skills/build-report/references/report-archetypes.md

# Optional report patterns

Use these as starting ideas, not required structures. Combine or omit parts according to the question and available evidence. The [public component examples](../../../templates/data-app/base/examples/README.md) show how to implement the pieces in the shared runtime.

| Reporting job | Useful composition | Common mistake |
| --- | --- | --- |
| One-metric brief | Answer, compatible comparison, one useful visual or exact values, material limitation | Inflating a narrow answer with forecasts or recommendations |
| Operating review | Current state, meaningful changes, selected drivers/segments, risks and feasible actions | Turning the sample's metric/source list into a quota |
| Experiment readout | Decision, design/population, effect and uncertainty, guardrails, interpretation | Calling an inconclusive result a win or inferring causality from observational data |
| Forecast/scenario | Observed baseline, labeled projections, assumptions, sensitivity, decision range | Drawing future values as if they were observed |
| Technical/data-quality assessment | Scope, test/method, affected data, findings, limitations, reproducible evidence | Hiding the method when it is the reason for the report |
| Opportunity/segment analysis | Relevant segments, comparable denominators, size/tradeoffs, evidence-backed prioritization | Ranking segments using incompatible populations |

The [substantial adoption/retention example](../../../templates/data-app/base/examples/reports/adoption-retention/README.md) is one operating-review composition. Its executive/technical variants and charts are available to borrow selectively; its business story does not define a valid report.