otitolayefelix14-pixel/mtn_customer_churn_cleanning

A beginner-to-professional data analytics portfolio project covering Excel, PostgreSQL, and Tableau, built around a real MTN Nigeria customer transaction dataset (974 transactions, 496 unique customers, 35 states). # Customer Churn Analysis for MTN Nigeria A beginner-to-professional data analytics portfolio project covering Excel, PostgreSQL, and Tableau, built around a real MTN Nigeria customer transaction dataset (974 transactions, 496 unique customers, 35 states). ## Project Overview MTN Nigeria's retention team lacked a structured, data-driven view of who churns, what they were using, why they left, and how much revenue is at risk. This project quantifies churn (29.16% overall), identifies the highest-risk states, devices, and plans, and translates the findings into six concrete retention recommendations. ## Tools Used- **Excel** — data cleaning, pivot tables, calculated columns, charts- **PostgreSQL** — relational modeling (4NF-normalized schema), 42 SQL queries, views, indexes- **Tableau** — interactive dashboard with KPI cards, filters, actions, and a story page ## Dataset `mtn_customer_churn.csv` — 974 rows / 17 columns. See the full data dictionary in the Project Guide (`Documentation/Project_Guide.pdf`, Section 3). ## Methodology Ask → Prepare → Process → Analyze → Share → Act. Full lifecycle documented in `Documentation/Project_Guide.pdf`. ## Results | KPI | Value | |---|---| | Total Revenue Analyzed | ₦199,348,200 | | Unique Customers | 496 | | Overall Churn Rate | 29.16% | | Highest-Churn State | Adamawa (61.1%) | | Top Stated Churn Reason | High Call Tariffs (54 customers) | ## Dashboard See `tableau/Dashboard_Screenshots/` for exported images of the live Tableau dashboard, or open `tableau/Dashboard.twbx` in Tableau Desktop / Tableau Public. ## Insights 1. Churn is heavily concentrated geographically (Adamawa, Imo, Akwa Ibom). 2. Pricing (tariffs + data cost) is the single largest addressable churn driver. 3. Satisfaction score and tenure are weak standalone predictors of churn. 4. Revenue exposure from churn is broad, not limited to low-value accounts. ## Recommendations See `reports/Final_Report.pdf`, Section 7.4, for the six prioritized business recommendatio …