A financial institution is migrating its customer transaction data from a relational database (RDBMS) to...

In Adobe Real-Time CDP, transaction data is inherently behavioral and time-bound. TheXDM ExperienceEventclass is the optimized choice for this data type because it is designed to capture immutable, point-in-time actions. Each transaction (containing amount, type, and date) should be treated as an event. By including thecustomer IDwithin this schema and marking it as anIdentity, the platform’s Identity Service automatically associates these events with the correspondingXDM Individual Profile.

This approach is superior to Option C because the Individual Profile schema is intended for stateful attributes (like "current balance" or "account level"), not a growing list of transactions. Storing transactions in the profile would lead to extremely large profile fragments, degrading performance. Option A is inefficient as it creates schema sprawl; instead, a single ExperienceEvent schema should use a "transaction type" field to differentiate between deposits, withdrawals, or transfers.

By leveraging the NoSQL architecture of the Real-Time Customer Profile, these events are stored in a way that allows theSegmentation Serviceto evaluate them in milliseconds. For example, a segment could instantly identify "customers who made a transaction over $1,000 in the last hour." Linking via the customer ID ensures that as soon as a transaction is ingested, it is immediately visible on the unified profile for real-time personalization.