How to use the Pareto chart maker
- Choose the input type. Categories with counts expects two columns — category and count — as you would have on a check sheet summary. Raw list to tally takes one label per line (for example one line per complaint or per rejected part) and counts them for you.
- Enter the data: paste from Excel or Google Sheets, type, or upload a .csv / .xlsx file. A CSV template and sample data are provided.
- Set the options: sorting, grouping of the small categories into “Other”, the cumulative threshold (80% by default), and the titles. If a third column holds a cost or weight per occurrence you can rank by count × cost instead of by frequency.
- Read the result: the chart, the list of vital-few categories and the table. Export the chart as PNG or SVG, download the table as CSV, or print the report.
What a Pareto chart shows
A Pareto chart is a bar chart of categories sorted from the most to the least frequent, with a line showing the cumulative percentage of the total. It is one of the seven basic quality tools and the standard way to answer the question “where should we start?”. It rests on the Pareto principle, which Joseph Juran brought into quality management as “the vital few and the useful many”: in most problem data, a small number of causes account for most of the effect — roughly 80% of the defects from 20% of the causes. The 80/20 split is a rule of thumb, not a law; the chart shows what the split really is for your data.
When to use it — and which measure to rank
- To prioritise defect types, scrap reasons, downtime causes, complaint categories or audit findings before starting improvement work.
- To compare before and after an improvement, using the same categories and the same time span.
- To stratify: draw separate Pareto charts by machine, shift, supplier or product to find where a problem concentrates.
Frequency is not always the right measure. Fifty cosmetic scratches may cost less than three cracked housings. If you add a cost, time or severity weight per occurrence in the third column and choose Count × cost / weight, the tool ranks categories by total impact, which is usually what management cares about. It is good practice to look at both rankings.
How the numbers are calculated
| Quantity | Formula |
|---|---|
| Value of a category | count, or count × weight |
| Percent of total | value ÷ total × 100 |
| Cumulative percent | (sum of this and all larger categories) ÷ total × 100 |
| Vital few | the smallest set of the largest categories whose cumulative percent is at least the threshold (80%) |
The left axis runs from zero to the grand total, so the bars and the cumulative line share one scale and the line ends exactly at the top of the chart; the right axis shows the same height as 0–100%. Cumulative points are plotted at the centre of each bar. The category that crosses the 80% line is counted among the vital few, so their combined share is always at least the threshold.
Grouping the tail into “Other”
A long tail of tiny categories makes the chart hard to read. You can combine every category below a chosen percentage of the total, or everything beyond the top N, into one “Other” bar. By convention “Other” is drawn last, even when it is larger than some named bars. If “Other” becomes one of the tallest bars, the classification is too coarse: break it down again.
Reading the chart
Look at the shape of the cumulative line. A line that rises steeply and then flattens means a few categories dominate — a clear Pareto pattern, and those categories are where improvement effort pays. A line close to a straight diagonal means the problem is spread evenly across categories; in that case the classification is not separating causes, and you should stratify the data by a different factor. Remember that a Pareto chart ranks categories by how often (or how expensively) they occur; it does not show whether the process is stable over time. For that, track the defect counts on a p, np, c or u chart, and use a histogram when the data are measurements rather than categories.
Worked example (the sample data)
A die-casting line records 465 defects in twelve categories over one month. The four largest are scratches (182), porosity (121), dimension out of tolerance (64) and burrs (31).
- Scratches: 182 ÷ 465 = 39.1%
- Scratches + porosity: 303 ÷ 465 = 65.2%
- + dimension out of tolerance: 367 ÷ 465 = 78.9% — still below 80%
- + burrs: 398 ÷ 465 = 85.6% — the 80% line is crossed
So four of twelve categories (33%) cause 85.6% of the defects. Ranked by count × cost the picture changes: porosity (121 × 22 = 2,662) and dimensional defects (64 × 18 = 1,152) move ahead of scratches (182 × 4.5 = 819).
After fixing the top causes, redraw the chart with new data: the former number two becomes the next target. To express the overall defect rate as DPMO or a sigma level, use the sigma level calculator.
Frequently asked questions
What is a Pareto chart?
A Pareto chart is a bar chart in which categories are sorted from largest to smallest, combined with a line that shows the cumulative percentage of the total. It makes it easy to see which few categories account for most of the problem.
What is the 80/20 rule in a Pareto chart?
The 80/20 rule, or Pareto principle, says that roughly 80 percent of the effects come from about 20 percent of the causes. On the chart, the categories to the left of the point where the cumulative line crosses 80 percent are the vital few to work on first. The exact split varies from one data set to another.
How do I calculate the cumulative percentage?
Sort the categories from largest to smallest, divide each count by the total to get its percentage, then add the percentages one after another. The cumulative percentage of the third category, for example, is the sum of the first three percentages. The last category always reaches 100 percent.
Should the Other category go at the end?
Yes. By convention a combined Other or Miscellaneous bar is placed last, even if it is larger than some of the named categories. If Other is one of the tallest bars, split it into more specific categories.
When should I use cost instead of count?
Use cost, downtime or another weight when categories differ a lot in consequence. A defect that occurs rarely but scraps an expensive part may deserve priority over a frequent cosmetic defect. Ranking by count times cost shows total impact.
How many categories should a Pareto chart have?
Between about 5 and 10 bars is easiest to read. If you have more, group the smallest ones into Other. If you have fewer than four categories, a Pareto chart adds little over a simple table.