Model
Segment Filter
Risk Threshold70%
10%95%
Date Range
Data Source
Connected to Snowflake
Last refreshed: Mar 18, 2026 08:15
Customer Churn Predictor
Real-time churn risk scoring using XGBoost v4 model
Total Customers
48,321
+2.4% vs last month
High Risk (>70%)
1,847
+12.1% vs last month
Avg Churn Prob
14.2%
-0.8% vs last month
Model Accuracy
94.1%
+0.3% vs last month
Churn Probability Distribution
0-10%
58.8%
10-20%
17.1%
20-30%
8.5%
30-40%
5.1%
40-50%
3.4%
50-60%
2.3%
60-70%
0.9%
70-80%
1.3%
80-90%
1.5%
90-100%
1.1%
Below threshold Above threshold (70%)
Feature Importance
months_since_last_activity
23%
support_tickets_90d
18%
usage_decline_pct
15%
contract_months_left
12%
nps_score
9%
login_frequency_30d
8%
feature_adoption_pct
6%
payment_delays_count
5%
Top Risk Customers
Threshold: 70%| Customer ID | Segment | Tenure (days) | ARR | Churn Prob |
|---|---|---|---|---|
| C-109234 | Enterprise | 412 | $284k | 96.1% |
| C-087451 | SMB | 89 | $42k | 94.9% |
| C-203918 | Enterprise | 1104 | $512k | 93.4% |
| C-055672 | Mid-Market | 237 | $128k | 92.2% |
| C-312045 | SMB | 34 | $18k | 91.1% |
| C-441290 | Mid-Market | 156 | $96k | 88.7% |
| C-198734 | Enterprise | 890 | $340k | 87.4% |
| C-562018 | SMB | 67 | $31k | 86.1% |
Data sourced from Snowflake · Model: XGBoost v4 (Production) · Threshold: 70%