Contact Point Phone expects one row per phone number with a type indicator (Mobile, Home, Work). Using streaming transforms, the Contact data stream can be exploded into a separate Phone DLO that produces three rows—one for each field—and adds the phone type attribute. This normalized DLO is then mapped to the Contact Point Phone data model object, making every phone number separately addressable for identity resolution and activation with a single ingestion.

A. Mapping three different fields in one row violates the one-row-per-number model and drops type metadata.

C. Calculated insights create metrics, not normalized records for identity and activation mapping.

D. Formula fields stay columnar inside the same row; they still fail the normalization requirement.

1. Salesforce Help, “Normalize Phone Data with Streaming Transforms,” Data Cloud Data Ingestion, section “Explode Columns,” para 2-4.

2. Salesforce Data Cloud Developer Guide, “Contact Point Phone Object Definition,” Table 1, Note on ‘one record per phone’.

3. Salesforce Architects, “Designing for Multi-valued Attributes in Data Cloud,” Tech Brief, p. 5-6.