Data science

Intelligence workbench module allows you create ready-to-use data science models to gain insights from the unified customer data and make informed decisions.

The following data science models are available within Intelligence Workbench.

Algo/ use case family	Algorithm Name	Model Description	Links
Audience selection models	Customer/ Contact Lookalike Account lookalikes	Identifies similar audience for a given seed segment by analyzing profile and behavioral attributes	Lookalike audience selection models
Lead/ Account scoring	Predictive Contact Lead Scoring Predictive Account Scoring	Scores Leads/ Accounts (B2B) based on their likelihood to convert based on engagement patterns	Predictive Contact Lead Scoring models Predictive Account Scoring models
Propensity models	Product Propensity Re-purchase Propensity Churn Propensity Engagement Propensity	Predicts the likelihood of purchasing/ re-purchasing a product(s) OR customer churning over OR engagement based on transactional, engagement or other patterns.	Product Propensity models Repurchase Propensity Models Churn Propensity models Engagement Propensity Models
Campaign/ Channel/ STO  recommendations	Campaign recommender Channel Recommender STO	Identifies the most effective campaign and/or channel based on past interactions	Channel Recommender models Send Time Optimization Campaign Recommender models
Next best recommendationmodels	Next Best Action Next Best Promotion Next Best Offer	Identifies the next best Action/ Promotion/ Offer to present to customers based on past interactions.	Next Best Offer models Next Best Action models Next Best Promotion models
CLTV	Customer Lifetime Value	Estimates the Customer’s value over a period of time based on profile and transaction patterns.	Customer Lifetime Value models
RFM/ Fatigue	Recency Frequency Monetary Fatigue segmentation	RFM models identify personas based on the recency, frequency or monetary value of transactions Fatigue – classifies customers into different levels of fatigue based on profile & engagement	Recency, Frequency and Monetary models Fatigue Segmentation models
Attribution	Campaign Attribution (Revenue & non-Revenue)	Measures effectiveness of campaign by measuring touchpoints that drive revenue and conversion	Campaign Attribution models

Accessing data science models

To access data science models:

1. Click the Oracle icon  in the bottom-right corner to open the navigation menu.

2. Select Intelligence workbench.

The page will display the list of existing models.

Creating data science models

To create data science models,

1. Follow the steps detailed in Creating data science models

2. After creating the model, follow the steps for Running training and scoring jobs

To create data science queries in SQL,

1. Follow the steps in Creating model queries in SQL

To view data science model outputs

1. Follow the steps in Viewing data science models

Next steps

Get more information about each model and learn how to create and use them.

• Customer Lifetime Value models

• Next Best Offer models

• Next Best Action models

• Campaign Attribution models

• Product Propensity models

• Contact lead scoring models

• Account Scoring models

• Recency, Frequency and Monetary models

• Channel Campaign Recommender models

• Fatigue Segmentation models

• Churn Propensity models

• Engagement Propensity models

• Lookalike audience selection models

• Repurchase Propensity models

data science, data science model, analyze data