric is only as honest as the recording culture.
• Treating recoverable rework as a yield-saver story. Yes, rework recovers parts. Yes, the parts ship. The cost of recovering them keeps showing up in lead time, labor, and morale. The lean answer is to remove the cause, not to celebrate the recovery.

Rework rate and related Lean tools

Rework rate is the complement of scrap rate inside the yield family. Together they account for the gap between first-pass yield and final yield. Rework is the most visible part of the hidden factory, the unmeasured loop of corrective work that lean is built to surface. Reducing rework is one of the highest-leverage quality projects in any shop and almost always pays back within a quarter.