Creating a churn prediction using a scorecard

2 Tasks

20 mins

Pega Customer Decision Hub '24.2

Pega Customer Decision Hub '24.1 Pega Customer Decision Hub '23 Pega Customer Decision Hub 8.8

Visible to: All users

Beginner

Pega Customer Decision Hub '24.2

English

Scenario

U+ Bank wants to predict and avoid potential customer churn before it happens. When customers leave a bank, the result is costly in terms of lost revenue and acquiring new customers. By detecting customers who might be at risk of churning, the bank can take proactive measures, such as providing incentives or personalized marketing offers, to satisfy them. To allow the bank to identify vulnerable customers, as a Data Scientist, you must create a Prediction based on a scorecard that evaluates churn risk and verify the accuracy of the scorecard for Barbara and Robert.

Use the following credentials to log in to the exercise system:

Role	User name	Password
Data Scientist	DataScientist	rules

Your assignment consists of the following tasks:

Task 1: Create a scorecard

To create a scorecard that predicts churn, edit the scorecard to include the conditions and scores of four customer fields:

1. CreditScore:
   Split the values for CreditScore into five ranges, and then assign scores to each range, as shown in the following table:

   Condition	Score
   <=200	65
   <=400	50
   <=700	35
   <=900	15
   Otherwise	5

2. Age:
   Split the values for Age into five ranges, and then assign a weight of 2 and scores to each range, as shown in the following table:

   Condition	Score
   <=21	90
   <=25	80
   <=30	50
   <=55	20
   Otherwise	10

3. RelationshipLengthDays:
   Split the values for RelationshipLengthDays into four ranges, and then assign a weight of 2 and scores to the ranges, as shown in the following table:

   Condition	Score
   <=180	75
   <=360	60
   <=720	30
   Otherwise	10

4. OwnershipStatus:
   Split the values for OwnershipStatus into three ranges, and then assign scores to the ranges, as shown in the following table:

   Condition	Score
   Rent	25
   Owner	5
   Otherwise	35

Configure the scorecard to output Churn if the churn risk score is equal to or greater than 228 and Loyal if the score is lower than 228.

Task 2: Create a new Prediction

Create a new Prediction in Pega Customer Decision Hub™ to calculate churn risk by using the new scorecard.

 

You must initiate your own Pega instance to complete this Challenge.

Initialization may take up to 5 minutes so please be patient.

• Log In

Challenge Walkthrough

Detailed Tasks

1 Create a scorecard

1. On the exercise system landing page, click Launch Pega Infinity™ to log in to Pega Customer Decision Hub™.
2. Log in as a Data Scientist:
   1. In the User name field, enter DataScientist.
   2. In the Password field, enter rules.
3. In the Work from 1:1 Operations Manager section, click CR-1 to open the change request.
4. In the Scope of changes section, click Create > Decision > Scorecard to create a new scorecard.
5. In the Scorecard Record Configuration section, enter Churn Risk.
6. In the Context section, enter or select UBank-CDH-Data-Customer.
7. Click Create and open to configure the scorecard.
8. In the Predictor expression field, enter or select .CreditScore, and then define the conditions with the values, as shown in the following figure:
   
9. Click the Add icon to add another predictor expression.
10. In the Predictor expression field, enter .Age, and then define the conditions with the values, as shown in the following figure:
    
    Note: Notice the weight value of 2. The weight value indicates the relative importance of a particular predictor in the outcome of the model.

11. Click the Add icon to add another predictor expression.
12. In the Predictor expression field, enter or select .RelationshipLengthDays, and then define the conditions with the values, including the weight, as shown in the following figure:
    
13. Click the Add icon to add another predictor expression.
14. In the Predictor expression field, enter or select .OwnershipStatus, and then define the conditions with the values, as shown in the following figure:
    
15. On the scorecard rule form, click the Results tab to edit the segmentation.
16. In the Result column, in the first row, enter Loyal.
17. In the second row, enter Churn. In the first row, in the Cutoff value field, enter 228.
    
18. Click Save to save the changes to the scorecard.
19. Click Actions > Run to test the scorecard.
20. Enter the test imputs:
    1. .CreditScore: 300
    2. .Age: 45
    3. .RelationshipLengthDays: 230
    4. .OwnershipStatus: Rent
21. Click Run, and inspect the execution details.
    

2 Create a new Prediction

1. In the navigation pane of Customer Decision Hub, click Intelligence > Prediction Studio to open the Prediction Studio landing page.
2. In the upper-right corner, click New to create a Prediction.
3. Ensure that Customer Decision Hub is the active selection, and then click Next.
4. In the Prediction name field, enter Predict Churn Risk.
5. In the Outcome field, select Churn.
6. In the Subject field, select Customer.
   The following figure shows the completed Prediction configuration:
   
7. Click Create.
8. On the Models tab, in the Predict Curn Risk row, click More > Replace Scorecard.
   
9. In the Introduce challenger model dialog box, select Scorecard, and then click Next.
10. In the Scorecard field, enter or select ChurnRisk, and then click Replace.
11. Confirm that the Churn Risk scorecard drives the Prediction, and then click Save.

Confirm your work

1. In the upper-right corner, click Run.
2. In the Run prediction window, select Barbara Data Transform as the data source, and then click Run.
3. Confirm that Barbara scores under the threshold of 228 and that the output predicts that she will remain loyal.
   
4. Repeat step 2 for Robert, and then confirm that he is likely to churn.
   
   Note: When you submit a Prediction for deployment, the application saves the prediction changes to a Change prediction change request. After a review by the Team Lead, a Revision Manager deploys the changes as part of a revision.

---

This Challenge is to practice what you learned in the following Module:

• Creating predictions v5

---

Available in the following mission:

• AI for 1:1 Customer Engagement v5