Customer Churn Analysis

Business Problem

Churn was being reported as a monthly percentage with no view of who was leaving or why. Leadership could see a 26.54% churn rate but had no way to translate it into a prioritised list of customers worth saving, so retention spend was spread evenly instead of where risk was concentrated.

Our Approach

We rebuilt the reporting layer around customer-level segmentation: Power Query standardisation, engineered contract, tenure and usage features, RFM-style scoring, and a one-page report combining headline KPIs, segment diagnostics, a service-mix heatmap and the contract-length relationship. The design principle was simple — segment first, explain second, recommend third.

Key Findings

• Month-to-month contracts are the highest-risk pattern — short commitment reduces switching friction and creates a natural retention window.
• New customers churn materially more than mature accounts; early lifecycle value reinforcement is the gap.
• Service configuration (fibre vs DSL vs none) explains distinct pockets of elevated churn.
• RFM segments turn 'churn' into a workable target list: high propensity plus low engagement is the urgency zone.

Business Recommendations

• Convert — targeted upgrade and annual-plan offers for month-to-month customers.
• Onboard — a 30/60/90-day journey focused on early value realisation.
• Personalize — use RFM and usage patterns to tailor retention messaging and outreach.
• Measure — track save rate, post-intervention churn and retention ROI by segment.