Enterprise AI Administration
Churn Propensity Model
Model Details · Telecom AI Model Registry
Model Details
Churn Propensity Model
Algorithm: XGBoostAPPROVED DESIGN
- Code
MDL-CHURN-PROP- Version
- v3.2
- Owner
- Head of Customer Analytics
- Governance owner
- Model Risk Committee
Governance configuration shown for review and validation. Read-only.
Business purpose
Identify high-risk churn segments for targeted retention
Production-recommended design specification. This record does not mean the model has been trained, fitted, deployed or validated.
Technical specification
- Model family
- Gradient Boosted Decision Trees
- Algorithm
- XGBoost
- Implementation library
- XGBClassifier
- Model type
- Binary Classification
- Explainability method
- SHAP TreeExplainer
Implementation governance
- Governance status
- APPROVED DESIGN
- Implementation status
- MODEL DEVELOPMENT REQUIRED
- Validation status
- VALIDATION REQUIRED
- Lifecycle stage
- Design
- Current performance
- Not yet measured
- Refresh cadence
- Weekly
- Last reviewed
- 14 Sep 2026
- Next review due
- 14 Dec 2026
- Historical sample reference
- AUC 0.84 — earlier sample figure, kept for provenance only; not a validated model result.
Business
- Mission area
- Churn & Retention
- Decision journey
- Churn & Retention
- Target variable
- Churn probability (next 30 days)
- Output
- Segment-level churn risk scores and revenue at risk
Input features
Governed variables, in feature-rank order
- 1.30-day usage decline %Change in usage vs prior 30 days
- 2.Recharge frequencyRecharges per 30 days
- 3.Tenure monthsMonths since activation
- 4.Complaint count 30dComplaints in last 30 days
- 5.network_quality_scoreExperienced network quality for the subscriber's serving sites (0–100).
- 6.Plan typePrepaid, hybrid or postpaid
- 7.Digital activity scoreApp and wallet engagement score
Related Intelligence
Linked governed records
Business rules (2)
Formulas (2)
Source mappings (8)
- CRM · subscription.status_change
- CRM · crm_subscribers.termination_date
- CRM · crm_subscribers.activation_date
- Mediation / CDR · cdr_data_daily.data_mb
- Recharge · recharge_transactions.recharge_date
- Contact Center · customer_complaints.complaint_category
- Billing · billing_service_revenue.amount
- CRM · customer.segment
Simulation models (1)