Journey Optimizer - Decision Management on the Edge blueprint
To learn more about Decision Management refer to the product documentation HERE and the Decision Management Overview HERE
Adobe Decision Management is a service provided as part of Adobe Journey Optimizer. This blueprint outlines the use cases and technical capabilities of the application and provides a deep dive into the various architectural components and considerations that make up Decision Management.
Decision Management can be deployed in one of two ways. The first is via the Adobe Experience Platform Hub, which is a single data center architecture. In the “hub” approach offers are executed, personalized, and delivered in second latency. Thus the hub architecture is best suited for customer experience that do not demand sub-second latency, examples include offer decisions which are provided for kiosks or agent assisted experiences such as in call centers or in person interactions.
The second approach is via the Experience Edge Network, which is a globally distributed geographically located infrastructure to serve fast sub-second and millisecond experiences. The end consumer experience being executed by the edge infrastructure closest to the consumers geo-location to minimize latency. Decision Management on the Edge is designed to serve real-time consumer experiences. These include experiences such as web or mobile inbound personalization requests.
This blueprint will cover the specifics of Decision Management on the Edge.
To learn more about Decision Management on the hub refer to the Decision Management on the hub blueprint.
Use cases for Decision Management on the edge
- Online personalization via web or mobile inbound experiences.
- Cross channel journey execution - offer consistency across web, mobile, email, and other interaction channels through Adobe Journey Optimizer.
|Decision Management with Adobe Target
||Decision Management can be integrated with Adobe Target such that offers can be tested and delivered as Target experiences.
Adobe Experience Platform
- Schemas and datasets must be configured in the system before you can configure Journey Optimizer data sources
- For Experience Event class-based schemas add 'Orchestration eventID field group when you want to have an event triggered that is not a rule-based event
- For Individual Profile class-based schemas add the ‘Profile test details’ field group to be able to load test profiles for use with Journey Optimizer
- For Journey Optimizer guardrails refer to the following Journey Optimizer Guardrails.
- For Decision Management guardrails refer to the following Decision Management Product Description.
- Requests per second = 5000.
- Latency of response < 250ms.
- Access to edge real-time profile. Only edge projected audiences and profile attributes will be available in the profile.
- If personalization is required in first time experiences, hub will be ideal as the full profile is available. The edge profile must sync from the hub for the first time edge experience. Hence the very first experience from the edge will not include prior uploaded profile data to the hub.
Data ingestion guardrails
- Use the Web or Mobile SDK for deployment on websites and mobile applications to implement Decision Management where the SDK deployed.
- For an API server to server based implementation use the Edge Network Service API for direct server to server implementation of Decision Management.
Adobe Experience Platform
- Configure individual profile, experience event, and multi-entity schemas in Experience Platform, based on customer-supplied data.
- Create datasets in Experience Platform for data to be ingested.
- Add data usage labels in Experience Platform to the dataset for governance.
- Create policies that enforce governance on destinations.
- Create any customer-specific namespaces.
- Add identities to schemas.
- Enable the schemas and datasets for Profile.
- Set up merge policies for differing views of Real-time Customer Profile (optional).
- Create segments for Journey usage.
- Ingest data into Experience Platform using streaming APIs & source connectors.