Marketing Decision Support System for Customer Retention ยท Project 4. Two validated boundaries decide who lands on the retention target list; the underlying SVM churn-risk and SVR predicted-tenure scores stay hidden and only drive the priority tiering โ exactly as in the original app.
Access Role
Controls which export options are available.
Target Criteria
Active customers at or below this tenure are considered.
Customers whose SVR-projected tenure is at or below this value are considered. The list recalculates immediately.
Why this Target Boundary?
Avg satisfactionChurn rate
Across all customers (active + churned): churn is highest for the newest customers and falls as tenure grows, while satisfaction rises. The dashed line is your Target Boundary โ the crossover zone where risk concentrates.
Customers Matching Current Target Criteria: 0
Customer ID
Gender
Current Tenure
Contract Type
Monthly Charges
Payment Method
Services
Priority
Select a customer row above to view full details.
Export
Service Satisfaction and Churn Rate by Tenure
Average Service Satisfaction (left) ยท Churn Rate (right) vs Tenure (months)
Average Service SatisfactionChurn Rate
Across all customers (active and churned), churn rate is highest for the newest customers and falls steadily as tenure grows, while satisfaction rises. The Target Boundary marks that crossover โ it's why the target list can include very-low-tenure customers: that's precisely where churn risk concentrates. (The Retention Boundary is applied separately, to each customer's SVR-predicted tenure โ it isn't a second point on this chart's tenure axis.)
This chart is a descriptive summary โ average Service Satisfaction and Churn Rate grouped by tenure across all customers โ used here to justify the boundary choice. The SVM and SVR models operate independently of it: each predicts from an individual customer's full profile rather than from this aggregate trend.