Most small job shops know they have a scrap problem long before they know where it really comes from. Operators report bad parts, material gets written off, and quality issues show up in customer complaints or margin erosion. But when the owner or production manager asks, “What should we fix first?” the answer is often guesswork. For a complete overview, see our manufacturing execution system software guide.
That is where a scrap Pareto becomes useful. Instead of treating all scrap as one bucket, you sort scrap dollars by the few dimensions that matter most: part number, operation, and defect code. In a typical shop, a small number of combinations drive a large share of total scrap cost. If you can identify those combinations using work-order data and actual material cost, you can focus corrective action where it will pay back fastest.
For small manufacturers, this does not require a complicated BI project. A basic spreadsheet or a simple manufacturing system can do the job if your data is structured correctly. The goal is not perfect accounting on day one. The goal is to make scrap visible enough that supervisors, quality, and ownership can stop chasing symptoms and start eliminating the top losses first.
What a scrap Pareto should show
A Pareto analysis ranks losses from largest to smallest so you can see the “vital few” causes ahead of the “trivial many.” The method is widely associated with quality improvement and the Pareto chart. In a job shop, the practical version is simple: list scrap events, attach a dollar value, group them, sort them, and review the biggest contributors first.
A useful scrap Pareto should answer three questions:
- Which parts are driving the most scrap dollars?
- At which operations does scrap happen most expensively?
- Which defect codes explain the loss?
Those three views matter because each points to a different type of action. Scrap by part may reveal chronic quoting or process issues. Scrap by operation may point to a machine, setup method, fixture, or training problem. Scrap by defect code shows what kind of failure is happening, such as out-of-tolerance dimensions, wrong material, burrs, finish defects, or damage during handling.
When you combine all three, you stop having vague conversations like “turning has too much scrap” and start having specific ones like “Part 4821 at Op 30 is generating most of its scrap dollars from oversize bores on second shift.” That level of detail is where improvement starts.
The minimum data you need
Small shops often overcomplicate this step. You do not need every cost category to begin. You need a reliable, repeatable record of each scrap event tied to the work order.
Core fields to capture
- Work order number
- Part number
- Operation number or work center
- Quantity scrapped
- Defect code
- Date or shift
- Actual material cost per piece or per unit of measure
If you can also capture operator, machine, supplier lot, and whether the issue was detected in-process or at final inspection, that adds useful context. But do not wait for perfect data design before starting.
Why actual material cost matters
Many shops track scrap only in pieces. That is better than nothing, but it often points attention in the wrong place. Ten scrapped blanks on a low-cost part may matter far less than one scrapped near-finished component made from expensive stock.
At minimum, assign each scrap event a material cost based on what was actually consumed. If your process adds significant outside processing or internal value before the part is scrapped, you can later expand the model to include conversion cost by operation. But for most small shops, starting with actual material cost creates a much better ranking than raw quantity alone.
If you want a quick way to estimate the dollar impact before building your full report, use a simple tool like FactoryOS’s scrap cost calculator.
How to build a simple scrap Pareto from work-order data
You can do this in a spreadsheet, an ERP export, or an MES-style system. The mechanics are straightforward.
Step 1: Log every scrap event at the point of occurrence
Do not rely on end-of-week memory. The operator, lead, or inspector should record scrap when it happens, against the active work order and operation. A delayed entry usually loses the true defect code or the actual point of failure.
Keep defect codes short and standardized. For example:
- DIM-OVR — oversize dimension
- DIM-UND — undersize dimension
- BURR — burr or edge condition
- FINISH — surface finish defect
- MATL — wrong or defective material
- SETUP — setup-related error
- DAMAGE — handling damage
A short controlled list beats free-text descriptions every time. If your rework process is muddy, it also helps to separate true scrap from recoverable nonconformance. This is closely related to rework tracking, which we cover in this guide to manufacturing rework tracking for small job shops.
Step 2: Calculate scrap cost for each row
Create a simple formula:
Scrap Cost = Quantity Scrapped × Actual Material Cost per Piece
Example only:
- Work order 10457
- Part A-221
- Op 20
- 3 pieces scrapped
- Material cost per piece: $42
- Scrap cost: 3 × $42 = $126
This is enough to build the first Pareto. Later, you may choose to add operation-added value if scrap occurs late in the routing and the material-only view understates the loss.
Step 3: Build three grouped summaries
Summarize scrap cost by:
- Part number
- Operation number or work center
- Defect code
Then sort each summary from highest scrap cost to lowest. Add a cumulative percentage column so you can see which few contributors make up most of the total.
Even without a chart, a ranked table is enough to drive action.
Step 4: Drill into combinations, not just totals
The single most useful follow-up is a second-level view such as:
- Part by operation
- Part by defect code
- Operation by defect code
This prevents misleading conclusions. For example, Op 40 might rank high overall only because one family of parts is unstable there. The fix is not “improve Op 40” in general. The fix is targeted process control for that specific part-operation pair.
An example scrap Pareto table for a small shop
Below is a simplified example showing how a scrap-by-part summary might look for one month. Numbers are illustrative.
| Rank | Part Number | Scrap Cost | % of Total | Cumulative % |
|---|---|---|---|---|
| 1 | A-221 | $2,940 | 29% | 29% |
| 2 | B-144 | $1,980 | 20% | 49% |
| 3 | C-778 | $1,360 | 13% | 62% |
| 4 | A-305 | $1,040 | 10% | 72% |
| 5 | D-090 | $820 | 8% | 80% |
| 6 | All others | $2,060 | 20% | 100% |
This tells a shop owner something important immediately: the top five parts account for 80% of scrap dollars. That means the first review meeting should not cover 47 different parts. It should focus on those five.
Now imagine that Part A-221 is then broken down by operation:
| Part Number | Operation | Defect Code | Scrap Cost |
|---|---|---|---|
| A-221 | Op 20 | SETUP | $1,260 |
| A-221 | Op 30 | DIM-OVR | $980 |
| A-221 | Op 50 | DAMAGE | $700 |
Now the action list becomes practical. Instead of launching a broad “quality improvement” effort, the team can review setup sheets at Op 20, inspect tool offset control at Op 30, and look at handling and packaging between Op 50 and final inspection.
How to choose the right defect codes
Defect codes should help you act, not just categorize. A poor code list creates vague reporting. A good list separates problems that have different root causes and different owners.
Keep the list short but meaningful
For most job shops, 8 to 15 defect codes is enough to start. If you have 40 codes, operators will choose inconsistently. If you have 3 codes, the data will be too broad to guide action.
Separate cause from symptom when possible
Some shops use only symptom codes like “oversize” or “scratch.” Others use cause codes like “setup” or “tool wear.” In practice, many need both. One practical approach is:
- Primary defect code for what failed
- Optional cause code for why it likely failed
If your team is not ready for two fields, start with one standardized defect code field and use review meetings to assign likely causes afterward.
Align codes with your corrective action process
If a defect code cannot trigger a clear owner, it is probably not useful. For example:
- SETUP may belong to manufacturing engineering or the area lead
- MATL may belong to purchasing, receiving, or supplier quality
- DAMAGE may belong to material handling or packaging
This alignment is consistent with broader quality management principles published by organizations such as NIST Baldrige, where measurement is useful only if it supports process improvement and accountability.
Common mistakes that make scrap Pareto reports misleading
Counting pieces instead of dollars
High-volume low-cost scrap can drown out expensive low-volume losses. Always review dollars first, then quantity as a secondary lens.
Mixing scrap and rework together
A part that can be recovered is not the same as a part that must be thrown away. Track them separately. Otherwise, your scrap numbers will be overstated and your corrective actions will blur together.
Using defect descriptions with no standard codes
Free text creates reporting chaos. “Burr,” “burrs,” “edge burr,” and “sharp edge” will scatter the same issue across multiple categories.
Not tying scrap to the operation where it occurred
If everything gets booked at final inspection, upstream causes stay hidden. The defect may be found at final, but the operation field should reflect where it was created whenever possible.
Ignoring schedule and workload context
Scrap often rises when jobs are rushed, setups are compressed, or WIP piles up between operations. If your top scrap areas also struggle with lateness or congestion, review them together with schedule adherence and WIP controls. Related reading: planned vs. actual start and finish times by work order and WIP inventory control for small manufacturers.
How to run a weekly scrap review that leads to action
A scrap Pareto is only useful if it changes behavior. The best cadence for many small shops is a short weekly review using the prior week and month-to-date totals.
Who should be in the meeting
- Owner or operations manager
- Production supervisor or area lead
- Quality representative
- Manufacturing engineer or process owner if available
What to review
- Top 5 parts by scrap cost
- Top 5 operations by scrap cost
- Top 5 defect codes by scrap cost
- Any large single scrap event from the week
- Status of prior corrective actions
What good action items look like
Avoid vague actions like “watch quality more closely.” Better examples:
- Revise setup sheet for Part A-221 Op 20 and require first-piece signoff
- Replace worn soft jaws on Lathe 3 for family B parts
- Change incoming inspection plan for supplier lot on material code 6061-PL-2
- Add protective tray between Op 50 and final wash to prevent handling damage
Each action should have an owner and due date. On the next review, check whether scrap dollars actually dropped. If not, the action was not enough or the suspected cause was wrong.
How this connects to profit, quoting, and capacity
Scrap is not only a quality issue. It consumes material, machine time, labor, inspection effort, and schedule capacity. In a small job shop, recurring scrap on one high-value part can quietly erase the margin on an otherwise good customer account.
Once your Pareto is working, use it in three places:
Quoting review
If specific parts repeatedly generate scrap, your quoted margin may be unrealistic. You may need to re-price, revise the process, or challenge the print tolerance with the customer.
Capacity planning
Scrap creates hidden load because replacement parts must be remade. If a work center is constantly remaking bad parts, its apparent utilization may look busy while productive output remains weak. That is why scrap analysis pairs well with runtime and performance analysis, such as machine utilization by shift.
Continuous improvement priorities
Small shops have limited engineering bandwidth. A scrap Pareto helps you spend that bandwidth on the few issues most worth solving, instead of the loudest complaint from the floor that day.
Start simple, then add detail as your process matures
The first version of your scrap Pareto does not need to be elegant. If you can capture work order, part, operation, defect code, quantity, and actual material cost consistently, you already have enough to uncover expensive patterns that are probably invisible today.
After a few months, you can improve the model by adding:
- Shift or crew comparison
- Machine-level trends
- Supplier lot traceability
- Late-stage value-added cost by operation
- Links between scrap and schedule misses, shortages, or rework
But do not wait on those enhancements before acting. The biggest win usually comes from basic visibility and disciplined follow-through.
If your shop is still managing scrap from paper notes and memory, now is the right time to put a simple system in place. FactoryOS helps small manufacturers track work orders, shop-floor activity, and loss data in one place so teams can see where margin is leaking and respond faster.
Start with the top scrap dollars, not the longest complaint list. Build the Pareto, review it every week, and assign real corrective actions to the parts and operations hurting you most. When the data gets specific, improvement gets faster.
Ready to make scrap visible and fix the biggest losses first? Start your free FactoryOS trial.
