Definition
A Pareto chart is a bar chart that arranges problem categories in descending order of frequency or impact, overlaid with a cumulative percentage line. It makes immediately visible which few causes account for the majority of a problem — the “vital few” versus the “trivial many.”
The principle behind it: in most situations, roughly 80% of the effect comes from roughly 20% of the causes. A Pareto chart turns this insight into a visual decision-making tool. Instead of attacking all problems equally, the team focuses resources on the one or two categories that dominate.
Japanese Origin
パレート図 (parēto zu) is written in katakana, indicating it is a borrowed term. Vilfredo Pareto was a 19th-century Italian economist who observed that approximately 80% of Italy’s land was owned by 20% of the population. Joseph Juran applied this observation to quality management, naming it the “Pareto principle,” and taught it in Japan during his 1954 visit.
図 (zu) means “diagram” or “chart.” The Pareto chart became one of the most widely used of the 7 QC tools precisely because it translates abstract data into an immediately actionable visual: the tallest bar is the priority.
History
Vilfredo Pareto, 1896 — Pareto published his observation about wealth distribution in Italy. He noted that the distribution was consistently uneven across countries and time periods.
Joseph Juran, 1940s-50s — Juran generalized Pareto’s observation into a universal principle: in any situation, a small number of causes produce the majority of the effect. He coined the phrases “vital few and trivial many” (later revised to “vital few and useful many”). Juran brought this concept to Japan during his 1954 JUSE lectures, where it became foundational to the Japanese quality movement.
Kaoru Ishikawa, 1960s — Ishikawa included the Pareto chart in his codification of the 7 QC Tools, establishing it as a standard technique for QC circles and shop floor quality improvement across Japanese industry.
At Toyota — Pareto analysis is routine in A3 problem-solving and QC circle activities. When a quality problem is identified — for example, through andon stop data or customer complaints — the first analytical step is often a Pareto chart to determine which defect types, which machines, which shifts, or which parts account for the largest share of the problem.
How to Construct a Pareto Chart
- Define the problem and data collection period — What are you measuring? Defect types? Causes of downtime? Customer complaints by category? Over what time period?
- Collect data — Use a check sheet to tally occurrences by category
- Rank categories — Sort from highest frequency to lowest
- Draw bars — Each bar represents a category, arranged left to right in descending order
- Calculate cumulative percentage — Add a line showing the running total as a percentage of the whole
- Identify the vital few — The categories that account for 70-80% of the total are the priority targets
How to Read and Use a Pareto Chart
The chart answers one question: where should we focus?
- The leftmost bars represent the biggest contributors. These are the targets for improvement.
- The cumulative line shows how much of the total problem is addressed by the top categories. If the top two bars account for 75% of defects, solving those two problems eliminates three-quarters of the total.
- After implementing a countermeasure, create a new Pareto chart with fresh data. If the countermeasure worked, the dominant bar should shrink or disappear, and a different category may now be the tallest bar — the new priority.
Before-and-after Pareto charts are a powerful way to demonstrate improvement. They show both what changed and what the next priority should be.
Common Mistakes
Creating a Pareto chart without acting on it. A Pareto chart is a decision tool, not a display piece. If it identifies the top problem category but no countermeasure follows, the chart is waste.
Using categories that are too broad or too narrow. “Quality problems” is too broad — it does not guide action. “Scratch on part 4507-B left rear surface, 3mm from edge” is too narrow — it fragments the data. Categories should be specific enough to guide investigation but broad enough to aggregate meaningful data.
Not stratifying the data. A Pareto chart that combines data across all shifts, machines, and operators may hide the real pattern. Stratifying — creating separate Pareto charts by shift, by machine, or by product line — often reveals that one shift or one machine accounts for the majority of the problem.
Pareto charts based on opinion instead of data. The chart must be built from counted, measured, observed data — not from team opinions about what the biggest problem “probably” is. This is why the check sheet comes before the Pareto chart in the QC tool sequence.
Never updating the chart. A Pareto chart is a snapshot. After improvement actions, the distribution changes. Teams that post a Pareto chart once and never update it are looking at old news.