Try it while you read
Open the free Pareto Chart tool and click Load sample: it reproduces the case study below.
Theory: the vital few and the useful many
The Pareto principle, named after the economist Vilfredo Pareto and brought into quality management by Joseph Juran, says that a small number of causes produce most of the effect. In manufacturing, typically three or four defect types make up 80 % of the scrap. Working on those first gives the largest improvement for the effort.
A Pareto chart sorts the categories by frequency from left to right and adds a cumulative percentage line. The categories needed to reach the cut-off (usually 80 %) are the vital few.
Count or cost?
The most frequent defect is not always the most expensive one. A scratch reworked in one minute costs less than a wrong dimension that scraps a finished part. When a cost per occurrence is known, ranking by count × cost often changes the priority.
Drill-down
A first-level Pareto is rarely specific enough to act on. Clicking the top bar shows its subcategories (for example scratches by location or by machine), which narrows the problem before the root cause analysis.
Key point
Step by step
- Define the categories clearly so every defect fits exactly one. Keep “Other” small.
- Collect the data from scrap reports, inspection or complaints.
- Enter category and count in the table; the same category on several rows is added up. Add subcategory and cost per item if known.
- Read the chart: the vital few are highlighted.
- Switch to cost ranking if costs are entered.
- Drill down by clicking a bar.
- Compare with data after the improvement to show the result.
Case study: scrap on a machining line
Six months of scrap records of a shaft line are grouped into 8 categories, 125 scrapped parts in total (the Load sample data).
| Category | Count | % | Cumulative % |
|---|---|---|---|
| Scratches | 48 | 38.4 | 38.4 |
| Dents | 31 | 24.8 | 63.2 |
| Wrong dimension | 17 | 13.6 | 76.8 |
| Burrs | 12 | 9.6 | 86.4 |
| Missing hole | 7 | 5.6 | 92.0 |
| Paint runs, wrong label, other | 10 | 8.0 | 100 |
4 of 8 categories make up 86.4 % of the scrap: scratches, dents, wrong dimension and burrs. By count, scratches are the clear first priority. Ranked by cost, wrong dimension moves up because those parts are scrapped after all operations. The team runs two parallel actions: separators in the transport boxes against scratches and dents (cheap, quick) and an SPC chart with tool probing against wrong dimensions (high cost per part).
Result
Common mistakes
- Categories that overlap or are too broad (“surface defects”).
- A large “Other” bar that hides real problems.
- Ignoring cost: the most frequent problem is not always the most important.
- Stopping at the chart without a cause analysis.
Frequently asked questions
Is it always 80/20?
No. 80/20 is a rule of thumb that says a few categories usually cause most of the problem. Sometimes it is 70/30 or 90/10. If all bars are about the same height, the categories are probably too broad or too narrow.
What period of data should I use?
Long enough to include normal variation in product mix and shifts, typically 1 to 6 months. Too short a period can make a one-off event look like the main problem.