The Data Warehousing and BI Plan: One Definition, One Report of Record
Every organization has had the meeting: two dashboards show two different numbers for the same metric, and the first half hour is spent arguing about which one is right. The cause is almost always the same â KPIs defined differently in different places, built on different sources, with no single report everyone agrees to trust.
The Data Warehousing & BI Plan â aligned with the data warehousing and business intelligence knowledge area of DAMA's DMBOK â prevents that. This article explains its contents and how ADPA generates it.
What the Plan Contains
| Section | Purpose |
|---|---|
| 1. Purpose and scope | Which analytics and reporting are covered |
| 2. Subject area map | The analytical subject areas, such as sales, finance, and operations |
| 3. Source traceability | Which governed sources feed each subject area |
| 4. KPI definitions | Precise definitions and calculation logic for each KPI |
| 5. Dimensional model overview | The facts and dimensions behind the reports |
| 6. Metric ownership | Who owns each KPI and its definition |
| 7. Report of record register | The single authoritative report for each KPI |
| 8. BI access tiers | Who may see which reports and data |
| 9. Review cadence | When definitions and reports are reviewed |
How ADPA Generates It
Inputs
- Systems of record and data domains from the architecture and governance charters.
- Master data entities from the Master Data Management Plan.
- The existing BI and analytics tool inventory and BI strategy documentation.
Grounding
ADPA's Governance Knowledge Graph is checked for governed domains and master data entities â so subject areas and KPIs are built on trusted, deduplicated data â and for KPI definitions that already exist elsewhere in the organization, so new definitions don't conflict with them.
Outputs
Structured entities for subject areas, dimensional model references, KPI definitions, and reports of record.
Quality Gates
- KPI traceability: every KPI names a source master data entity or system of record (blocking).
- No duplicate KPI logic: no KPI may be defined twice with conflicting logic in the same draft (blocking).
- Single report of record: no KPI may have more than one report of record (blocking).
These three rules capture the essence of trustworthy BI: every number has a known source, one definition, and one official home.
Where It Fits
The BI plan sits downstream of governance, architecture, modeling, and master data. Its KPI definitions then feed the Metadata Management Plan, where they become part of the business glossary and lineage â so anyone can see exactly where a number comes from.
Conclusion
Business intelligence is only as valuable as the trust people place in it. Defining each KPI once, tracing it to governed sources, assigning an owner, and naming a single report of record turns dashboards from a source of debate into a basis for decisions.