Glossary term
Data Quality Management
Subject: Technology, Cyber and Data, IT and Data Governance
Also known as: Data integrity, Data quality
Data quality management is the ongoing work of defining, measuring and improving how accurate, complete, consistent, timely and valid data is, so that it is fit for the decisions and reports that rely on it.
Explanation
Poor data leads to wrong decisions, failed reports and customer errors, often without anyone noticing at the time. The management cycle begins by agreeing what good means for each critical data item, using dimensions such as accuracy, completeness, consistency, timeliness, uniqueness and validity. Rules are then built into checks, for instance that a date of birth is a real date or that a customer has only one record. Results are scored and tracked, issues are logged with an owner, and root causes are fixed at the source rather than patched downstream. Data integrity, in this sense, means data stays correct and unaltered through its life, and it is a common regulatory theme in pharmaceuticals and finance. Data quality sits within data governance, which names the owners accountable. A risk model fed incomplete or duplicated records produces confident-looking figures that are wrong, and nothing in the output flags it.
Example
The compliance team at a wealth manager notices that many customer records have no country of birth, so screening rules cannot work properly. The data owner sets a completeness target of 99%, adds a mandatory field at onboarding, and runs a clean-up for existing customers. Monthly dashboards track progress against the target.
Not to be confused with
- Data Security
- Security protects data from unauthorised access or loss; quality concerns whether the data is correct, complete and usable.