What it is
A Pareto chart sorts categories — causes of complaints, types of defect, sources of cost — from the largest to the smallest and adds a line showing the running total as a percentage. Where the line climbs steeply, a few categories account for most of the effect.
Its right-hand axis is a percentage of the same total as the bars, which makes this the one case where Chart Buffet draws a second axis.
✓ Use it when
- You are prioritising: which few problems should be fixed first?
- Your data is counts or costs per category, all positive.
✗ Avoid it when
- The categories have a natural order of their own (months, sizes).
- There are only two or three categories.
Tips for a better chart
- Group the long tail into “Other” and keep it at the end even if it is large.
- Mark the 80% level, or note how many categories it takes to reach it.
- Use counts or costs, not percentages, for the bars.
The maths
With values sorted so that v1 ≥ v2 ≥ … ≥ vn, the cumulative line is Ck = 100 · Σi≤k vi / Σ v. It is concave and reaches 100% at the last category.
The curve is a close cousin of the Lorenz curve used to measure inequality (Lorenz 1905). The “80/20 rule” is a rule of thumb, not a law: real data may be 70/30 or 95/5.
A little history
In 1896 the economist Vilfredo Pareto observed that land and income in Italy were concentrated in few hands, with a distribution that followed a power law (Pareto 1896). Decades later the quality pioneer Joseph Juran generalised the idea as separating “the vital few from the trivial many”, named it after Pareto, and made the sorted, cumulative chart a standard tool of quality control (Juran 1951). Juran later conceded that Pareto never applied it beyond wealth.
Questions
Why does a Pareto chart have two axes?
The left axis counts the bars; the right axis is the cumulative percentage of the same total. Both describe one quantity, so the second axis doesn’t mislead the way two unrelated scales do.
References
- Pareto, V. (1896). Cours d’économie politique. Lausanne: F. Rouge.
- Lorenz, M. O. (1905). Methods of measuring the concentration of wealth. Publications of the American Statistical Association 9(70), 209–219.
- Juran, J. M. (1951). Quality-Control Handbook. New York: McGraw-Hill.