A Pareto chart ranks problem categories from highest to lowest frequency, using bars for frequency and a line for the cumulative percentage, embodying the Pareto principle (the 80/20 rule) that a few vital problems account for most of the effect. This helps focus improvement resources on the most important few problems, and it is one of the seven basic QC tools. It gives resource allocation a data basis and is the starting tool for selecting improvement topics; combining bars and the cumulative line makes the key problems immediately visible.
Use it for defect, complaint or cost-loss attribution; for selecting improvement topics (one of the seven basic QC tools); in quality meetings to set improvement priorities; and for before-and-after comparisons of improvement effects. A Pareto chart quantifies the relative importance of problems so the team can focus on the vital few first. Producing a Pareto chart before and after improvement makes the effect comparison obvious, and it applies to any categorical count data such as complaints, defects or downtime.
Look at the cumulative percentage line: typically the first 20% of categories contribute about 80% of the problems, and these vital few are the priority improvement targets. Also rank by loss amount (for example, by cost) rather than frequency alone. The tool supports sorting by frequency or amount and AI interpretation of the key categories. Combining cost and frequency perspectives avoids being misled by superficially frequent problems, and the inflection point of the cumulative line marks the boundary of the priority improvement zone.
Keep the categories mutually exclusive and collectively exhaustive (MECE); if the other category is too large (above 20%), re-classify. Sort by frequency or amount, whichever best reflects business impact. A Pareto chart only ranks problems, it does not explain causes, so pair it with a fishbone diagram for root-cause analysis; after improvement, use a Pareto chart again to verify the effect. Combine it with the other QC tools to move from ranking to root-cause analysis, and keep the other category small.