Churn & NPS Driver Tree Analytics
for a Southeast Asia Telco
Industry:
Telecommunications
Service Pillar: Data &
AI → Data & Analytics
Executive
Summary
A leading telecom operator in Southeast Asia was facing high customer
churn and declining Net Promoter Scores (NPS) due to service quality
issues and fragmented customer insights. Munvel Business Solutions
delivered a predictive analytics solution using driver tree modeling and
advanced visualization to identify root causes of churn and improve
customer retention strategies.
Problem Overview
Rising churn rates and
limited visibility into customer sentiment drivers
Data scattered across CRM,
CDRs, ticketing, and QoS systems
No unified analytics layer to
correlate service performance with customer behavior
Reactive customer engagement
rather than predictive retention campaigns
Our Solution
Unified customer and service
data into a centralized analytics model
Applied SHAP-based driver
tree analysis to reveal top churn influencers
Implemented predictive churn
and satisfaction models using historical data
Integrated insights into CRM
for proactive retention offers and service improvements
Technologies Used
Python, Power BI, Azure Synapse, Scikit-learn, SHAP, SQL
Server
Key Outcomes
01
14% reduction in churn within two quarters through targeted
retention actions
02
9-point increase in NPS across high-value segments
03
Actionable insights into service-level issues driving
dissatisfaction
04
Empowered marketing and service teams with self-service analytics
dashboards