Dimension Pattern

Identity Resolution

Identity Resolution connects different identifiers that represent the same business entity across systems and over time.

Problem

The same real-world entity can have different IDs across systems.

A customer may have one ID in CRM, another ID in billing and a different ID in a policy or contract system.

Historical reporting becomes difficult when these identities change, merge, split or become visible at different times.

Duplicate entitiesBroken joinsIncorrect attributionLate identity corrections

Example

One customer appears under different IDs in CRM, billing and policy systems.

CRM

C-1029

Billing

B-8841

Policy Holder

P-5510

Historical challenge

The mapping may become available after reports were produced, or one ID may be merged into another in a later source correction. The model must decide how identity mappings affect historical reports and cross-system joins.

Why it happens

Identity mappings are often historical facts themselves.

Cross-system IDs are not always stable. They can be created late, corrected later, merged into canonical IDs or split into multiple business entities. In historical models, identity is not only a matching problem — it is also a time-dependent modeling problem.

CRM + billing integrationMaster data managementSystem migrationsMerge historySplit historyConformed dimensions

Common modeling approaches

Model identity mappings as historized relationships.

Cross-reference table

Store source IDs and their canonical business entity ID in an explicit mapping table.

Historized mapping

Add valid-time intervals so ID mappings can change without rewriting the past.

Canonical entity ID

Introduce a stable surrogate business entity ID for reporting across systems.

Bitemporal mapping

Track when mappings were valid and when they became known if identity corrections arrive late.

Validation checks

Validate both identity uniqueness and historical coverage.

✓ Detect multiple source IDs for the same business entity✓ Validate one current canonical ID where required✓ Check historical identity mapping coverage✓ Detect identity mapping overlaps and gaps✓ Measure impact of late identity corrections

Why it matters

Reliable identity mapping is the foundation for cross-system historical reporting.

Without stable identity mapping, even perfectly historized source tables cannot be reliably joined.

Identity Resolution determines whether customer-level, contract- level or policy-holder-level reporting can be trusted across systems and over time.

Related Patterns

Historical Conformance Relationship History Snapshot Reproducibility Historical Correction State ↔ State Alignment

Try it

Explore cross-system historical relationships in the Workbench.

Use the Historical Modeling Workbench to reason about historized sources, cross-system joins, relationship history and temporal validation risks.

Open Historical Modeling Workbench →