Cohort Triangle

A cohort triangle is a table of acquisition cohorts by periods since acquisition, triangular in shape because newer cohorts have lived through fewer periods.

The table must be read in two directions and each answers a different question. Moving along a row traces one cohort's decay, while moving down a column lines up cohorts at the same age, and the column is where it becomes visible whether the product improved over time. The empty lower-right corner is not missing data; it is a future that has not happened yet. A frequent error is reading the last column as a trend, since it is filled by the oldest cohort alone, so a drop or jump at the right edge is a sample-size artefact rather than a behavioural change. The diagonal represents the most recent calendar period, and an external event such as an outage leaves its mark along that diagonal.