In this article
Predictive Analytics is most effective when the views are used together. Each view answers a different question and builds on the previous step.
A typical workflow might be:
1. Start with the full population (Summary View)
Use the Summary view to understand how predicted outcomes are distributed.
Identify a group of members with higher likelihood that is large enough to act on.
2. Validate the group (Comparison View)
Use the Comparison view to confirm that the selected group differs meaningfully from the overall population or from other groups.
3. Understand associated patterns (Prediction Drivers View)
Use the Prediction Drivers view to understand what factors are most associated with that group.
These patterns provide context; they are not causes.
4. Review the members (Member List)
Open the Member List to examine the specific members in the cohort.
Use this view to validate the group and refine it if needed.
5. Iterate
This process is iterative. You may move between views to refine the cohort, validate your selection, and confirm that it remains meaningful and actionable.
6. Take action
Use the cohort to support outreach, prioritization, or quality improvement efforts.