Employee retention metrics measure how many employees stay with an organisation over a defined period and are central to strategic HR decision-making.
Tracking these KPIs turns raw HR events—hires, promotions, transfers and terminations—into insight that links to cost savings, continuity of customer experience and team productivity.
This guide promises four deliverables: clear formulas you can use today; recent benchmarks and interpretation guidance; a cost-of-turnover model with a worked example; and a practical path to production using MiHCM analytics (Lite, Enterprise and Data & AI). Expect actionable templates, a retention-rate calculator walkthrough and an operational dashboard checklist to move from reporting to preventative action.
Why tracking retention metrics changes HR from reactive to strategic
- Retention KPIs quantify the business impact of people changes and prioritise interventions by financial value.
- When linked to manager- and role-level data, they reveal where targeted actions (onboarding, manager coaching, compensation) will have the greatest effect.
- Accurate metrics enable scenario modelling (for example, the savings from reducing voluntary turnover by X percentage points) that executives understand.
What to track and why
Retention rate and turnover rate are the baseline KPIs every HR team should track. Use cohort splits (new hires 0–6 months, manager, role, location) to find hotspots quickly. Model the cost of turnover so interventions can be prioritised by financial impact — replacing a mid-level hire commonly ranges from roughly 50% to 150% of annual salary according to industry analyses (SHRM, 2025; NIST, 2022).
Predictive signals let HR act earlier: engagement drops, repeated unplanned absence, missed 1:1s and recent role or manager changes are strong early indicators. Quick checklist:
- Retention rate formula (see section 4)
- Voluntary/involuntary exit split
- Average and median tenure
- Cost-of-turnover model
- Flight-risk indicators
Use rolling metrics (12-month rolling retention) for more stable trend analysis and cohort-level snapshots for tactical interventions.
What are employee retention metrics? Definitions & scope
Employee retention metrics measure the proportion of people who remain employed with an organisation over a defined period. They are distinct from turnover metrics, which count separations. Together they provide a holistic view of workforce stability and where HR effort should be concentrated.
Core definitions:
- Overall retention rate: percentage of employees who remain employed through a period.
- Turnover rate: percentage of separations during a period (can be overall, voluntary or involuntary).
- New-hire retention: retention measured for cohorts at 30, 90, 180 and 365 days.
- Average and median tenure: measures of how long employees stay, useful for trend detection and forecasting.
Complementary metrics:
- eNPS/engagement scores: leading indicators of retention risk; eNPS = % promoters − % detractors (NPS research, 2022).
- Months-to-first-promotion: proxy for career progression opportunities.
- Absence and unplanned leave rate: early signal of disengagement.
Core data you must have to calculate reliable retention metrics:
- Authoritative hire date and termination date fields (with timezone-normalised timestamps).
- Reason-for-leave categories standardised across teams (voluntary, involuntary, retirement, fixed-term end, internal transfer out).
- Manager, department, role and location tags for cohort slicing.
- Headcount snapshots or average headcount calculations for the period under analysis.
Which HR events to capture and how to standardise reason-for-leave:
Standardisation means mapping free-text reasons into a controlled taxonomy at the moment of exit and keeping a rehire flag for accurate cohort interpretation. Capture rehires with new employee IDs or a clear rehire date to avoid double-counting. Ensure part-time and FTE equivalents are recorded if comparisons must reflect workload rather than headcount.
How to calculate retention vs turnover (formulas & examples):
This section gives exact formulas, worked examples and Excel-ready steps to compute monthly and annual retention and turnover rates.
Retention rate: standard formulas
Two commonly used formulas (both acceptable; choose one and be consistent):
- Formula A (common in HR analytics): Retention rate = ((Employees at period end − New hires during period) ÷ Employees at period start) × 100.
- Formula B (alternate): Retention rate = (Employees who remained during period ÷ Employees at period start) × 100.
Worked example — annual retention (Formula A)
Assume 1 January headcount = 200, hires during year = 30, headcount 31 December = 190.
Retention = ((190 − 30) ÷ 200) × 100 = (160 ÷ 200) × 100 = 80%.
Interpretation: 80% annual retention means 20% of the start-year population left during the year.
Turnover rate: standard formula
Turnover rate = (Number of separations ÷ average headcount) × 100.
Average headcount = (headcount at period start + headcount at period end) ÷ 2, or use monthly snapshots for higher fidelity.
Worked example — annual turnover
Using the same dataset: separations = 40 (count of exits), average headcount = (200 + 190) ÷ 2 = 195.
Turnover = (40 ÷ 195) × 100 = 20.5%.
Monthly vs annual rates and conversions
- Monthly turnover = (Monthly separations ÷ average monthly headcount) × 100.
- To approximate annualised turnover from monthly rate r: Annualised ≈ 1 − (1 − r)^12 (for probability-based view), or multiply average monthly rate by 12 for rough estimates.
Retention-per-category (by department, manager, hire cohort)
Use the same formulas applied to the cohort population. Example: new hires in first 12 months — cohort start = hires with start date in period; survivors = those still employed at 12 months after hire; retention = (survivors ÷ cohort size) × 100.
Excel-ready tips
- Store one row per employee with hire_date, termination_date, manager_id, department, FTE, reason_for_leave.
- Use pivot tables or calculated columns to count cohort sizes and survivors.
- Use consistent inclusion rules for internal transfers and rehires; exclude internal transfers from separations if they remain employed in organisation.
Pitfalls to avoid
- Using raw snapshots without accounting for new hires during period.
- Misclassifying rehires or transfers as separations.
- Failing to adjust for part-time / FTE when required.
Top retention KPIs HR should track
Below are the primary KPIs HR should monitor. Group them into lagging and leading indicators for clarity.
Must-track KPIs
- Overall retention rate — annual and rolling 12-month.
- Overall turnover rate — split into voluntary and involuntary.
- New-hire retention — 30/90/180/365-day retention for onboarding evaluation.
- Average and median tenure — trend over time.
- Retention by manager, department, location and job level — identifies hotspots.
- Cost of turnover — per-exit and total cost model (see section 7).
- Engagement metrics tied to retention — eNPS, pulse survey change and manager effectiveness.
- Absence & unplanned leave rate — early signal of disengagement.
- Flight-risk score — composite model: pay competitiveness, engagement, time-in-role, manager change.
- Promotion and internal mobility rates — higher internal mobility generally correlates with better retention.
Which KPIs are leading vs lagging indicators?
| Metric | Type | Actionability |
|---|---|---|
| Retention rate | Lagging | Strategic review, budget planning |
| Turnover rate | Lagging | Root-cause analysis |
| eNPS / pulse | Leading | Manager coaching, engagement programs |
| Absence rate | Leading | Early engagement, wellbeing interventions |
Use leading indicators to triage interventions and lagging KPIs to measure program impact over quarters and years.
Benchmarks: What is a good employee retention rate?
Benchmarks depend heavily on industry, role seniority and geography. Use them as directional targets rather than absolute pass/fail thresholds.
High-performing organisations commonly target annual retention rates of 90% or higher (equivalently turnover ≤10%) — a practical, aspirational benchmark for many professional services and stable-product firms (research compilation, 2021).
Rule-of-thumb ranges (directional)
| Industry | Typical annual retention range |
|---|---|
| Technology / Professional services | 85%–95% |
| Healthcare / Education | 80%–92% |
| Retail / Hospitality | Lower retention; multiple reports cite substantially higher turnover vs corporate averages (often below 75%) |
How to choose the right benchmark for your organisation:
- Compare rolling 12-month retention to your past 3-year average to assess trend.
- Segment benchmarks by hire cohort (new hires vs tenured staff) and manager to find program targets.
- Target improvement where gaps exceed about 5 percentage points versus chosen target; those are highest-priority cohorts for intervention.
Benchmarks must be adapted: for example, high-churn frontline roles will have lower baselines than mid-office roles; measure both to avoid misleading averages.
Calculating the cost of turnover
Quantifying the cost of turnover lets HR convert retention improvements into hard savings for leadership. Break costs into direct and indirect categories and build an Excel model that accepts role-level assumptions.
Cost components:
- Separation costs — exit interviews, final payroll, administrative overhead.
- Recruitment costs — agency fees or internal recruiter time, advertising, assessment tools.
- Onboarding & training — formal training, manager time and peer training hours.
- Lost productivity — vacancy time and ramp to full productivity.
- Manager time — time spent interviewing, onboarding and knowledge transfer.
- Replacement pay differential — if replacement is paid at a premium.
Default quick assumptions (recommended):
- Recruitment cost = 20% of annual salary (agency/higher-cost hire) or 10% for in-house recruiting.
- Lost productivity = 3 months of productivity at 50% capacity (0.25 FTE-year equivalent) for mid-level roles.
- Onboarding and training = 1–2 months of manager and peer time.
- Separation admin = fixed amount (e.g., $1,000).
Worked example — mid-level role:
Assumptions: annual salary = $80,000; recruitment cost = 20% × salary = $16,000; onboarding & training + lost productivity ≈ $20,000; manager time and separation costs = $4,000. Total per-exit cost = $40,000 (50% of salary). Reducing voluntary exits by 2 roles per year saves $80,000 before intervention cost.
Scaling and scenario analysis:
Scale per-exit cost across team or organisation by multiplying per-exit cost by number of separations to estimate company-level impact. Run what-if scenarios: e.g., reducing voluntary turnover from 12% to 9% across a 1,000-headcount saves X hires avoided × per-exit cost = avoided cost. Use A/B testing for interventions to validate assumptions.
For guidance on industry per-role assumptions and a downloadable Excel template, adjust default assumptions above to reflect organisation-specific ramp times and recruiter costs. For wide roles or senior hires, expect per-exit costs to approach 100–200% of salary in some literature (SHRM, 2025).
Early warning signals & flight-risk metrics
Leading indicators allow HR teams to act before separations occur. Construct a composite flight-risk score from multiple signals and use it to prioritise supportive interventions.
Leading signals to monitor:
- Sharp drop in engagement or pulse scores over two consecutive surveys.
- Repeated unplanned absences or increased sick days.
- Missed 1:1s, declining participation in team rituals or training.
- Performance score declines or stalled progression.
- Recent manager change or extended recruitment activity for the individual’s role.
Composite flight-risk model inputs:
- Tenure and time-in-role.
- Pay competitiveness vs market benchmarks.
- Engagement trend (eNPS and pulse deltas).
- Recent role or manager change flag.
- Absence pattern and performance signals.
How to validate a flight-risk model:
- Run a retrospective analysis: compare historical scores to actual exits to assess precision and recall.
- Calibrate thresholds to control false positives — focus on top X% highest-risk where interventions are scalable.
- Continuously monitor model drift and update features with fresh data.
Privacy and ethical considerations:
Use flight-risk models to enable supportive interventions, not punitive actions. Limit sensitive feature access, document model logic and obtain governance sign-off. Communicate purpose to managers and protect individual privacy through aggregated reports where possible.
Sample early-intervention playbook
- Top-risk → immediate manager stay conversation within 7 days + tailored development plan.
- Mid-risk → pulse survey, career-path check-in and targeted upskilling.
- Low-risk → monitor engagement and include in next coaching cycle.
How to use retention metrics to design interventions:
Translate insights into prioritised interventions by combining attrition rates with per-exit cost to focus spend where it yields the highest ROI.
Prioritisation approach:
- Identify cohorts with highest attrition and highest per-exit cost (e.g., senior technical roles with high recruiter fees).
- Estimate annual avoidable exits = current exits × desired percentage-point improvement.
- Compute potential savings = avoidable exits × per-exit cost.
- Rank interventions by expected savings net of implementation cost.
Examples of interventions:
- Improved onboarding: structured 30/60/90 plans and manager checkpoints to boost new-hire retention.
- Manager training: coaching and calibration to improve manager effectiveness scores and retention by team.
- Career pathways and promotion clarity: reduce flight risk by shortening months-to-first-promotion.
- Targeted retention bonuses or counter-offers for high-value at-risk employees.
- Regular stay interviews with focused action items.
Simple ROI calculation:
If an intervention costs $30,000 to implement for a team and is expected to avoid 2 mid-level exits (per-exit cost $40,000), avoided cost = $80,000; net benefit = $50,000. Use this approach to make the case to leadership.
Experiment design:
When testing interventions, run controlled experiments where feasible. Example: randomly assign new hire cohorts to enhanced onboarding vs standard onboarding and compare 6–12 month retention. Use statistical significance checks and monitor secondary impacts such as engagement and performance.
MiHCM feature example: Turnover Management identifies peaks and at-risk employees using analytics to target interventions where the ROI is highest.
Tracking retention metrics with MiHCM
MiHCM captures the authoritative HR events and produces operational dashboards so teams can move from ad-hoc spreadsheets to repeatable workflows.
Authoritative data capture:
- Hire/termination dates, reason-for-leave, manager hierarchy and role metadata are recorded in MiHCM to ensure accurate cohort calculations.
- Rehire flags and FTE handling remove double-counting and enable cohort-level accuracy.
Dashboards to set up:
- Rolling retention dashboard — 12-month rolling retention with cohort filters.
- Voluntary vs involuntary split — monthly trend and manager-level drilldowns.
- Cohort retention — new-hire retention at 30/90/180/365 days.
- Cost-of-turnover modelling — role-level per-exit cost and scenario analysis.
MiHCM Data & AI runs flight-risk scoring and what-if cost scenarios while SmartAssist prompts managers with recommended actions (stay conversations, development nudges) when risk thresholds are met.
Operational workflow example: Daily data sync → weekly retention dashboard refresh → monthly leadership risk review → targeted interventions (manager prompts via SmartAssist) → follow-up pulse surveys to measure impact.
Access controls and governance: Define who sees manager-level retention (managers and HR business partners) vs executive summaries (C-suite). Log data access and use role-based controls to protect employee-level sensitivity.
Step-by-step: from raw HR events to an operational retention dashboard
- Ensure clean source data: hire/termination/manager fields validated.
- Define your retention formulas and cohort rules in the analytics layer.
- Build dashboards with filters for manager, role, tenure and location.
- Embed cost-of-turnover model and flight-risk outputs for prioritisation.
- Automate alerts and manager workflows for at-risk cohorts.
Comparison: traditional KPIs vs predictive analytics
Traditional KPIs and predictive analytics are complementary. Use the former for governance and reporting and the latter to prioritise proactive interventions.
Key contrasts
| Type | Examples | Lag/Lead | Actionability | Data needs |
|---|---|---|---|---|
| Traditional KPIs | Retention rate, turnover, tenure | Lagging | Strategic planning and post-facto assessment | Hire/termination dates, headcount |
| Predictive models | Flight-risk score, propensity-to-leave | Leading | Targeted interventions and prioritisation | Engagement, compensation, time-in-role, absence patterns |
Trade-offs:
- Predictive models require more data and maintenance and can generate false positives; they must be governed and validated.
- Traditional KPIs are simpler to compute and essential for compliance and trend reporting.
Recommended hybrid approach:
- Keep robust traditional measurement while piloting predictive models with clear guardrails.
- Use retrospective validation and gradually expand model scope as reliability improves.
When to add predictive models (maturity checklist)
- Clean and consistent event-level HR data exists.
- Sufficient historical turnover events (recommended: hundreds of exits) to train and validate models.
- Governance processes for model use and manager actions.
Examples — retention rate calculator & template walkthrough: A downloadable retention-rate Excel template should include one row per employee with fields: employee_id, hire_date, termination_date, reason_for_leave, manager_id, department, job_level, FTE. Use pivot tables to create cohorts and calculate survivors.
Walkthrough – sample calculation:
- Load dataset into Excel and create a pivot table grouped by hire cohort (e.g., hires in Jan 2024).
- Create calculated columns: tenure_days = IF(termination_date, termination_date − hire_date, TODAY() − hire_date).
- For 365-day retention, count hires with tenure_days ≥ 365 and divide by cohort size.
Monthly retention calculator: Monthly formula: (Employees at month end − New hires during month) ÷ Employees at month start × 100. To convert to a rolling 12-month retention, compute the value for each trailing 12-month window.
Scenario demonstration: Assume voluntary turnover is 12% annually for a 500-person team; per-exit cost = $30,000. Exits/year = 0.12 × 500 = 60; cost = 60 × $30,000 = $1.8M. If an intervention reduces voluntary turnover to 9%, exits = 45; cost = 45 × $30,000 = $1.35M; avoided cost = $450,000. Subtract program cost to compute net savings.
Use the template to generate pivot-ready columns and integrate role-level assumptions for per-exit cost. The template supports scenario analysis and simple charts for executive reporting.
Conclusion and next steps
Must-track metrics: retention rate (overall and rolling), turnover split (voluntary vs involuntary), new-hire retention, average tenure, engagement indicators and cost-of-turnover. Implementation path: clean event data, define formulas and cohorts, build dashboards, and pilot predictive scoring.
90-day action checklist for HR leaders:
- Clean data: validate hire/termination dates, manager mapping and reason-for-leave taxonomy.
- Implement core dashboards: rolling retention, voluntary/involuntary split and new-hire retention.
- Run cost-of-turnover analysis for top 3 roles to prioritise interventions.
- Pilot flight-risk model with one business unit and document governance.
How to prioritise first pilot cohorts:
- Start where per-exit cost is high and attrition rate exceeds internal benchmark by >5 percentage points.
- Prefer cohorts with manageable size so experiments are statistically tractable (e.g., teams of 50–300).
คำถามที่พบบ่อย
What is employee retention rate?
How do you calculate retention vs turnover rate?
What is a good retention rate?
Varies by industry. Many organisations aim for ≥90% annual retention (≤10% turnover) as a high-performing benchmark (research, 2021).
How is the cost of turnover determined?
Sum separation, recruitment, onboarding/training, lost productivity and manager time. Literature commonly cites replacement costs in the range of 50%–150% of annual salary depending on role seniority (SHRM, 2025).
