Customer churn analysis for ABC Communications — AnalystLab Africa Week 1
# Telco Customer Churn Analysis — ABC Communications Ltd.
**AnalystLab Africa — Week 1 Data Analytics Project**
## Business Problem
ABC Communications, a telecom provider, is losing a significant share of its customer base to churn. This project investigates *who* is churning, *why*, and what the company can do about it — using the Telco Customer Churn dataset (7,043 customers, 21 attributes).
Customer churn is costly beyond the lost sale: acquisition costs (marketing, onboarding) are only recovered over a customer's lifetime, so early churn means that investment is never fully recouped — making retention significantly cheaper than reacquisition.
## Key Findings
| Driver | Finding |
|---|---|
| **Contract type** | Month-to-month customers churn at 43%, vs. 11% (one-year) and 3% (two-year) |
| **Payment method** | Electronic check churns at 45% — more than double any other payment method |
| **Online security** | Customers without security add-ons churn at 42%, vs. 15% with it |
| **Senior citizens** | Churn at 42%, nearly double non-seniors (24%) |
| **Tenure** | Strongest numeric predictor of churn (correlation: -0.35); churned customers have a median tenure of 10 months vs. 38 months for retained customers |
Overall baseline churn rate: **27%** (1,869 of 7,043 customers).
## Repository Structure
```
├── data/ Raw dataset (CSV)
├── analysis/ Excel workbook — data cleaning, pivot tables, 9 charts, correlation heatmap
├── report/ Full business report (Word) — business case, analysis, insights, recommendations
├── charts/ Individual chart images (PNG)
└── presentation/ Summary slide deck
```
## Methodology
1. **Data inspection** — checked row/column counts, data types, missing values (found 11 blank `TotalCharges` values tied to new customers, corrected to 0), and duplicates.
2. **Exploratory analysis** — built 9 visualizations (3 bar charts, 2 pie charts, 2 histograms, 1 box plot, 1 correlation heatmap) to examine churn across contr …