Scoring Framework Predictions
Pitch the right products to customers and boost your revenue by analyzing customers’ likelihood of purchasing financial assets. Decrease customer attrition by taking steps based on the customer’s likelihood of churning. Sell an asset to an interested customer based on predictions and increase your customer’s assets under management. Get prediction scores based on the data of your accounts and contacts.
Required Editions
| Available in: Enterprise, Performance, and Unlimited Editions of Lightning Experience where Financial Services Cloud, CRM Analytics for Financial Services Cloud, AI Accelerator, and Scoring Framework are enabled |
To get predictions about customers’ likelihood of purchasing financial assets, likelihood of churning, or likelihood of adding assets, create a CRM Analytics template configuration in Scoring Framework with Prediction Scores (Financial Services Cloud) as the template configuration type.
- If you want to define a predefined target variable to get predictions about customers’ likelihood of purchasing financial assets, select Accounts Have Newly Purchased Financial Accounts (Likelihood to Purchase) or Accounts Have Associated Opportunities That Are Closed (Likelihood to Purchase).
- If you want to define a predefined target variable to get predictions about customers’ likelihood of churning, select Account Status Has Changed to Inactive (Likelihood to Churn).
- If you want to define a predefined target variable to get predictions about customers’ likelihood of adding assets, select Assets Under Management Have Increased for Accounts (Likelihood to Add Assets).
To get predictions based on the data of your accounts and contacts, create a CRM Analytics template configuration in Scoring Framework with Prediction Scores for Accounts or Contacts (Financial Services Cloud) as the template configuration type and define a custom target variable.
The apps created based on the template configuration contain these preconfigured recipes.
| Recipe | Description | Output |
|---|---|---|
| Get Data to Generate Example Dataset | The recipe creates a dataset that’s used to generate an example dataset by evaluating account snapshot data, financial account snapshot data, tasks, events, cases, and data from the object selected to train the model. | Example Dataset |
| Get Example Dataset | The recipe creates a dataset that Einstein learns from by evaluating account snapshot data, financial account snapshot data, the dataset created by the Get Data to Generate Example Dataset recipe, and data from the object selected to train the model. | Example Dataset |
| Prediction Dataset | The recipe creates a dataset based on which Einstein generates predictions by evaluating financial accounts, financial account charges and fees, leads, tasks, cases, and data from the object that you want to get predictions for. | Prediction Dataset |
| Get Financial Services Cloud Predictions | The recipe generates predictions and the top three factors that possibly impact the predictions by evaluating the prediction dataset. | Writes prediction details and the top three predictors to the writeback object’s field that’s selected when setting up the template configuration in Scoring Framework. |
You can modify the recipes in these scenarios.
- Your schema deviates from the Financial Services Cloud schema.
- A custom field of an existing entity changes.
- The storage of feature data changes from an existing entity to a custom entity.
- The data doesn’t load properly.
- The app stops working because of incorrect data values.
