A Pareto chart ranks causes by their contribution and plots them in descending order, usually with a cumulative line. It rests on the observation that a small share of causes accounts for most of an effect — often quoted as 80/20, though the actual ratio varies and does not matter.
What it settles is where to start. Faced with twenty problems and capacity for three, the chart converts an argument about priorities into an arithmetic one. That is most of its value: not the insight, but the end of the debate.
The choice of unit decides the answer. Ranking defects by count and ranking them by cost frequently produce different orders, and the frequent-but-cheap category is a classic way to spend a year on the wrong problem. Rank by whatever the organisation actually cares about.
The other failure is treating the long tail as noise. A Pareto reflects the period it was drawn from; a small category that is growing will not look important and will be the next chart’s first bar.
