Executive Dashboard
E-Commerce Churn Overview
High-level summary of churn behaviour across a 4,000-customer e-commerce cohort. Analysing purchase patterns, loyalty engagement, and behavioural signals to surface actionable retention insights.
↑ Risk
7.2%
Overall Churn Rate
289 of 4,000 customers lost
Stable
3,711
Retained Customers
92.8% retention rate
Avg AOV
$73.40
Avg Order Value
$831 average lifetime spend
Per cust.
8.3
Avg Orders / Customer
Median tenure: 10.4 months
Key Finding: Recency Is the Dominant Churn SignalCustomers inactive for 90+ days account for 64% of all churned accounts. Days Since Last Order is the #1 predictor (24.4% model importance). Automated win-back flows triggering at 60-day inactivity is the highest-ROI retention action available.
Churn Distribution
Churned vs. Retained — Overall Split
Proportional breakdown of the 4,000-customer cohort. The 7.2% churn rate aligns with the industry benchmark of 5–10% for e-commerce platforms with active loyalty programmes.
Churn Rate by Customer Tenure
New customers (first 90 days) churn at the highest rate. Risk drops sharply after year one, confirming that onboarding quality is the primary early-lifecycle retention lever.
Behavioural Signals
Days Since Last Order — Churn Escalation
Churn risk escalates rapidly beyond 90 days of inactivity. Past 180 days, over 30% of customers churn. Win-back sequences should trigger at 30, 60, and 90-day inactivity marks.
Order Value Distribution — Churned vs. Retained
Low-AOV customers (under $50) churn disproportionately, reflecting price-sensitive behaviour. Customers spending $200+ per order show strong loyalty, validating premium segment investment.
24.4%
Top Feature Importance
74.8%
Model ROC-AUC
4,000
Customers Analysed
7.2%
Overall Churn Rate
92.8%
Retention Rate
Feature Analysis
What Drives E-Commerce Churn?
Feature importance scores derived from the trained Random Forest model using Gini impurity. Each score quantifies how much a given customer attribute contributes to the churn prediction decision, directly informing where to invest retention resources.
How to Read This PageImportance % = the share of all Random Forest decision splits attributed to that feature. The top 5 features account for 72.6% of all predictions. Features tagged "High" priority represent the clearest opportunities for immediate retention intervention.
All 15 Features Ranked
Feature Importance — Full Ranking
All 15 model features ranked by Gini importance. Red bars indicate critical predictors, amber moderate, grey minor refinements to the model's decisions.
Importance Distribution — Horizontal Bar
Visual breakdown of importance scores. The top 6 features collectively drive over 75% of all churn predictions made by the model.
Business Insights & Recommended Actions
Feature Impact Table — All 15 Features
Every feature documented with its CSV column name, importance score, priority level, a plain-language business interpretation, and a specific actionable recommendation. Use this as the foundation for your retention roadmap.
Loyalty Segmentation
Churn by Loyalty Tier
Loyalty programme membership is a powerful protective factor against churn. Customers enrolled in even the entry Silver tier show significantly lower attrition, validating investment in tiered rewards, points programmes, and exclusive member benefits.
Priority: Convert No-Loyalty Customers to SilverCustomers with no loyalty membership represent 50% of the base and churn at 10.4% — over 4× the rate of Gold members. Enrolling even a fraction of this segment into Silver would materially reduce the overall company churn rate.
Churn Rate by Loyalty Tier
Churn risk decreases consistently as customers ascend the loyalty ladder. Each tier step delivers a meaningful reduction in attrition probability, confirming the ROI of a well-structured loyalty programme.
Volume — Churned vs. Retained by Tier
Stacked absolute counts show that the No-Loyalty and Silver tiers contribute the majority of churned customer volume, primarily due to their larger base sizes.
Tier-by-Tier Retention Strategy
Specific intervention recommendations for each loyalty tier, prioritised by churn risk level and potential impact on the overall retention rate. Each tactic is directly actionable from the
loyalty_tier field in the dataset CSV.Category Analysis
Churn by Product Category
Churn rates vary meaningfully across the five primary product categories. These differences reflect distinct purchase motivations, price sensitivity, and competitive dynamics — each requiring a tailored retention approach informed by the primary_category field in the dataset.
Electronics & Fashion Drive the Most Churned VolumeElectronics has the highest churn rate (9.1%), while Fashion contributes the second-largest absolute volume of churned customers. Both benefit most from price-match guarantees, post-purchase trust-building sequences, and stronger returns policies.
Churn Rate by Primary Category
Electronics leads in churn rate (9.1%), followed by Fashion (8.4%). Sports and Beauty show the strongest retention, likely driven by habitual repurchase behaviour and lower price sensitivity in those verticals.
Volume — Churned vs. Retained per Category
Stacked absolute counts reveal that high-volume categories generate more churned customers in total even when their per-rate is similar. Use this to prioritise by business impact, not just churn percentage.
Category Summary & Retention Tactics
Per-category churn metrics from the dataset with prioritised retention recommendations. Each tactic addresses the specific purchase psychology and competitive dynamics of that product vertical, mapped directly from the
primary_category CSV column.