Analytics Dashboard

Analytics dashboards do not exist in any documented form. Business questions are answered by engineers running ad-hoc SQL queries and pasting results into Slack. When the CEO asks 'how are we doing on retention?' someone scrambles to write a query from scratch. There is no persistent visualization, no saved report, and no documented definition of what 'retention' means in this context.

None — AI cannot generate or maintain business intelligence because no analytics dashboard definitions, metric calculations, or visualization specifications exist anywhere.

Analytics dashboards exist but are scattered and inconsistent. Different team members have created personal dashboards in Looker, Metabase, and Google Sheets. The VP of Product has a dashboard showing 'Weekly Active Users' while Marketing has a different dashboard showing 'Active Users (Weekly)' — both claim to measure the same thing but produce different numbers. Nobody knows which analytics dashboard is authoritative.

AI could index the scattered analytics dashboards, but cannot reconcile conflicting metric definitions or determine which dashboard is authoritative because no documentation explains the differences or establishes precedence.

Analytics dashboards live in a single BI platform with consistent formatting and a naming convention. Each dashboard has a documented owner and description. Metric definitions are written in the dashboard's description field. But the definitions are prose ('retention = users who come back') without links to the underlying SQL queries, source tables, or calculation methodology. When two analytics dashboards show different retention numbers, tracing the discrepancy requires reading the SQL.

AI can catalog analytics dashboards and surface their documented descriptions, but cannot validate metric accuracy or trace discrepancies because dashboard definitions lack formal links to underlying calculation logic and source event schemas.

Analytics dashboards are well-documented with full traceability. Each chart element links to a named metric definition with documented calculation logic, source tables, refresh schedule, and data lineage. A stakeholder can click on a retention chart in the analytics dashboard and see exactly how retention is calculated, which events feed it, and when it was last refreshed. Dashboard documentation is current and findable through search.

AI can audit analytics dashboard accuracy by tracing metrics to source calculations, detect when underlying data sources change in ways that affect dashboard reliability, and recommend dashboard improvements. Cannot yet auto-generate optimal dashboard layouts because dashboards lack formal semantic relationships between their component visualizations.

Analytics dashboards are formal entities with machine-readable specifications. Each dashboard has a structured definition including layout schema, metric dependency graph, filter relationships, intended audience, and decision context. An AI agent can answer 'which analytics dashboards would be affected if we changed the retention metric calculation?' by traversing the metric dependency graph across all dashboard specifications.

AI can autonomously design analytics dashboard layouts for new business questions based on the metric ontology, detect dashboard redundancies, and generate dashboard specifications that match stakeholder needs without manual chart-by-chart configuration.

Analytics dashboards are self-documenting. When a metric calculation changes, every affected dashboard's documentation updates automatically. Dashboard specifications include auto-generated descriptions of what each visualization shows, how it is calculated, and who uses it. New analytics dashboards can be auto-generated from business questions — a stakeholder asks 'how is our enterprise expansion pipeline performing?' and the system generates a dashboard specification with appropriate metrics, filters, and layout.

Can autonomously generate, document, and maintain analytics dashboards from business questions. AI designs layouts, selects metrics, configures filters, writes documentation, and updates specifications in real-time as underlying metric definitions and usage patterns evolve.

Ceiling of the CMC framework for this dimension.