Across 1470 employees, the overall attrition rate is 16.1% (237 departures). The strongest single predictor of attrition is Job Role: Human Resources — controlling for all other factors in the logistic regression model. The model achieves an AUC of 0.838, indicating good discriminatory power. 27 employees currently exceed the high-risk probability threshold and represent the priority cohort for targeted retention action.

The highest attrition is in Sales at 20.6%, compared to the company-wide average of 16.1%. Departments significantly above the average should be prioritised for engagement surveys and manager effectiveness reviews. Rates below the company average may indicate practices worth replicating in higher-attrition functions.

Sales Representative has the highest attrition rate at 39.8%, well above the company average of 16.1%. 4 of 9 job roles exceed the company-wide average, suggesting that attrition risk is concentrated in specific functions rather than evenly distributed across the organisation. Roles at the top of this chart warrant pipeline-review and targeted compensation benchmarking.

Attrition risk peaks at Level 1 (26.3%), compared to the company-wide average of 16.1%. Junior levels (Level 1–2) average 18% attrition, confirming the well-documented new-hire vulnerability pattern. Senior levels (Level 4–5) typically show much lower attrition as long tenure and higher compensation create stronger retention anchors.

After controlling for all other predictors, Job Role: Human Resources has the highest odds ratio (772171.969x), meaning employees with this attribute are approximately 772172 times more likely to leave. Years at Company is the strongest protective factor (OR = 0.987), significantly reducing attrition odds. 8 predictor levels have confidence intervals entirely above 1.0, confirming statistically significant risk-elevation.

The random forest ranks Monthly Income as the most important predictor, followed by Age. This partially diverges from the logistic regression ranking, suggesting that non-linear effects or interaction terms are at work — these are captured by the random forest but not by the linear model. Predictors appearing in the top five of both models represent the most robust retention levers available to HR leadership.

27 employees exceed the high-risk attrition probability threshold. The attribute that most distinguishes them from the average employee is Stock Option Level — high-risk employees score/earn 90.7% lower than the overall workforce average. The delta column shows the percentage difference between high-risk and overall averages: large positive deltas for scores (e.g. years at company) suggest retention anchors are absent, while large negative deltas for satisfaction ratings signal disengagement.

Attrition risk peaks at 0 year(s) of tenure (36.4% annualised rate). The average attrition rate in the first three years is 26.9%, confirming a critical early-career vulnerability window that onboarding and early-career development programmes should address. Points above the company-wide average line represent the tenure milestones where HR should concentrate proactive retention outreach.

Employees who left had a median monthly income of $3,202 compared to $5,204 for those who stayed — a 38.5% gap. The compensation difference is statistically significant (p < 0.05). While salary is rarely the sole driver of attrition, a consistent compensation gap below market rates, especially at lower job levels, provides HR with a quantifiable salary lever for targeted retention offers.