SF-Data-Arch Master Data Management Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Use Salesforce Duplicate Management with matching rules and duplicate rules to identify and block duplicate accounts at entry.

  Why it's wrong here

  Duplicate Management prevents and flags duplicate records during creation and update, which improves data quality but does not define which source is trusted per attribute or how to merge conflicting values into a golden record. It is a complementary control, not the mechanism for attribute-level survivorship, so it does not meet the stated requirements on its own.

• ✓

  Define source ranking and attribute-level survivorship rules so that the highest-ranked source wins per field, with documented tie-breakers.

  Why this is correct

  Source ranking with attribute-level survivorship lets the architect specify which system is authoritative for each field, such as mainframe for legal name and Salesforce for billing address. Deterministic tie-breakers ensure conflicts resolve consistently. This directly satisfies the requirement that the golden record reflect the most trusted source per attribute and that conflicts be resolved without ambiguity, making it a correct choice.

• ✗

  Configure an external object via Salesforce Connect to read mainframe customer data in real time without replication.

  Why it's wrong here

  External objects provide read-only access to remote data and do not participate in survivorship or golden record creation within Salesforce. They cannot be updated by Salesforce processes, so they cannot contribute to a merged master record. This approach addresses access, not conflict resolution, and therefore does not satisfy the attribute-level trust and deterministic resolution requirements.

• ✗

  Implement a data steward approval workflow that routes every field-level conflict to a human for manual selection before the golden record is written.

  Why it's wrong here

  Manual conflict resolution does not scale for high-volume master data and does not provide deterministic resolution as required. It also introduces latency and inconsistency because different stewards may choose differently. While stewardship is valuable for exceptions, relying on it for all conflicts contradicts the need for deterministic, rule-based survivorship, so it is not a correct capability here.

• ✓

  Establish a unique global identifier and cross-reference table that maps each source system key to the Salesforce master record.

  Why this is correct

  A global identifier with a cross-reference table enables reliable linking of the same entity across Salesforce, the mainframe, and the partner portal. Without consistent identity, survivorship rules cannot be applied to the correct records, so this is essential. It supports deterministic merging by ensuring all source records for an entity are grouped before attribute-level rules execute, making it a correct choice.