Audience Manager Fundamentals and Implementation
The core model
Adobe Audience Manager (AAM) is a data management platform built around four verbs: collect signals, classify them into traits, group traits into segments, and send segments to destinations. Understand the identity backbone, because most exam traps live there. The Audience Manager UUID is the primary key; declared IDs (customer keys such as CRM IDs or hashed emails) and Experience Cloud IDs (ECID/MID) are linked into an identity graph through ID syncs (server-side and pixel-based). Without a successful sync, profiles fragment and overlap reports look wrong.
Implementation building blocks
The Data Collection Server (DCS) ingests event calls (the /event endpoint). The Adobe Experience Platform Identity Service issues the ECID. Server-side forwarding from Analytics moves hit data into AAM without extra tagging. DIL (Data Integration Library) is the legacy on-page library; modern implementations often use the Web SDK or tags. Destinations include cookie, URL, and server-to-server types.
Traps to expect
- Confusing trait realization (a device qualifies for a trait) with segment qualification (the profile is placed in a segment).
- Assuming ID syncs are instant and universal; they are asynchronous and partner-dependent.
- Mixing edge vs batch (onboarded) data timing.
- Forgetting that subdomain and partner IDs must be configured before data flows.
Rule of thumb
If a question describes "data not appearing," check the chain in order: tag fires → DCS call → signal in Unused Signals report → trait rule → segment rule → destination mapping. The first broken link is the answer.