Who this is for: Manufacturers investigating recurring losses of quality, time and productive capacity.
Scrap and rework consume resources in different ways. Discarding a part creates a loss; recovering it can still require extra time, inspection and another trip through production. Looking only at the amount of material thrown away can leave a substantial part of the operational problem unexplained.
The analysis should connect the defect to where it originates, when it is detected and the activities required to deal with it. This makes it possible to compare changes to design, process, equipment, information or inspection without selecting a solution solely because it addresses the most visible symptom.
Agree definitions before comparing numbers
Agree how to classify scrap, rework, sorting and material awaiting a decision. If one department records a part as scrap while another recovers similar parts without a record, the monthly comparison becomes unreliable. Keep the period and calculation basis consistent: piece counts, weight and financial value answer different questions.
For a reasonably consistent product family, a scrap-piece rate can be expressed as scrapped quantity divided by produced quantity within the defined scope. That percentage does not explain cause or cost. Retain the product, batch and stage information so that an aggregate does not hide a defect concentrated in a particular configuration or operating condition. Document what the denominator includes.
Build the cost picture without double counting
Separate incurred cost from estimated consequences. If the part’s standard cost already includes an operation, do not add that operation again without a clear accounting reason. Agree the valuation method with the people responsible for the business’s cost information, and state where a practical estimate is being used instead of a recorded expense.
The table helps define the scope of the assessment; not every category applies to every case. In particular, capacity consumed by rework is not automatically a lost sale. Demand, the production constraint and opportunities to recover the time must be assessed before translating occupied capacity into a commercial loss. Keep these assumptions visible in the comparison.
Scroll horizontally to see every column.
| Cost area | Information to collect |
|---|---|
| Embedded material and operations | Affected quantity and valuation method |
| Part recovery | Additional hours, consumables and operations |
| Inspection and sorting | Time spent and quantities examined |
| Handling and administration | Additional transfers and resources involved |
| Capacity or delivery effects | Specific evidence, separate from assumptions |
Look for patterns before assigning a cause
Group observations by defect type, product and stage while retaining date, batch and known operating conditions. A concentration guides investigation but does not by itself establish causation. A defect detected at final inspection may originate much earlier, and the last operator involved may simply be the person who discovered it.
Discuss the hypotheses with design, production and quality staff. Priorities may depend on frequency, cost, customer impact and the practicality of intervention. NIST’s process-improvement resources provide general methodological context, but the specific action should follow evidence from the actual process. Results from other manufacturers cannot be treated as evidence of what this business will achieve.
References: NIST — Lean and Process Improvement
Check that the improvement has not moved the loss elsewhere
Define measures before the intervention. Afterwards, compare suitable periods and products, noting changes in mix, materials and inspection practices. Fewer recorded scrapped parts could result from a classification change or additional rework. Read quantities, recovery activity and cost together so that the conclusion reflects the whole loss rather than one favourable metric.
Useful outputs include a loss register, examined hypotheses, selected actions and verification criteria. Where the solution needs design changes, automation or software, translate those needs into specific requirements. Measure the financial result in the actual business context. A percentage reduction cannot be promised before the conditions and causes are understood, and an early improvement needs a suitable observation period before being treated as sustained.
Prepare for the first review
- Choose a product family and a defined analysis period.
- Gather produced, scrapped and reworked quantities using shared definitions.
- List defect types and detection stages with batch references.
- Record time and material used in recovery, inspection and sorting.
- Describe changes already attempted and their observed outcomes.
Common questions
Where do we start if scrap records are unreliable?
Start with a limited scope and a few categories that operators can apply consistently. Reliable observation of one product or stage can be more useful than a large, inconsistent archive. The collection can become more structured as its purpose is clarified.
Will adding inspection reduce scrap?
Inspection may detect a defect earlier or prevent it from progressing, but it does not automatically remove the cause. Assess its cost and position alongside changes to the process that creates the defect.
Sources and further reading
Apply this to your project.
Tell us which defects recur and how they are handled today. We can connect production information and costs to define an improvement that can be assessed against evidence.
Discuss your production process