Pareto Analysis: What Good Looks Like vs What Bad Looks Like
Pareto analysis is built on an empirical observation: in most processes, a small number of causes — typically around 20 percent — account for the large majority of effects, typically around 80 percent. The 80/20 ratio is not a law of nature; it varies. The underlying principle, which Vilfredo Pareto documented and Joseph Juran applied to quality management, is that effects are not equally distributed. A small number of defect types, suppliers, complaint categories, or cost drivers create most of the problem. Pareto analysis is the tool that identifies which ones — the vital few — so that improvement effort can be concentrated where it will have the most impact. In 2022, with organisations managing inflation, labour shortages, and materials delays simultaneously, the ability to focus limited resources on the causes that matter most has rarely been more valuable.
What Good Looks Like
Builds the Pareto chart from actual frequency or impact data, not from perceptions or committee opinions. The y-axis represents a meaningful quantity — number of defects, cost of rework, complaint volume, lost production hours — not a subjective rating.
Separates the vital few from the trivial many and focuses improvement effort on the top one to three categories. A well-run Pareto analysis does not try to fix everything; it selects the bar or bars on the left of the chart that collectively account for 60 to 80 percent of the total effect, and works there first.
Drills down with a secondary Pareto. Once the dominant category is identified, the team runs a second Pareto analysis within that category to find the dominant sub-cause. Two rounds of Pareto analysis often produce a specific, actionable root cause that a single round would miss.
Updates the chart after corrective action to confirm effect. A Pareto chart is a before-and-after tool. Running the same analysis after a countermeasure is implemented either confirms that the distribution has shifted (the intervention worked) or shows that it has not (the root cause was not actually addressed).
Pairs with a 5 Whys or fishbone diagram to move from "what is the biggest problem" to "why is it happening and how do we fix it."
What Bad Looks Like
Uses a ranking based on gut feel or stakeholder opinion rather than data. "We think supplier quality is the main issue" is not Pareto analysis; it is a guess. Pareto analysis requires a dataset.
Treats all categories as equally important and spreads improvement resources across all of them. The entire point of Pareto analysis is to break the assumption that all causes deserve equal attention. A team that identifies ten causes and assigns one person to each has wasted the insight.
Builds the chart but does not act on it. Pareto analysis is a prioritisation tool, not a report. A chart produced for a steering committee presentation that does not lead to a focused corrective action is wasted effort.
Stops at the first chart and does not drill down. If the top category accounts for 60 percent of defects, the question is not "good, we know what to fix" — it is "which sub-type within that category accounts for 60 percent of those defects." Stopping at the surface level misses the actionable insight.
Conflates frequency with impact. A defect type that occurs 50 times per month at a cost of $10 each is less important than one that occurs five times per month at a cost of $2,000 each. A frequency Pareto and a cost Pareto can lead to different priorities. Build both when data permits.
XNM supports public-sector and capital-project clients in applying Lean Six Sigma analysis tools — including Pareto, fishbone, and control chart methods — to focus improvement effort where it creates the most value. Connect with XNM's strategic advisory team to discuss how structured analysis could help your organisation address its most persistent quality and process problems.