Q: Assuming that there's a data governance organization already in place, the question now arises how can one measure the quality of data - both on master data and transactional data? And what KPIs can be established for monitoring these?
Sid Adelman's Answer:
First determine what data is most important - the data owners on the business side should be able to tell you what is most important - and then profile that data for completeness, valid values, conformance to business rules, uniqueness, and redundancy.
The business owners should help you develop the KPIs. They will tell you the importance of the quality of each of the data elements and they will tell you - assuming you ask the right questions - the types of data quality that are important to them, e.g., completeness, currency, and they will tell you how they want the data measured. Be careful not to expect perfection but recognize that some data elements have a higher data quality threshold than others.
The KPIs should include the data that are the sources to the data warehouse. It's always cheaper and easier to capture the data correctly rather than trying to fix it later in the cycle so there should be an emphasis on edit checks and incentives in the data entry system to capture the data correctly, not to capture it fast.
Anne Marie Smith's Answer:
Validity Percentage: Percentage of occurrences in a source element that contain a value that is within the accepted domain of the target element.
Completeness Percentage: Percentage of source element occurrences that contain values that considered to be "not missing."
Accuracy Percentage: Percentage of occurrences of a source element that contain values that are considered to be correct and have business integrity.
Consistency Percentage: Percentage of occurrences of a source element that have values in agreement with those contained in the same element in another source.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access