Business Intelligence
Using data to inform strategic decisions.
Workflow
- 01Query Formulation
- 02Data Visualization
- 03Trend Identification
- 04Decision Support
Data in and out
- IN Reporting extracts from General Ledger · dailyWithout it: Management reporting diverges from the ledger.
- IN Trial balance extract from General Ledger · daily
- IN Cash position and forecast from Cash Management · daily
If you change this
It implies 4 topics across one module.
See the path ahead →Why it matters
BI is where data either earns trust or loses it. It is also where the same measure most often appears with different values, because each report was built by someone solving their own problem.
- It is the visible surface: Everything upstream is judged by it.
- Duplication is the failure mode: Six dashboards with five definitions of revenue.
- Adoption is the measure: An unopened report is an unrealised cost.
The workflow
- 01
Question definition
What decision the report supports, and who makes it.
- 02
Measure selection
Metrics chosen from the register rather than invented.
- 03
Build
The report or dashboard constructed against the shared model.
- 04
Validation
Numbers checked against the source of truth before release.
- 05
Publication
Released with owner, definitions and as-at date visible.
- 06
Review
Usage monitored; unused reports retired rather than accumulated.
Retire reports. A BI estate that only grows becomes a maintenance burden where nobody knows which of the four similar dashboards is current.
What you need in place
- Definitions from the register: Not re-derived per report.
- A certified layer: Reports that have been validated, distinguished from ad hoc exploration.
- Self-service with guard rails: People can explore, but certified numbers stay certified.
- Usage telemetry: So retirement is evidence-based.
- An owner per report: Named, and asked annually whether it is still needed.
Questions to ask
01Certification
Can a user tell a validated report from someone's experiment?
If not, all of them get quoted as fact.
02Duplication
How many reports show revenue, and do they agree?
Check before trusting any.
03Usage
Which reports are opened, and by whom?
Most estates have a long tail nobody uses.
04Definitions
Do report measures come from the register?
Otherwise the register is decorative.
05Performance
How fresh is the data behind the dashboard?
Users assume live unless told otherwise.
Metrics and KPIs to track
Certified coverage
Key measures available in a validated report.
Report usage
Opens per report, with the unused tail identified.
Definition conformance
Reports using register definitions.
Data freshness
Lag between event and availability.
Alongside this