Our March 4 Community Show explored how Reltio grouping features help teams move beyond managing records one-by-one to understanding—and activating—connected context across customers, households, and organizations. Suchen Chodankar (Principal Product Manager) walked through how grouping started with automated householding and where it’s headed next as part of our longer-term Intelligent Data Graph vision.
From custom code to configurable grouping
Suchen opened with a familiar challenge: many organizations want “household” or “related entity” views, but often rely on third-party data, complex integrations, and custom logic to keep those relationships current. Reltio’s automated grouping was built to reduce that manual effort by leaning on the matching engine and introducing a grouping-focused match rule type—so you can automatically link related entities (without merging them).
In a householding scenario, for example, you can use attributes like last name, address, ZIP code, or shared contact data to group individuals into a household. That shift enables teams to:
-
View customers in context (e.g., household-level relationships)
-
Identify cross-sell and upsell opportunities based on connected entities
-
Derive new insights via aggregation (e.g., total household income, member counts, min/max values)
How it works today: simple steps, real outcomes
The walkthrough emphasized how straightforward the configuration can be:
-
Create a grouping type and define where the new grouped entity should live
-
Select which member entity types should be grouped
-
Choose the match rule that determines group membership
-
Specify the relationship type used to connect members to the group
-
Define aggregation and mapping rules to populate group attributes
In the demo, Suchen showed how a basic configuration can generate thousands of household entities from hundreds of thousands of individual records—without building a custom pipeline to stitch and maintain those relationships.
Grouping vs. deduplication: linking is not merging
A key section clarified a common question: grouping is designed to link related entities, not consolidate duplicates into a single golden record. Deduplication (match-and-merge) creates one surviving representation of a real-world entity. Grouping creates a new entity (the group) and relates members to it—preserving individual records while adding context.
What’s next: enhanced grouping and multi-level hierarchies
Suchen previewed where grouping is going next: expanding beyond householding into larger-scale, multi-level hierarchies (think stores → states → countries → global), with higher group-size limits, richer aggregation, and the ability to group “groups of groups.”
He also shared how pairing grouping with Agent Flow can unlock new ways to summarize, analyze, and ask questions of your data—like identifying top-performing regions or calculating rollups without manually defining every segment.
Please be aware this feature is currently being worked on, and is not yet ready for GA.
Q&A highlights: control, updates, and governance
The discussion surfaced practical considerations customers care about:
-
Auto-updates: Group membership can dynamically reassess as data changes—members can be removed, reassigned, or trigger new groups when needed.
-
Manual additions: Users can add members manually, and system-authored vs. user-authored associations remain distinct.
-
Governance and downstream impact: Groups are real entities and relationships, but you can control visibility and readiness through publication practices and security rules. You also choose the relationship type—so downstream consumers can ignore grouping relationships until they’re ready.
Keep the conversation going
If you joined live, thank you for the great questions and feedback. If you’re exploring grouping beyond householding—segmentation, hierarchies, rollups, or advanced analytics—share your ideas in the community. Your input directly shapes where grouping goes next.
#CommunityWebinar
#Featured