This functionality is currently in limited testing and not generally available.
Customer AI, as part of Adobe Experience Platform Intelligent Services, provides marketers with the power to generate customer predictions at the individual level.
With the help of influential factors, Customer AI can tell you what a customer is likely to do and why. Additionally, marketers can benefit from Customer AI predictions and insights to personalize customer experiences by serving the most appropriate offers and messaging.
Customer AI relies on individual behavioral data and profile data for propensity scoring. Customer AI is flexible in that it can take in multiple data sources, including Adobe Analytics, Adobe Audience Manager, Consumer Experience Event data and Experience Event data. If you use the Experience Platform source connector to bring in Adobe Audience Manager and Adobe Analytics data, the model automatically picks up the standard event types to train and score the model. If you bring in your own Experience Event dataset without standard event types, any relevant fields will need to be mapped as custom events or profile attributes if you’d like to use it in the model. This can be done in the Customer AI configuration step in Experience Platform.
Customer AI integrates with Customer Journey Analytics (CJA) to the extent that Customer AI-enabled datasets can be leveraged in data views and reporting in CJA. With this integration, you can
Some of the steps are performed in Adobe Experience Platform prior to working with the output in CJA.
Once you have prepared your data and have all your credentials and schemas in place, start by following the Configure a Customer AI Instance guide in Adobe Experience Platform.
In CJA, you can now create one or more connections to Experience Platform datasets that have been instrumented for Customer AI. Each prediction, such as “Likelihood to upgrade account”, equates to one dataset. These datasets appears with the “Customer AI Scores in EE Format – name_of_application” prefix.
Each Customer AI instance has two output datasets if the toggle is turned on to enable scores for CJA during the configuration in Step 1. One output dataset appears in Profile XDM format and one in Experience Event XDM format.
Here is an example of an XDM schema that CJA would bring in as part of an existing or new dataset:
(Note that the example is a profile dataset; the same set of schema object would be part of an Experience Event dataset that CJA would grab. The Experience Event dataset would include timestamps as the score date.) Every customer scored in this model would have a score, a scoreDate, etc. associated with them.
In CJA, you can now proceed to create data views with the dimensions (such as score, score date, probability, and so on) and metrics that were brought in as part of the connection you established.
In CJA Workspace, create a new project and pull in visualizations.
Here is an example of a Workspace project with CAI data that trends propensity scores for a segment of users over time, in a stacked bar chart:
Here is a table that shows reason codes for why a segment has high or low propensity:
This flow diagram shows the entry flow for customer propensity over different scoring runs:
This bar chart shows the distribution of propensity scores:
This Venn diagram shows the propensity overlaps over different scoring runs: