Mixed Scenarios — Resolution Key 07–12
7 — Valid Status, Rising Defects
Primary: the expected rule is already defined; the new need is condition/root-cause diagnosis.
Leading KA: Data Quality.
Supporting: Governance/Stewardship; Metadata; source operations.
Roles: DQ Analyst leads analysis; Data Steward supplies/owns business expectation; process/system owner remediates cause.
Best action: profile/monitor the defect spike, isolate root cause, then change the defect-producing process/control; retain the approved status domain unless evidence shows it is wrong.
Weaker: add new valid statuses merely because invalid values increased—the volume does not prove the rule is wrong.
Changed fact: if stakeholders actually disagree on what statuses should be valid, Governance/Stewardship becomes leading.
Sources: Steward vs DQ Analyst; DQ activity boundaries.
8 — Product Means Something Different in Every Mart
Primary: analytical areas lack consistent shared descriptive meaning.
Leading KA: DW/BI, specifically dimensional conformance.
Supporting: Master/Reference Data; Governance; Metadata; Data Quality.
Roles: BI/DW Architect, Product Data Owner/Steward, MDM/RDM team, Metadata specialists.
Best action: govern and establish a reusable/conformed Product dimension/meaning across Sales and Inventory, drawing on trusted master/reference sources and preserving mappings/history.
Weaker: one-off dashboard mapping—it may make one report agree while marts remain semantically inconsistent.
Changed fact: if the real problem is that multiple Product records cannot be determined to represent the same product entity, MDM entity resolution becomes leading.
Sources: Fact/Dimension/Conformed Dimension; Reference vs Master.
9 — Perfect Format, Wrong Person
Primary: value is rule-conforming but does not correspond with reality.
Leading KA: Data Quality — Accuracy.
Supporting: source process/system ownership; DW/BI as consumer; Governance/Stewardship.
Roles: DQ Analyst, Data Steward, source process owner.
Best action: verify source/reality evidence and repair the identity/value association; investigate the process that assigned the wrong email.
Weaker: stricter regex—the existing problem already passes format/domain validation.
Changed fact: if the email violates an approved format/domain rule rather than belonging to the wrong person, Validity becomes the better dimension.
Source: Validity vs Accuracy.
10 — Governed Codes vs Warehouse Conversion
Primary: enterprise permissible values/mappings have not been governed; the warehouse is about to invent local semantics.
Leading KA: Reference Data Management / Governance.
Supporting: DW/BI transformation; Metadata; Data Quality.
Roles: Reference Data Steward/Owner, governance body, integration/DW team.
Best action: define authoritative code set/mappings and change process first, publish the governed mapping, then implement target transformation in the warehouse.
Weaker: choose a local ETL conversion without enterprise authority—this can create another unofficial reference scheme.
Changed fact: if the governed mapping already says source US should become target 840, the remaining task is DW/BI transformation implementation.
Sources: Reference vs Master; Mapping/Remediation/Transformation.
11 — Temporary Load Area or Operational Store?
Primary: staging and ODS are being confused because both can sit between sources and analytical stores.
Leading KA: DW/BI.
Supporting: Data Architecture; Integration.
Roles: DW/BI Architect, integration engineers, operations.
Best action: classify the transient truncated preparation table as Staging; classify the integrated retained near-current operational-reporting store as ODS.
Weaker: classify by physical database/location rather than persistence and purpose.
Changed fact: if the retained store becomes durable integrated multi-year analytical history, Data Warehouse becomes the closer role.
Source: Staging vs ODS vs DW vs Mart.
12 — Clean the Addresses Again
Primary: repeated correction is substituting for causal process improvement.
Leading KA: Data Quality.
Supporting: Governance/roles; source operations; Metadata if rules/definitions matter.
Roles: DQ Analyst facilitates diagnosis; Data Steward clarifies expectations; source process/system owner owns preventive/root-cause change.
Best action: analyze and remove the defect-producing workflow/cause, add prevention/control, then correct the existing backlog.
Weaker: automate monthly cleansing only—efficient correction can still leave the cause untouched.
Changed fact: if the underlying source cause was already fixed and only historical bad records remain, Correction becomes the primary remaining activity.
Source: Prevention vs Correction vs Root Cause.