PEGA Decisioning - Avoiding Redundant Product Offers to Customers

How to make the Offer recommended to the Customer is Personalized? One basic eligibility check is by avoiding Redundant Product Offers to Customers

• If customer already have the Products, don't offer them.
• We can extend the Customer DataModel by defining the 1-1 OR 1-n relationship of the customer owned Products/services. For example, a list of purchases made, a list of accounts owned, a list of family members, and a list of services owned.
• You can define the Relationship between Customer class & Product class
• You can build the Custom DataFlow to populate the customer Profile information composing other Data properties
• Nested Data Type also possible. We need to discuss and agree with the Data Model & compose the same with the respective class object
• When u use the Data Join, the target mapping always be the SR class properties and the source can be customer class properties OR another SR class properties
• Data-Join component allows you to join 2 sets of data and map to the SR class properties
• Filter component Filter can be plain Filter condition (Takes Customer class properties & SR class properties) OR it can be a Proposition Filter.
• @and(booleanA, booleanB) validates to TRUE if both the conditions met
• When u connects with the newly created DB table, create the Table and respective classes through DB Class mapping wizard.
• Go to Data Model -> Classes & Properties -> Database Class Mappings landing page. Click New External Database Table Class Mapping.
• Use DataImport component to Import the Data from External DB Table contents.

To avoid the same product recommendation which customers own already, from PEGA we need to define the data relationship using DataFlow. For example, to cater to a use case where a customer owns multiple products and details of all products are available in a Product master table. So, there is a one-to-one relationship between the ProductHoldings and Product entities.

Now to implement the business case discussed (Filter the products already owned by the customer), you must filter out the products already owned by the customer. For this, you use a strategy pattern that is composed of three components: Data Import, Data Join and Filter, we can filter out the products which customer is holding already.  DataImport+DataJoin+Filter is the Strategy design pattern used to achieve.

 

Update the Strategy to include a component that retrieves Interaction History data for previous phone offers. We then filter out phone offers for customers that have accepted a phone offer in the past. We can of course make the filtering process much more sophisticated by considering phone upgrades, etc. But for the sake of our demonstration, we will keep the strategy simple; anyone who has been made a phone offer within the last 120 days will not be made another phone offer of any kind. IH+GroupBy+Filter is the Strategy pattern used to prevent Redundant offer recommendation in PEGA Strategy logic.

Happy Learning!

PEGA Decisioning - Balancing Business Objectives and customer needs

To Balancing Business Objectives and customer needs, its good to remember the following points.

• Always-On customer brain stands behind the NextBestAction
•  Business Logic is an iterative development. We will develop with the current requirement and based on the customer behavior, we will change the Strategy
• The Key Elements in Implementing the Business Logic is:
• Business Issue * Group. All propositions will be added under any specific Group only
• Propositions - Proposition Catalogue will be maintained
• Prioritization - BL-Strategy
• Use the NBAM components to implement the logic Strategy. ADM is used to predict the customer Propensity and learning the data real-time.
• Predictive models will be used to learn from Existing models
• All Decisioning Strategy business rules are classified under Hard & Soft Rules. Eligibility rules are Hard rules and all Decision Data (Control param), Arbitration components are Soft rules
• Break the whole logic into Small sub strategies logically and process the Propositions in each stage to apply the proposition selection process
• Eligibility Rules are always determined as Hard Rules
• Prioritization component will be used to determine the Top offer at any given time.
• BusinessIssue and Group is something we want to think first before implementing NBAM project
• NBAM project is not a finished product. We want to learn and make the continuous changes to make the better decision 
• Apply BML methodology to keep optimizing the Strategies to drive the business outcome 
• B-Build - Make the Strategy/Logic changes
• M-Measure - Measure the performance of the Strategy 
• L-Learn - Learn the feedback from Customer 
• Repeat the BML method again. 

Happy Learning!

PEGA Decisioning - Decision Strategy execution

 To get the detailed understanding of PEGA Decisioning Strategy execution,

• Understand the purpose of each application component
  
• While populating the proposition itself, pxRank will be updated. Each source component is data source and destination is the final source. All propositions will be recomputed and pxRank will be updated again. Initially pxRank will be computed, the order in which the propositions received.
  
• When u call set property, new property will be set to the proposition
  
• Solid arrow means to copy the data from source to destination. Dotted arrow means only reference the data source. To refer the property from the referencing data source, use the reference component Name.PROPERTYName
  
• Aggregate component always refers the only one data (pxRank=1)
  
• Eligibility rules are Hard rules * Contextweight rules are soft rules
  
• use a Group By component to compute the average margin and a Filter component to filter out propositions with margin less than the average.

Happy Learning!

PEGA Decisioning - Assessing the Impact of the new product Offer

Assessing the Impact of the new product Offer is very critical in Decisioning & marketing prior to launch in Production. We can assess the Impact of the new product Offer using the following steps. 

• Simulation will be used to assess the new product offer. It’s more like what-if analysis
• Use the simulation and analyze the results using VBD to understand the forecast of the product sale. Identify the eligibility rules&context weight to identify the new offers to be proposed.
• Based on the simulation result, update the strategy and control parameters to give the better outcome for new product. It is an iterative process.
• Propositions are anything that can be offered to a customer. This can include advertisements, products, offer bundles and even service actions
• The actual definition of a proposition is “The proposed course of action.”
• Simulation enables you to execute a complex or simple Decisioning strategy using large customer sample both reference and new simulations.
• In Pega Decision Management we create a simulation run using decision data flows which can be run in a batch mode.
• Before adding the new proposition data in Decision Date, understand the structure of the proposition and update the static contents accordingly
• Understand the Strategy and modify the control parameters influence the offer and make the amendments accordingly. After changing the Marketing strategy, execute the product launch simulation strategy referring the proposition distribution report.

PEGA Decisioning - Managing the Decisioning Business Rules (Revision Management)

Has 3 envs in PROD (build-Sandbox, test-sandbox and prod).Revision Management enabled business users to make the change with the limited riles

Post every release, AppOverlay reset will be done in the dev environment (PROD-AUTH). ie, new version of the app overlay will be created. Post creation, completed package will be deployed in business sandbox (Simulation) followed by PROD like regular BAU release.     

• AppOverlay release & package deployment as part of BAU release has to be explored. 
• The process to create a version & export is based on the dynamic case management process.
• As part of Revision management, CRs will be created by Revision Manager and the change will be implemented by Strategy Designer. Revision manager will approve/Reject the CR

New AppOverlay version will be installed like below:

•      The System Architect can use the revision activation landing page which can be found at Decisioning -> Infrastructure -> Revision Management -> Revisions.
• These revisions are primarily used to implement the config parameter to alter the decision application behavior.
• To sendback the CR to the RevisionManager, use SendBack CR option incase any additional RuleSet access required.
• Once you submit the Revision, jar package file will be generated. We need to download them.
• To implement the changes in pre-prod & Prod environment, Decisioning -> Infrastructure -> Revision Management -> Revisions->Import revision.
• Select the Operator from the available list to test the changes for the selected operators before promoting the change.Decisioning -> Infrastructure -> Revision Management -> Revisions->"Activate" the change. Change will be activated to all the users. We can roll-back if needed.
•  After activating the package, we need to logout and login again

Working with the Revision Management is 2 set processes. 1. Create Revision. 2. Create CR. We can have more than 1 CR in a given Revision. Once Revision Submitted, Jar will be generated. If the change is not working, we can revert the change by doing "Roll-back", otherwise activate the Revision.

More details about PEGA Revision management is available here https://community.pega.com/knowledgebase/articles/decision-management-overview/revision-management-decisioning-rules-pega-platform-pega-73 

Enriching the PEGA business application with the Decision Management (Next Best Action)

To Enriching the PEGA business application with the Decision Management (Next Best Action)

·         Open the flow rules which are associated with the Business process. Add Decision flow shape OR Run-Interactionflow shape to introduce the DSM capability in business process applications.

Decision flow Shape ->We can add Decision Table, Tree, Map, Predictive Model and score card model. 

Run-Interaction flow shape ->Add a DSM Strategy and store the results into clipboard memory to refer the results clipboard in the business flow.

Working with Revision Management:

·         Create a new revision, a new change request, include all business rules.

o    On the Dashboard, click New revision. 2 decision flow shares in a decision flow.

·         Decision  - various decision rules can be selected including (DecisionTable, DecisionTtree, predictive model & Scoring model)

2. Interaction Rule Type - to call the Strategy from the flow (it’s a flow shape)

Proposition Management:

•        3 level hierarchy. Business issue (NBA), Group(Treatment), offerID (Treatment identifier)
•        To create the property in a group, open the group->Form Tab, create Property, come back and add the newly created property

Building the Strategy:

•  Pega application is running with the context of the Customer Class.
•  We have to define the Strategy & other required business rules under the customer class. ie. Apply to Class
•  Use the copy/paste option available in the Strategy canvas. Also, use the grid layout to make sure all the components are aligned
•  Expression Builder. How to test? Quick option. Learn articles
•  Properties you are seeing in Expression Builders are Strategy properties. Not AppliesTo class properties
•  Applies to class properties can be used through Primary.
•  Use the Test Canvas made available in the Right side to test the Strategy for a customer/Batch of customers
•  In a Strategy, to refer SR class properties, use dot (.). To refer the AppliesTo class Properties, Refer Primarydot(.)

Making NBA more relevant:

ADM: ADM are Predictors and defined under customer. Decision->Create->AdaptiveModel

• Under Predictors Tab, define the Predictors you want to learn the ADM.Click Add predictors and add the Customer Class Properities
• On the response Tab, define the responses from the customer (Positive/Negative). Values will be retrieved from IH
• Open the ADM landing page by DS->Decisions->ADM. ADM influences the propensity.
• To add a predictive model, Just add the ADM component defining the Customer attributes (Key ones which can be modeled and compared against other data)
• ADM used to predict the customer behavior and make an offer accordingly instead of Customer coming back to us and requesting an offer. Propensity will be stored in .pyPropensity property.
• .pyChannel & pyDirection property should be set for ADM before ADM to invoke
• Churn behavior can be predicted immediately using Predictive analysis and provide an offer. .pyPropensityScore will be used to understand the acceptance from IH date for ADM models defined

Predictive Models:

•          Build the predictive model using the old customer data.
•          Decision->Create->Predictive Models
•          Add Predictive Model component into Strategy. Refer the Predictive model already made available for us> Predictive models will have the details of determining the Churn & Loyal customers based on the data defined.
•         Output of the Predictive model will be stored under .pxSegment.
•          Redefine the Strategy to read the outcome from Predictive model and make the alternative flow accordingly.

IH:

• IH data will help us to understand the past responses and predict the future and helps business not to server the same offer again
• While loading IH, define last X days record to load from IH table rather than all the records
• Define issue & Group to fetch the IH for the channel, otherwise all the IH will be loaded
• Add GroupBy and define the function to set the property from the aggregation
• Dotted lines (.....) refers this is not the main flow. But the data will be referenced from the Strategy component.

Simulation:

• Simulation allows the Decisioning to execute for a large set of customers.
• In the simulation run, a decision strategy is applied against each customer in the population.
• Simulation testing can be used for A/B testing. B4&A4 decisioning change simulation testing. It's more like what-if testing
• VBD will be used to see the Champaign&Challenger simulation results & delta impact
• DataFlow will be used for both old and new Strategy changes against the same CustomerDataSet and the results delta will be referred in the result
• Post the DataFlow execution, open the results/reports Tab to see the results. "Proposition Distribution" will give the details of the proposition offered to the customers
•  VBD will be used to compare the results of 2 different Datasets and see the outcome change & delta of the result. 
• To create a simulation Run, create a DataFLow. You can open the Dataflow under Simulation, DataFLow Runs to see the components
• DataFlow usually will have CustomerData-DataSet, followed by Strategy and results will be stored in result DataSet (VBD/DB). To write the results to VBD, use the DataSet Type as VBD
• Goto manage->DataSources to explore the VBD results
• Channel is a dimension in Visual Business Director in Y Axis.

AppOverlay & Decision Manager Portal configuration:

• Decision Manager portal mainly used for PROD Simulation & PROD. Use an applicationOverlay for development
• Business sandbox & AppOverlay env & Direct deploy for Decision Data update
•  Need to configure ADM monitoring, IH, VBD

Building An app overlay application:

• Create New App Overlay wizard and create an app
• Define the accessgroup & list of available rules to expose to AppOverlay to make the changes. Primarily on Decision data.
• Default application Overlay will be created after Landing page. The name of the overlay starts with ‘RTC-‘.
• Start the Application Overlay wizard via the Decisioning > Infrastructure > Revision Management->Application Overlays landing page.
• Define the AppOverlay application, Revision ruleset, accessgroup & the Rules to be included as part of AppOverlay
• Select the rules to make available for the revision Management, select the rules and click "Include for Revision Management"
• Review the AppOverlay and credit the AppOverlay application
• Setup new accessgroups & their roles and the required operator IDs to use.