WHAT PARETO ACTUALLY TELLS YOU_
Rank the categories from largest to smallest and calculate their cumulative share. The shape tells you whether impact is concentrated in a few categories or spread more evenly.
CAPYBARA RULE: Pareto answers “where is the recurring impact concentrated?” It does not answer “why is this happening?”
USE COMPARABLE DATA_
CAPYBARA CASE #003_
The Chaotic Capybara Café logged 100 incidents:
The top three categories account for 80% of the incidents. That is a strong prioritization signal. It still says nothing about why watermelon gets dropped or coffee gets cold.
COMMON FAILURE MODES_
Treating 80/20 as a law
The classic pattern is a useful heuristic, not a requirement. Your data may be more or less concentrated. The workspace lets you inspect a threshold rather than worship one ratio.
Mixing measures
Ten expensive failures and ten minor complaints are not comparable if one row means cost and another means count.
Using tiny or unstable samples
A ranking from a handful of events may move dramatically next week. Record the time window and data quality.
Confusing frequency with severity
The most frequent category is not automatically the highest-risk category. Choose the measure that matches the decision you need to make.
HAND THE PRIORITY TO A CAUSAL TOOL_
Once the highest-value category is clear, take that category into Fishbone to map competing causes or into 5 Whys when a coherent causal chain is already plausible. That handoff is where prioritization becomes investigation.
GOOD CHECK: If you ranked total downtime instead of incident counts, would you still choose the same priority? If not, be explicit about why this measure is the right one.