Propensity Agent
Propensity Agent enables media teams to build predictive models without code, using historical data to score and segment users by likelihood to take action. It supports advertising use cases such as churn prevention, retention, and conversion—producing performance-ready audiences for immediate activation.
How It Works
Step 1: Launch a Predictive Workflow Select from pre-built modeling workflows—like churn, retention, or loyalty—to get started without setup or code.
Step 2: Define Model Parameters Choose your dataset, set the target behavior and thresholds, and customize inputs based on the modeling goal.
Step 3: Preview & Run the Model Review the modeling logic, selected features, and performance preview. Then run the model instantly—no technical expertise required.
Step 4: Review Results & Insights See what’s driving user behavior with transparent outputs and SHAP-based explainability. Adjust confidence thresholds to refine segments.
Step 5: Activate Segments Propensity Agent scores your audience and auto-generates targetable segments—ready to use in downstream workflows or platforms.
Last updated