Lesson 6 — The Seven Formal Data Quality Activities
Memorize the sequence and understand why the order matters.
The spine
FRAME → DEFINE HIGH QUALITY → DIMENSIONS/RULES → INITIAL ASSESSMENT → PRIORITIZE → IMPROVEMENT GOALS → OPERATE
1. Define a Data Quality Framework
Create the repeatable enterprise approach: roles, governance alignment, methodology, standards, communications, measurement, and extension of good practices.
2. Define High Quality Data
Identify important data and articulate what “good enough” means for defined consumers/purposes. Use business context, processes, systems, Governance, Metadata, and stakeholders.
3. Identify Dimensions and Supporting Business Rules
Translate expectations into dimension-aligned, testable rules. Some rules may have to be reverse engineered from policy, code, edits, triggers, workflows, standards, or SME knowledge.
4. Perform an Initial Data Quality Assessment
Actually inspect critical data. Profile/query it, compare reality with rules/expectations, discover relationships and anomalies, and validate findings with Stewards/SMEs.
Assessment can reveal more than defects: undocumented dependencies, implied rules, contradictory/redundant data, and data that genuinely conforms.
5. Identify and Prioritize Potential Improvements
Combine data evidence with business impact, criticality, stakeholder needs, risk, cost/benefit, and feasibility. Do not rank solely by defect count.
6. Define Goals for Data Quality Improvement
Turn priorities into desired outcomes. Mix justified quick hits with strategic root-cause/preventive changes.
7. Develop and Deploy Data Quality Operations
Make quality business-as-usual through rule management, monitoring, issue workflow, SLAs, and response/reporting.
Important distinction
Prioritization decides what matters most. Improvement goals define the result you intend to achieve. They are not the same decision.
Source: pp. 441–450.