Real-Time CDP Overview and Architecture
Adobe Real-Time CDP is built on the Experience Platform foundation, and exam questions test whether you can separate the layers. At the base sits the data lake (Experience Data Lake) storing raw ingested data; on top of it, the Profile Store materializes unified customer profiles via the Real-Time Customer Profile service. Ingestion services (batch and streaming), Identity Service (identity graph), the Query Service, and Dataset Service feed and manage that store. Activation happens through Destinations, and governance is enforced by the Data Usage Policy layer.
Know the difference between the three RTCDP editions: B2C (people-based profiles, marketing activation), B2B (account and person profiles, account-based segmentation, relationship types), and B2B2C - the most capable edition, which adds the B2B capabilities to B2C. A frequent trap is assuming B2C supports account-based audiences; it does not.
Understand edge vs. hub. Edge services (Edge Network, edge segmentation, edge profile) enable low-latency personalization for web/mobile use cases, while hub-based (streaming) segmentation runs in the platform on streaming data arriving through the Edge Network. Edge segmentation has constraints: it evaluates only a subset of data available at the edge and does not support batch data.
Another trap: confusing Real-Time CDP with Adobe Audience Manager or Adobe Analytics. RTCDP is the customer data platform for profile unification and activation; it is not a reporting tool and not an email execution tool - it hands off to destinations.
Rule of thumb: ask “where does the data live, where is it evaluated, and where does it go?” - lake → profile → destination. If a question asks for sub-second web personalization, the answer involves edge services; if it asks for a nightly batch audience to a cloud storage destination, it is hub-based batch segmentation.