Create a Cohort Retention Analysis from Transaction Data

ROLE Act as a lifecycle analytics specialist. CONTEXT Use [TRANSACTION WORKBOOK] to analyze [RETENTION DEFINITION] with [COHORT INTERVAL] cohorts in [SPREADSHEET PLATFORM]. INSTRUCTIONS Verify customer identifiers, transaction dates, qualifying activity, and the available observation window. Define cohort assignment and elapsed periods precisely, including the handling of refunds, duplicate transactions, and customers first observed before the dataset begins. Build a customer-period activity table and a cohort retention matrix using unique customers rather than transaction counts. Keep periods that have not yet occurred blank instead of treating them as zero retention. Distinguish period-specific, rolling, and cumulative retention, and implement only the requested definition. Compare cohorts cautiously when sizes or acquisition mixes differ. CONSTRAINTS Avoid fabricated statistics and unsupported explanations for retention changes. Flag assumptions, left-censoring, and incomplete follow-up. Do not imply that newer cohorts outperform older ones when observation periods differ. OUTPUT FORMAT Provide a workbook layout, helper-column formulas or pivot instructions, a heatmap specification, interpretation guidance, and validation tests. Suggest three follow-up analyses that could explain patterns without presenting hypotheses as findings.