---

name: attrition-audit
description: "Synthesize turnover data and culture signals to find the root causes of attrition — not the exit survey answers, which are almost always sanitized."

/attrition-audit

Exit interviews are largely useless. Departing employees say "better opportunity" and "compensation" because they don't want to burn a reference. The real drivers — a specific manager, a dysfunctional team dynamic, work that doesn't match what was promised — surface three data sources later. This skill triangulates departure patterns, pulse survey trends, and 1:1 themes to find what people won't say on the way out, then prioritizes the intervention with the highest ROI before another quarter passes.

Departure Pattern Analysis

• Attrition rate by department, role level, and tenure band — where is it highest?
• Voluntary vs involuntary split — is this a retention problem or a hiring accuracy problem?
• Time-to-departure pattern: are people leaving at 3 months (onboarding failure), 18 months (growth plateau), or 3+ years (burnout/leadership)?
• Manager-level attrition correlation: does attrition cluster under specific managers? (This is the most common root cause that goes unaddressed)
• Regrettable vs non-regrettable: of voluntary departures, what percentage did you want to keep?

Cross-Reference: What The Surveys Are Actually Saying

• Pulse survey scores by department and manager — where are the outliers (high and low)?
• Which specific questions have trended down over the last 3 surveys? ("My manager gives me useful feedback" and "I see a path to grow here" predict attrition 6 months ahead)
• 1:1 themes flagged by managers over the last quarter — what recurring concerns appear?
• Glassdoor/Blind patterns if available — what themes appear in negative reviews that aren't in exit surveys?

Root Cause Identification — distinguish root causes from symptoms:

• Symptom: "compensation complaints." Root cause: "compensation is below market for roles that can easily move" OR "compensation is fine but perceived as unfair internally"
• Symptom: "lack of growth." Root cause: "no internal mobility program" OR "managers don't know how to have career conversations" OR "the role was misrepresented in hiring"
• Name the 2-3 root causes with supporting evidence from at least 2 data sources each

Cost of Attrition — quantified:

• Loaded replacement cost per role type (standard formula: 0.5-2x annual salary depending on seniority)
• Current annualized attrition cost in dollars
• Cost of losing institutional knowledge vs. hiring replacements: estimate the ramp time and productivity loss
• Cost of remaining team morale drag — 1 regrettable departure impacts 3-5 peers' engagement scores

Intervention Prioritization

For each root cause, identify: the intervention, the estimated cost, the time to measurable impact, and the leading indicator that tells you it's working before attrition numbers improve (which always lag 6-9 months).

Success Criteria at 6 Months

• Which specific metric changes first (pulse score for manager feedback, offer acceptance rate, 90-day voluntary departures)?
• What does the data need to show in 6 months to confirm the intervention worked?
• What would indicate you misdiagnosed the root cause?

Rules

1. Exit survey data is a starting point, not a conclusion. Always cross-reference with pulse and 1:1 data.
2. Manager-level analysis is mandatory. Aggregated department data hides the real driver.
3. Every root cause needs evidence from at least 2 independent sources.
4. Cost of attrition must be a real number. "Significant" or "high" is not an input to a budget conversation.
5. Interventions must have a leading indicator — attrition data lags the actual problem by 6-9 months.
6. Distinguish what you know from what you're inferring. Label both.

The output of this skill is a 1-page brief for your leadership team: root cause, cost, intervention, ROI estimate, and the metric you'll watch to know it's working.