Short answer
Data quality makes important data fit for its intended use by defining acceptable conditions, detecting failures, tracing causes, and verifying repairs.
About Data quality
Data quality defines, monitors, and improves the accuracy, completeness, and fitness of data. Quality is judged against a stated use, not against an abstract goal of perfect data.
Use this competency for
- Roles accountable for datasets, pipelines, metrics, or operational records used by others.
- Work where missing, stale, duplicated, or invalid data can change decisions or system behavior.
Do not use this competency for
- Use data governance when the issue is who sets policy, owns definitions, or approves access rather than whether data meets a use-specific condition.
Important distinctions
Analytics engineering
Analytics engineering builds and operates transformations, while data quality defines and verifies whether outputs are fit for use.
Data governance
Data governance establishes accountability and rules, while data quality measures and improves the condition of data.
Expectations by level
IC1
IC1: Quality checks
Applies defined quality checks to a known dataset with guidance. Investigates common failures, documents impact, and verifies the correction.
Observable behaviors
- Runs checks for required fields, valid ranges, freshness, and duplicates.
- Traces a failed check to the affected source or transformation step.
- Confirms corrected records satisfy the original condition.
Examples
- Finds duplicate invoice rows after a load, identifies the retry that created them, and verifies the deduplication result.
- Flags a stale daily extract before a report refresh and records which metrics are affected.
IC2
IC2: Quality controls
Independently defines controls for a team data domain based on consumer needs. Prioritizes incidents by impact and prevents repeated failures.
Observable behaviors
- Defines measurable quality conditions with dataset consumers.
- Adds monitoring at the point where a failure can first be detected reliably.
- Completes root-cause analysis and implements a prevention step.
Examples
- Works with finance to define completeness and reconciliation checks for monthly revenue data.
- After an identifier change breaks joins, adds a contract check and a controlled backfill process.
IC3
IC3: Quality strategy
Leads data quality across connected domains used by multiple teams. Sets risk-based standards, clarifies ownership, and reviews whether controls cover critical uses.
Observable behaviors
- Maps critical data uses to owners, quality dimensions, and response targets.
- Creates shared incident and remediation practices for high-impact data failures.
- Reviews recurring quality trends and directs structural fixes across systems.
Examples
- Establishes tiered controls for customer, billing, and product data based on operational impact.
- Coordinates a cross-team repair when late events affect reporting, model features, and customer workflows.