The Query Service is the correct tool for performing custom, join-based analysis on raw datasets within AEP. Joining the experiment exposure data (who saw which variant) with the conversion event data provides a direct, accurate, and scalable way to calculate the conversion rate for each test group, enabling a proper causal investigation beyond surface-level metrics.

This scenario describes a sequence of events over time (add to cart, then a period of no purchase), which requires the Sequence container. The logic must check for the absence of a Purchase event *within* the 1-hour window following the Add to Cart event. Option 0 correctly implements this temporal sequence with an ‘is absent’ condition.

This is a fundamental AEP concept. A Profile is the real-time, unified customer record that aggregates data from all channels. An Identity is a specific piece of information (e.g., CRM ID, email, cookie ID) that acts as a key to link disparate data fragments together into that single Profile within the Identity Graph.

For data to participate in the Real-Time Customer Profile, two things are required: 1) The schema must be built on a class that is compatible with Profile (e.g., XDM Individual Profile), and 2) The specific dataset must have the ‘Profile’ toggle switched on during configuration. Ingestion into the Data Lake alone does not make it available for real-time segmentation.

Segment Builder allows for the use of existing segments as building blocks. Placing a segment (like ‘High-Value Customers’) inside a ‘NOT’ logic container is the direct and maintainable way to create its inverse. This ensures the suppression logic automatically updates if the definition of ‘High-Value Customers’ changes.

AEP requires incoming data to adhere to the structure of the assigned XDM schema. If the JSON payload from the streaming source uses different field names (e.g., ‘pageName’ vs. ‘web.pageDetails.name’) or a different nesting structure than the schema expects, the data will be ingested into the Data Lake but will fail to map correctly to the Profile fields, resulting in missing data.

Destinations like social media platforms identify users based on their own identifiers (hashed emails, phone numbers, etc.). By selecting the corresponding identity namespace (e.g., Email) from AEP’s Identity Graph, you are specifying the actual key that the destination platform will use to find the right user in their system and deliver the activation.

Calculated attributes are computed on a scheduled batch basis. A segment rule that depends on such an attribute inherits that batch evaluation cadence. The segment population will refresh after the calculated attribute job runs (in this case, nightly), not in real-time based on individual profile events.

The primary purpose of the Identity Graph is identity stitching. By linking the anonymous browser cookie (ECID) with known identifiers (CRM ID, email), AEP can unify behaviors from web sessions, mobile apps, and offline systems into one profile. This allows segmentation and activation to address the individual, not just separate identifiers.

The Query Service is the correct tool for performing custom, join-based analysis on raw datasets within AEP. Joining the experiment exposure data (who saw which variant) with the conversion event data provides a direct, accurate, and scalable way to calculate the conversion rate for each test group, enabling a proper causal investigation beyond surface-level metrics.0