Multiple records within the same XDM Individual Profile class-based dataset are all related to each other through the identity graph. At the time of profile assembly, what would be expected in terms o

Question

Options

• AProfile will keep only one fragment and ignore the other fragments completely.
• BProfile will randomly pick a fragment as truth and use the other fragments in timestamp order to fill
• CProfile will use the last processed fragment as truth and use the other fragments in timestamp
• DProfile will merge all fragments and overwrite previous data randomly.

• A
  5% (1)
• B
  5% (1)
• C
  79% (15)
• D
  11% (2)

Explanation

Option C is correct because Adobe Experience Platform's Real-Time Customer Profile uses a "last processed wins" merge strategy: the most recently processed fragment becomes the authoritative source of truth, and remaining fragments contribute attribute values in timestamp order to fill any gaps. This is deterministic and predictable - not random - ensuring a consistent, reproducible merged profile.

Why the distractors fail:

• A is wrong because no fragments are discarded; all linked records contribute to the final unified profile.
• B is wrong because fragment selection is not random - it is always the last processed fragment that anchors the merge, not an arbitrary one.
• D is wrong because overwriting happens in a defined timestamp order, not randomly - the process is governed by merge rules, not chance.

Memory tip: Think of it as a "latest draft + revision history" model - the last processed record is the current draft, and older fragments act like revision history that fills in anything the draft left blank.

No community discussion yet for this question.