Headline

Month-to-month contracts churn at 43.5% — 15× higher than two-year contracts at 2.9% — putting $38.4M in monthly recurring revenue at immediate risk.

Purpose

This analysis examines churn patterns across 5,000 active SaaS subscriptions to identify which customer segments and contract types face the highest risk of cancellation. The objective is to quantify revenue exposure and pinpoint the strongest predictors of churn so retention efforts can be prioritized where they matter most.

Key Findings

Overall churn rate: 27.1% (1,354 of 5,000 subscriptions have churned) — well above the typical SaaS benchmark of 3–7% monthly, indicating a significant retention problem
Month-to-month churn: 43.5% vs. Two-year churn: 2.9% — a 15-fold difference showing contract length is the dominant churn driver
MRR at risk: $47.9K (21.4% of active revenue) is exposed to churn in the next period, with month-to-month contracts accounting for $38.4K (39.8% of their revenue)
Median tenure at churn: 10 months — customers are leaving before establishing long-term loyalty
At-risk accounts identified: 20 high-probability churners, all on month-to-month contracts with electronic check payments, averaging 78.2% churn probability
Top protective feature: Two-year plan reduces churn odds by 74% (odds ratio 0.26, p<0.001); fiber optic service increases churn odds by 227% (odds ratio 3.27, p<0.001)

Interpretation

The data reveals a tiered churn structure driven primarily by contract commitment. Month-to-month customers face structural instability — they can leave with minimal friction, and 43.5% do. Two-year customers are locked in and rarely churn (2.9%). The survival curves show month-to-month subscribers experience steep early attrition (90% survival at month 1, declining to 80% by month 5), while two-year subscribers remain stable throughout the observation window. The $47.9K MRR at risk represents real revenue leakage; if current churn rates persist, the company will lose approximately $100.8K in annual revenue from already-churned customers alone. The 20 flagged accounts represent immediate intervention opportunities — all are month-to-month, all use electronic check payments (a known friction point), and all show 70–84% predicted churn probability.

Context

No data quality issues were detected; all 5,000 rows were retained. The analysis includes 8 predictors and uses logistic regression with odds ratios for interpretability. The feature importance table shows two-year plans and one-year plans as the strongest protective factors, while fiber optic service and electronic check payments increase risk. Survival analysis confirms the protective effect of longer contracts across the full 72-month observation window.