Generating Next Best Actions for omni-channel execution and engagement

From PegaWiki
Generating Next Best Actions for omnichannel execution and engagement / This is the approved revision of this page, as well as being the most recent.
Jump to navigation Jump to search

Generating Next Best Actions for omni-channel execution and engagement

Description Integration of Pega Customer Decision Hub with Adobe Real-time Profile in Experience Platform
Version as of 8.5
Application Pega Customer Decision Hub
Capability/Industry Area Partners and Integrations



Integration of Pega Customer Decision Hub™ with Adobe Real-time Profile in Experience Platform[edit]

Applicable Solutions and versions[edit]

  • Pega Customer Decision Hub (CDH) 8.5 and above
  • Adobe Experience Platform (AEP) Real-time Profile (CDP) and (by extension, to) Adobe Execution solutions such as Adobe Journey Optimizer (AJO)

Integration Type[edit]

  • File based / batch
  • Periodic / scheduled

Introduction[edit]

Reasoning and goal of integration[edit]

Larger enterprise customers often use multiple software solutions, often from different software vendors, to solve specific business problems. About 30% of Pega Customer Decision Hub customers use Adobe solutions. Many Pega and Adobe customers have requested an integration between the two solutions, where Pega delivers next best actions to Adobe Experience Platform (AEP) for implementation and delivery to customer audiences. The actual implementation and channel can be done through Adobe or through other third-party solutions with which AEP is integrated, which is an integration and under the responsibility of / on the Adobe side.


This article describes a best practice approach for the integration of Pega Customer Decision Hub with AEP, to deliver up to date next best actions for customer engagement as described above.

Leveraging Pega Customer Decision Hub - scheduled batch next best actions, can be sent to AEP to update the AEP Real-time Profile. This article describes the integration of the scheduled batch next best actions to AEP.

As a best practice, Pega Customer Decision Hub will be the center for customer decisioning. It connects the decision-making process for outbound messages to customer data used for all other channels to ultimately deliver the next best action, message, or conversation to the customer. In doing so, the full customer context, like browsing behavior on digital properties, previous service interactions, and customer propensity, are balanced with the strategic business goals to optimize every customer interaction regardless of channel.

From a 1:1 standpoint, Pega Customer Decision Hub leverages the latest available data to evaluate the best message to send to each customer, rather than creating a series of segments based on pre-defined business criteria. Customer-specific determination of the best message enables the business to develop a series of journey actions or offers (named Actions), and manage them with strategies and configuration of eligibility criteria, contact policies, and constraints from within the Next-Best-Action Designer. This approach enables dynamic orchestration of fluid customer journeys while establishing guardrails and controls, to ensure the message is not arbitrarily confused or disjointed.

In this model, a decision is made in real-time based on business criteria as translated into a set of strategies that evaluate customer eligibility, combined with customer relevance using adaptive learning models. Pega Customer Decision Hub also recommends the best collection of experiences and content (named Treatments) inside the message, based on the business criteria.

AEP can update and segment customer profiles because of the individuals’ specific Pega next best actions. Marketers can define actions to be taken once they receive next best actions and update the customer profiles. For example, outbound email messages are compiled by selecting a template that manages the overall structure and layout of the message, and then are populated by specific experiences and content with the next best action determined by Pega. Depending on Adobe and other 3rd party solutions integrated with AEP, one or more use cases can be implemented.

Integration Details[edit]

Personas[edit]

Marketer / NBA Designer, SW Engineer, Technical Architect, Security specialist / Admin

Use Cases[edit]

“As a marketer

  • I want to benefit from Pega Customer Decision Hub to define the right set of next best actions for my customers, and have the next best actions populated in the Adobe Experience Platform Customer Profile
  • so, I can engage with my customers with the right action, increasing my chances of relevancy, retention, and sales.”

Pega Customer Decision Hub sends scheduled batch next best actions to update Adobe Experience Platform (AEP) Real-time Customer Profiles by the following means:

  1. Pega Customer Decision Hub sends batch Customer next best actions to Cloud store on a scheduled basis.
  2. AEP cloud store source connector is configured to ingest next best actions to AEP Customer Profiles
  3. AEP leverages other Adobe or 3rd party solutions to execute on the next best actions as received by Pega Customer Decision Hub

Visual overview of integration[edit]

To schedule batch sends of next best actions to Adobe Experience Platform requires a standard Adobe Experience Data Model (XDM) profile schema, data set, and file structure to be established and invoked between Pega Customer Decision Hub and Adobe Experience Platform. The transmitted decision data can then enrich customer profiles in Adobe Experience Platform.

Diagram showing the integration process between Pega Customer Decision Hub and Amazon Experience Platform.
Overview of the integration process


The following diagram shows the core components of the AEP, providing more detailed insights into the Real-time Profile

Diagram showing core Amazon Experience Platform Components
Core AEP components

The following diagram shows the Real-time Profile vs the CDP. The CDP can be considered Adobe’s segment syndication app.

Diagram showing integrations between Adope Experience Platform and Pega Customer Decision Hub

Description of the process[edit]

Step 1: Schedule Next-Best-Action Designer to output the next-best-action results into AWS S3 Bucket.

Personas: Marketer / NBA Designer, Technical Architect

  1. Create Output Template rules to define the content to write to an external source as part of making an action.
    • Configure to create a new file template for file exports that uses Amazon S3 as the repository.
    • Configure the output content for this template from Next Best Action data attributes that will later be used to enrich customer profiles in Adobe Experience Platform.
    • Configure to finalize the template at the end of each outbound run.
  2. Configure one or more Actions in Next-Best-Action Designer to use the newly created file template in a generic flow shape.
  3. Configure the Primary Schedule that outputs next best action at a scheduled time, and the interval for outbound channels. If the scheduled next-best-action run picks up any of the configured actions , it will result in next-best-action data attributes output into S3 repository. If you use Pega Cloud, you need to automate secure file transfer protocol to move the file from Pega Cloud instance to client AWS S3 bucket as and when data is available from Pega Customer Decision Hub. 
Step 2: Configure Amazon S3 Source Connector to ingest next best actions to be used in Adobe Real-time Customer Profiles

Personas: SW Engineer, Technical Architect, Security specialist / Admin

  1. Create a new XDM Individual Profile Schema in AEP to represent the next-best-action fields that enrich AEP customer profiles. To enrich customer profiles, the schema must be compatible for use in Real-time Customer Profile. A compatible schema satisfies the following requirements:
    • The schema has at least one attribute specified as an identity property.
    • The schema has an identity property defined as the primary identity.
    • A mapping within the data flow exists wherein the primary identity is a target attribute.
  2. Create a new dataset from previously created XDM profile schema for next best actions.
  3. Create a new Amazon S3 account connection in Adobe Experience Platform Amazon S3 Source.
  4. Configure a dataflow for Amazon S3 source connection:
    • Map data fields in CSV file to Adobe XDM profile schema for next best actions.
    • Ingest data into newly created dataset.
    • Schedule ingestion runs.
Step 3: Customer profiles are updated with next best actions

Personas: SW Engineer, Technical Architect, Security specialist / Admin

  1. Inbound data from Amazon S3 source connector enriches and populates Real-time Customer Profiles with the Next Best Actions

Step 4: Campaign run results in an email with next best action sent to customers.

Personas: Marketer / NBA Designer

  1. Adobe applications can access the Real-time Customer Profile in AEP to get next best actions from customer profiles and target them in outbound channels such as email or SMS (for example, using Adobe Journey Optimizer, as described in the following use case).

Implementation Details[edit]

This is an example of an outbound communication driven by Pega Customer Decision Hub next best actions, leveraging the Adobe Experience Platform Real-time Customer Profiles and Adobe Journey Optimizer for external communication:

A U+ Bank marketer wants to utilize Next Best Action capabilities to identify customers with the highest likelihood to respond to targeted emails. The likelihood will be determined using the most up-to-date customer profiles and customers behavior.

The marketer has configured the next best action in Pega Customer Decision Hub that will run daily. The next best action is configured to use propensity, context weighting, action value, and business levers to prioritize offers. The next best action will determine if Adobe Journey Optimizer delivers an offer to a customer by using email channel or, if targeted previously, whether it should no longer be considered a next best action.

A data flow configured by a technical user will enrich next best actions for customer profiles in Adobe Experience Platform, that Adobe Journey Optimizer can access.

The marketer has created a credit card campaign email and an associated email campaign in Adobe Journey Optimizer. The campaign will access the customer profile with the next best actions to determine who should receive the targeted email.

Process/Steps to achieve objective[edit]

Configuring email delivery based on Pega Customer Decision Hub next best actions:[edit]


Note: The following examples and instructions assume the installation of the Pega Customer Decision Hub sample application. The recipient or customer data used in Pega Customer Decision Hub and Adobe Experience Platform are the same sources. The IDs of the actions should be the same as delivery codes used in Adobe journey Optimizer.

  1. Configure an Amazon S3 repository in Dev Studio, if one does not exist.
    Dev Studio configuration of Amazon S3 repository type
    Configuring Amazon S3 repository
  2. In Next-Best-Action Designer, create a new file template for file exports that uses Amazon S3 as the repository. Configure the template to finalize at the end of each outbound run, and select delimiter type as comma separated. As a best practice, configure the file name to include the run ID, the data and time of the template finalization, or both. This prevents the output file from being overwritten by subsequent runs.
    Definition tab of template configuration window showing how to configure file export to use Amazon S3.
    Template for Amazon S3 file export
    • Configure NBAToEnrichAdobeCDPTemplate file template.
      Configuring NBAToEnrichAdobeCDPTemplate
      Configuring NBAToEnrichAdobeCDPTemplate
    • Configure next best fields that are relevant to enrich Customer profile in AEP
Next best action fields relevant to enrich customer profile in AEP
Name Content
RecipientId .Customer.CustomerId
Direction .OfferData.pyDirection
Channel .OfferData.pyChannel
ActionId .OfferData.pyName
ActionName .OfferData.pyLabel
InteractionID .OfferData.pxinteractionID
Treatment .OfferData.pyTreatment
Identifier .OfferData.pxIdentifier
Priority .OfferData.Priority
Issue .OfferData.pyIssue
Group .OfferData.pyGroup
SubjectID .OfferData.pySubjectID


3. Configure all actions output that need to be shared with Adobe Experience Platform.

As a best practice, configure actions to use same dynamic template that reference the file template for file exports.

Example: configure an action Increase your Credit line v2 in Next-Best-Action Designer to use a dynamic template Services_ChannelAction_Template_V2 that uses the newly created file template NBAToEnrichAdobeCDPTemplate in a generic flow shape.

Dynamic template showing an example of how to configure an action flow.
An example of an action flow configuration.
Action “Increase your Credit line v2”
Action “Increase your Credit line v2”
Dynamic template “Services_ChannelAction_Template_V2” used in Action.
Dynamic template “Services_ChannelAction_Template_V2” used in Action


Dynamic template uses File template “NBAToEnrichAdobeCDPTemplate” in a generic shape.



4. Configure the primary schedule that outputs the next best action at a scheduled time, and the interval for outbound channels.

The following output file is created in S3 (this is an example with mocked up data for the email channel):

Amazon S3 xls output file sample.
S3 output file sample.


Note: Recipient ID and Customer ID will be the same if both, Pega Decision Hub and Adobe Experience Platform, use the same source.

5. Create XDM Individual Profile Schema in Adobe Experience Platform to represent the next best actions selected by Pega Customer Decision Hub.

The schema will look like the example below:

XDM CDH Next Best Action Schema structure in AEP.
Example Adobe Experience Platform XDM Individual Profile schema
Selecting a schema for a new dataset in Adobe Experience Platform.
Creating CDH Next Best Action Dataset from schema.


5. Create a new dataset CDH Next Best Action Dataset, from previously created XDM schema.

6. Configure Amazon S3 Source Connector in AEP to ingest next best actions to enrich customer profiles in Real-time Profile.

Create a new dataflow by selecting the first table within any selected folder structure, or single file selection in S3 bucket, and map the source data to target fields in the data set.

Creating a new dataflow by mapping source data.
Configuring Amazon S3 Source Connector in AEP.

7. Configure dataflow to ingest next best action files from Amazon S3 Source bucket “pega-cdh-next-best-actions”

Mapping source field to target dataset fields in Amazon S3 "Add data" tab.
Mapping source field to target dataset fields.


8. Create a campaign in Adobe Experience Platform Journey Optimizer than can perform one of the options based on the NBA field in the profile.

  • The workflow will deliver an email that represents an action determined by Pega Customer Decision Hub. This email is only sent to the recipients recommended to receive the email based on the actions selected by Pega Customer Decision Hub. Use the attribute action ID to determine if the email associated with the action should be sent or not.
  • Alter a journey path for the customer.
  • Alter the action/message payload for the customer.

More information and sources[edit]