Data Quality

Technical explanation

Quality is evaluated against explicit requirements and context. Controls include schema validation, reference checks, reconciliation, deduplication, anomaly detection, data contracts, issue workflows, ownership, and monitoring across sources and transformations.

Business relevance

Poor-quality data produces unreliable reporting, failed automation, weak customer experiences, regulatory exposure, and unsafe or ineffective AI decisions.

Implementation example

A revenue team defines required CRM fields, valid lifecycle transitions, uniqueness rules, and freshness targets, then monitors violations and assigns remediation owners.

Limitations and common misconceptions

Quality is purpose-dependent: data suitable for one decision may be inadequate for another. A dashboard score can hide critical field-level failures, and cleansing downstream does not solve defective source processes.

Topics

Sources