Skip to main content
sessoms25
Adobe Champion
Adobe Champion
September 11, 2026
New

Many-to-One (N:1) schema relationships between non-B2B XDM schemas

  • September 11, 2026
  • 1 reply
  • 40 views

I would like to share a key feature enhancement request for Adobe Experience Platform (AEP) Real-Time CDP architecture regarding data modeling flexibility. Ideally for B2C, B2B & B2P editions as we have consolidated all into single Prod & Dev sandboxes as advised by Adobe Ultimate Support.


The Feature Upgrade:

  • Enable native Many-to-One (N:1) schema relationships between non-B2B XDM schemas. This will allow multiple custom data records to link directly to a single parent account record within standard XDM profiles, without requiring both B2B XDM classes.

Use Case & Business Need:

  • Current Scenario: We have custom data entities (e.g., competitor products, telemetry, certifications) where multiple rows of data map to a single core business account.
  • The Problem: Standard non-B2B XDM schema relationships are restricted to lookup options, which force data into arrays. This results in complex architecture and confusing schema building UI.
  • Proposed Solution: A flexible parent-child relationship model that natively links N custom data rows to 1 parent account row in the Profile Service.

Value & Impact:

  • Cleaner Data Architecture: Eliminates the need for complex array manipulation.
  • Smarter Audience Segmentation: Allows segment builders to have distinct paths to attributes, and creates cleaner backend data models.
  • Speed to Value: Avoids architectural re-engineering/forced schema workarounds just to support custom relational data models.

1 reply

Level 2
September 29, 2026

Another important consideration is avoiding forced denormalization.

In many enterprise architectures, Account, Customer, Product, Certification, Subscription, and Telemetry data are maintained as independent entities in source systems. Requiring customers to replicate parent-level attributes into every child record increases data volume, creates synchronization challenges, and introduces potential data-quality inconsistencies.

A native relationship layer would allow AEP to preserve the semantic relationship while keeping the underlying data model normalized. ​@PavanBabuGandla_Analytics