Telecom Customer Churn Analysis

A cross-operator study that traces subscriber attrition back to the network itself, bridging RF KPIs and customer-experience signals to explain why people actually leave.

The brief.

Pakistani telecom operators were treating churn as a marketing problem when most of the signal was technical. The goal was to combine network KPIs with customer demographics to show product and engineering teams where the real drivers sat.

How it was built.

1. STEP 01Consolidated subscriber, billing, and network-quality datasets across four providers in SQL.
2. STEP 02Segmented users by age, tariff, region, prepaid vs postpaid, and reported satisfaction.
3. STEP 03Tested correlations between technical KPIs (signal strength, voice quality, dropped calls) and 90-day churn.
4. STEP 04Built provider-comparison dashboards in Power BI with per-segment drill-downs.

What it moved.

• →Signal strength explained 58% of churn variance; voice quality contributed another 45%.
• →Zong showed an 18% lower churn rate than peer operators across comparable segments.
• →Prepaid users churned 8% more than postpaid, concentrated in two age brackets.
• →Findings reframed the retention conversation from discounts to network investment.

Tools used.