Data quality business rule types

1. 01

   Definitional conformance

   Same understanding of definitions implemented + used consistently (incl. calculated fields, rollup rules)

   Build this rule
2. 02

   Value presence & record completeness

   Conditions under which missing values are acceptable/unacceptable

   Build this rule
3. 03

   Format compliance

   One or more patterns specify valid values (e.g., telephone number formatting)

   Build this rule
4. 04

   Value domain membership

   Assigned value must be in the defined domain (e.g., 2-char US postal codes for STATE)

   Build this rule
5. 05

   Range conformance

   Value must be within defined numeric/lexicographic/time range

   Build this rule
6. 06

   Mapping conformance

   Value must correspond to one from a domain mapping to equivalent domain(s) (e.g., 'AL' + '01' both map to 'Alabama')

   Build this rule
7. 07

   Consistency rules

   Conditional assertions maintaining relationships between 2+ attributes (e.g., postal codes correspond to correct states)

   Build this rule
8. 08

   Accuracy verification

   Compare value against system of record or verified source

   Build this rule
9. 09

   Uniqueness verification

   One and only one record exists for each real-world object

   Build this rule
10. 10

    Timeliness validation

    Expectations for accessibility and availability of data

    Build this rule

Aggregation checks (sets of data)

• Reasonableness of record count (statistics over time)
• Reasonableness of average amounts (thresholds + trends)
• Expected variance in transaction counts (statistics + thresholds)