Correlation and causation

Correlation is the observation that two measures move together; causation is the claim that one produces the other. The distinction matters in measurement systems because every dashboard, by construction, shows correlations: the initiative launched in March and the indicator improved in April. Whether the initiative caused the improvement is a different question — the improvement may come from seasonality, from another change made the same month, or from the measurement itself changing.

A measurement system preserves the distinction by being honest about what it asserts. Relating an initiative to an indicator and reviewing them together is legitimate and useful; declaring that the initiative moved the indicator requires ruling out the alternatives, which is analysis and judgement, not a chart. The practical test when reading any board: ask what would have happened without the intervention. If the system cannot help answer that — with history, with segments, with the record of what else changed — then what it shows is a coincidence of dates, however encouraging.