Enhancements and Features

Audience exploration allows you to explore your potential audience from all entities. You can now use the predefined arrays of objects, e.g., emails, phone numbers, and addresses, as a filter option in Audience.

When building an audience by using arrays of objects or their attributes, a pop-up appears. You can give a name for the filter and then add attributes from the array. The filter only checks the conditions once. Conditions can be as complex as wanted by using more than one attribute.

For additional details refer to: Audience Exploration

When adding the "Fulfills condition" decision to a CX Flow, a "Create Condition Step" popup appears. On the left side schemas, you can use attributes from the relationship schemas to build the condition criteria.

The relationship attributes from the relationship schemas can also be used in CX Journeys when you define the entry condition for a milestone in the journey.

For additional details refer to: Triggers

A new date type indicator has been added. This indicator can take a date from the profile or groups and calculate re-occurrences for that date.

You can define when the next occurrences happen and how many times. Once an event is created with the indicator, the next occurrence is shown in the calculated indicators of the profile.

This provides additional capabilities for contextual segmentation based on the next reoccurrence, trigger CX flows, and activate audiences. This also enhances our segmentation capabilities by providing a 'moving window' to capture only those in the relevant dates configured.

For additional details refer to: Creating a Reoccurring Date Indicator

The Multi-business capability allows each business unit to be divided into business areas. This can allow your users access to specific customers and configurations in CX flows, CX journeys, and Audiences. Having your business unit divided into areas can make your users focused on their prospective customers and limit the configuration to their target customers.

Only administrators can set up areas. It allows administrators to operate with multiple entities that require access to the CDP console and to apply business areas within a business unit so that not all data and configurations are exposed to all CDP users. A business area can reflect market, brand, or other qualifications according to operational requirements.

Areas are defined by an attribute chosen from the Customer Profile schema and/or Activity schema, e.g., Countries. If even one profile has this value, an area is suggested for activation. You can also create areas if no profiles have the value, such as testing purposes. If any areas are removed from the active list and there are no profiles in that area, the area is deleted. Otherwise, it will return to the suggested list.

For additional details, refer to: Multi Business Areas

When you build a CX flow with the "Customer Created or Updated" decision or trigger, you can choose a specific profile attribute or group attribute. So, only when the specified attribute is created or updated will the CX flow be triggered.

For additional details refer to: Triggers

Data Retention allows you to delete profiles according to set conditions you create and is only applied if the value exists. Activities related to the profiles are also deleted, and if the activity is associated with other profiles or groups, it won't be deleted. Once a rule is made, it can take up to 24 hours to see it executed. If the profile has more than one rule that applies to it, the rule with the longest retention period is the rule that is followed.

When you first use Data Retention, Simulation mode is enabled. Simulation mode lets you use Data Retention, but the shown deleted profiles are only simulated. The profiles aren't deleted in Simulation mode. Once you switch to Active mode, the total number of simulated deleted profiles is deleted in the next deletion period.

After going into Active mode, you can stop Data Retention by Deactivating it. You can still set up rules when Deactivated, but the profiles won't be deleted until Data Retention is in Active mode. If profiles weren't deleted due to data retention being inactive, the profiles are deleted in the next deletion period (shown as now in the graph).

Data Retention allows you to respect your customers' privacy and meet privacy regulations. In addition, you'll be able to cut costs by keeping data of active and engaged customers.

For additional details refer to: Data Retention

Create dashboards that provide analytical insights into your customer data and predictive indicators for customer activities, which enable you to make predictions and view the results of business actions.

The AI Workbench provides an entry point for business users, marketing analysts, and data scientists.

It allows the ability to slice and dice data easily. For example, you can easily create dashboards with different insights for specific populations, segments, etc. As well as update and configure dashboards, view out-of-the-box analytical insights, and more.

In addition, it empowers users to train models by using out-of-the-box and customizable models and artificial intelligence algorithms. For example, it enables users to predict Churn probability and risk classification for different targeted populations and predict Customers' Lifetime Value for specific topics and time frames.

Configure models, also known as predictive indicators, to make predictions and view the results.

A model is a mathematical representation of objects and their relationships. More specifically, in the AI Workbench, a model is an algorithm that runs on data and finds patterns to make predictions. For example, you can create a model that predicts the probability that a VIP customer will churn.

Model Types

You can design models for the following model types:

• Churn (Profile): Predict churn probabilities for profiles.
• Churn (Group): Predict churn probabilities for groups.
• Customer Lifetime Value (Profile): Predict customer lifetime value for profiles.
• Customer Lifetime Value (Group): Predict customer lifetime value for groups.

The SAP Customer Data Platform AI Workbench offers some out-of-the-box churn insights.

The following churn insights are provided out-of-the-box.

Churn Probability by Risk Level

The Churn Probability By Risk Level insight displays a pie chart representing the churn risk probability values (high/medium/low distribution). You can use this insight to evaluate the current general churn risk level.

Churn Risk Over Time

The Churn Risk Over Time insight displays a column bar chart comparing the churn risk values (high/medium/low) within a specific time. You can use this insight to evaluate general churn risk trends.

Churn Risk by Segment

The Churn Risk By Segment insight displays a column base chart representing a distribution of the churn risk probability values (high/medium/low) according to segments. You can use this insight to identify segments at a specific churn risk level.

Churn Probability Distribution

The Churn Probability Distribution insight displays a histogram representing the distribution of amount of profiles according to their churn probability. You can use this insight to evaluate how many profiles are at risk of churning.

The SAP Customer Data Platform AI Workbench offers out-of-the-box customer lifetime value insights.

The following customer lifetime value insights are provided out-of-the-box.

Average Customer Lifetime Value (CLV)

The Average Customer Lifetime Value insight displays a pie chart representing the high-level distribution between CLV levels and the average value of each level.

Customer Lifetime Value (CLV) Profiles

The Customer Lifetime Value Profiles insight displays a pie chart representing the distribution between CLV profiles.

Customer Lifetime Value (CLV) Profiles Over Time

The Customer Lifetime Value Profiles Over Time insight displays a chart representing the trend of profiles level values over time.

Average Customer Lifetime Value (CLV) Over Time

The Average Customer Lifetime Value Over Time insight displays a chart representing the trend of average CLV per level over time.