strategist-predictions

What problem does it solve?

Predictive modeling for churn and event likelihood helps teams identify users who are likely to leave or convert before those outcomes occur, enabling timely, data-driven interventions that reduce churn and improve conversion efficiency.

Core Features & Use Cases

• Model Configuration & Training: Guidance for defining churn/event windows, choosing prediction audiences, and understanding training constraints and limits.
• Segmentation & Targeting: Actionable patterns for using Churn Risk Score, Churn Category, Event Likelihood Score, and Likelihood Category in segments, campaigns, and Canvas flows.
• Analytics & Troubleshooting: How to read Prediction Analytics, choose score thresholds, interpret prediction quality, and resolve common training or data-quality errors.
• Use Cases: Tiered retention Canvas for re-engagement of at-risk users and likelihood-based conversion campaigns for targeted upsell or purchase prompts.

Quick Start

Segment users by Churn Risk Score > 70 and build a Canvas re-engagement flow targeting the high-risk cohort.