People analytics and workforce analytics are used interchangeably in many organizations, but in a contact center they answer different questions, use different data, and are owned by different teams. Conflating them leads to two problems: HR uses operational data without understanding its context, and operations ignores HR data that explains why the operational metrics look the way they do.
People analytics answers questions about the employee lifecycle: Who should we hire? Why are agents leaving? What predicts whether a new hire will succeed? Is compensation competitive? Are we developing people into the roles we need?
Workforce analytics answers questions about operational performance: Do we have enough agents on the phones? Is the schedule matching the demand? Are agents productive? What is our cost per call? Are we meeting SLAs?
Both are data practices. Both matter. But they serve different functions, and understanding where each starts and stops is essential for using either one effectively.
The distinction
| Dimension | People analytics | Workforce analytics |
|---|
| Focus | The employee — as an individual with a career, compensation, skills, and trajectory | The operation — as a system that must produce output at a target quality and cost |
| Primary questions | Who should we hire? Why do people leave? What predicts success? Is compensation fair? | How many agents do we need? Are we hitting service level? Where is productivity lost? What does it cost? |
| Time horizon | Long-term: hiring cycles, tenure patterns, career development, year-over-year trends | Short-term and medium-term: daily adherence, weekly forecast accuracy, monthly labor cost |
| Typical owner | HR, talent management, or people operations | WFM team, operations management, or contact center leadership |
| Data sources | HRIS, applicant tracking, exit interviews, compensation benchmarks, engagement surveys, performance reviews | ACD, WFM system, time tracking, QA evaluations, CRM |
| Output | Hiring strategy, retention programs, compensation adjustments, training plans, succession planning | Staffing plans, schedules, coaching priorities, overtime decisions, cost reports |
The questions it answers
| Question | Data required | Decision it drives |
|---|
| What predicts whether a new hire will succeed? | Applicant source, assessment scores, training performance, 90-day retention, 90-day QA scores | Refine hiring criteria. If agents from Source A have 70% 90-day retention and agents from Source B have 40%, invest more in Source A |
| Why are agents leaving? | Exit interview data, tenure at departure, supervisor, shift, performance rating at departure, compensation vs. market | Targeted retention interventions. If 60% of exits cite schedule inflexibility, the fix is scheduling policy — not a pay increase |
| Where in the lifecycle do we lose people? | Attrition segmented by tenure: 0–30 days, 31–90 days, 91–180 days, 180+ days | If early attrition (0–90 days) is 3× tenured attrition, the problem is onboarding or training, not general dissatisfaction |
| Is our compensation competitive? | Internal pay by role and tenure vs. market benchmarks for the same geography and role | If the operation pays $13/hour and the local market rate is $15/hour, attrition will continue regardless of other interventions |
| Are we developing internal talent? | Internal promotion rate, time from agent to team lead, career path completion rate | If zero agents have been promoted to supervisor in 18 months, top performers will leave for advancement opportunities elsewhere |
| What is the cost of attrition? | Replacement cost per departure (recruiting + training + ramp productivity loss) × departures per period | Justifies retention investments. If each departure costs $6,000 and attrition drops from 50 to 40 agents/year, the savings are $60,000 — budget available for retention initiatives |
Data sources
| Source | What it provides |
|---|
| HRIS (HR Information System) | Hire dates, separation dates, job history, compensation, demographics, supervisor assignments |
| Applicant tracking system | Applicant source, assessment results, time-to-hire, offer acceptance rate |
| Exit interviews | Stated reasons for leaving, categorized by theme (pay, schedule, management, career, personal) |
| Engagement surveys | Satisfaction scores by category (schedule, compensation, management, development), response rate, trend over time |
| Performance review records | Rating history, development goals, improvement plan outcomes |
The questions it answers
| Question | Data required | Decision it drives |
|---|
| Do we have enough agents? | Volume forecast, AHT, shrinkage, target service level | Staffing plan — how many agents to schedule per interval |
| Is the schedule matching demand? | Actual agents logged in vs. agents needed per interval | Schedule redesign — adjust shift patterns, add part-time peak coverage, stagger breaks |
| Are agents productive? | Occupancy, calls per hour, AHT, FCR, QA scores | Agent-level coaching — identify who needs help and what kind |
| What does it cost? | Total labor hours by category (regular, overtime, training) × rates | Labor cost management — is overtime structural? Is training cost reasonable? |
| Where is time going? | Time by activity category: on-phone, ACW, break, training, coaching, admin | Shrinkage diagnosis — if actual shrinkage is 32% but the model assumes 25%, every schedule is wrong |
| Is the forecast accurate? | Forecasted volume vs. actual volume by interval, by day, by week | Forecast method adjustment — correct for bias, anomalies, or missing event data |
Data sources
| Source | What it provides |
|---|
| ACD | Call volume, AHT, service level, agent states, queue data, FCR indicators |
| WFM system | Forecast, schedule, adherence, staffing requirements |
| Time tracking | Clock-in/out, break times, activity categorization, overtime hours |
| QA system | Evaluation scores by agent, rubric category, and period |
| CRM | Disposition codes, case resolution data, callback frequency |
For a deeper look at the workforce analytics practice — what to measure, how to connect metrics to decisions, and how to build the review cadence — see the workforce analytics guide and the implementation guide.
Where they overlap
Some questions sit at the intersection of people analytics and workforce analytics. These are the areas where HR and operations must collaborate — neither team has the full picture alone.
| Overlap area | People analytics contribution | Workforce analytics contribution | Combined insight |
|---|
| Attrition impact | Why agents leave (exit data, satisfaction surveys, compensation analysis) | What attrition costs operationally (overtime to cover gaps, service level impact, training replacement cost) | The full cost of attrition and the right retention investment — HR knows the cause, operations knows the cost |
| Training effectiveness | Training completion rates, trainee satisfaction, instructor evaluations | Post-training performance: AHT, QA scores, FCR for agents who completed training vs. those who did not | Whether training actually produces better agents — not just whether agents completed it |
| Performance management | Performance rating, development goals, career trajectory, compensation decisions | Performance metrics: AHT, FCR, QA scores, adherence, calls per hour | A performance review that combines objective metrics with behavioral assessment and development context |
| Schedule satisfaction and attrition | Exit data showing schedule-related departures, engagement survey responses on schedule satisfaction | Schedule patterns: which shifts have highest attrition, whether agents on fixed shifts stay longer than those on rotating shifts | The specific schedule changes that would reduce attrition — not generic "schedule flexibility" but data showing that agents on Shift C leave at 2× the rate of Shift A |
| Workload and burnout | Engagement survey data on stress, burnout indicators, sick leave patterns | Occupancy data, consecutive high-intensity intervals, overtime hours per agent | Whether overloaded agents are the ones leaving — connecting operational workload data to HR attrition data reveals whether staffing levels are causing turnover |
Who should own what
| Practice | Owner | Why this team |
|---|
| People analytics | HR / talent management | HR has access to compensation data, exit interviews, engagement surveys, and applicant tracking. They understand employment law constraints on how data can be used. They own the hiring pipeline, compensation structure, and retention programs |
| Workforce analytics | WFM team or operations management | Operations has access to ACD data, time tracking, QA, and scheduling systems. They understand the operational context (what service level means, why occupancy matters, how shrinkage works). They own the staffing plan, schedule, and coaching process |
| Overlap areas | Joint — with defined handoff | HR provides the "why" (why people leave, what they want). Operations provides the "how much" (what it costs, what it does to service level). Both are needed for decisions like retention investment, schedule redesign, or training overhaul |
The handoff in practice
| Situation | HR provides | Operations provides | Joint decision |
|---|
| Attrition spike in Q2 | Exit data: 65% of departures cite "better pay elsewhere." Compensation analysis shows the operation is $1.50/hour below market | Replacement cost: $6,000 per departure × 15 departures = $90,000. Overtime cost to cover gaps: $4,200/month | Is a $1.50/hour raise (cost: $1.50 × 40 hrs × 52 weeks × 100 agents = $312,000/year) justified by the attrition cost savings ($90,000 replacement + $50,400/year overtime)? The math says no for a blanket raise — but a targeted raise for agents with 6+ months tenure (where attrition is lower and replacement cost is higher) may work |
| QA scores declining across the operation | Training records: no refresher training in 6 months. New hires received shorter training (3 weeks vs. historical 4 weeks) | QA data: scores declined 4 points in rubric categories "resolution accuracy" and "process compliance." FCR dropped 3 points | The training reduction is the likely cause. Restore the 4th training week and add monthly refreshers targeting the specific rubric categories that declined |
Common mistakes
| Mistake | What happens | Fix |
|---|
| Using workforce analytics terms for people analytics questions | Operations asks "what is our attrition rate?" and calculates it from headcount data. But they do not segment by tenure, reason, or voluntary vs. involuntary — which are people analytics questions | Attrition rate is a workforce analytics metric (operational impact). Attrition diagnosis (why, who, when) is people analytics. Both are needed |
| Treating engagement surveys as workforce analytics | An engagement survey shows satisfaction at 3.2/5. Operations responds by adjusting schedules. But the low satisfaction may be about compensation or career development — not scheduling | Engagement surveys are people analytics. Use the survey to identify the theme, then apply workforce analytics (schedule data, workload data) or people analytics (compensation data, promotion data) depending on the finding |
| Ignoring people analytics entirely | Operations manages by metrics only: AHT, adherence, QA scores. Agents who meet targets are "fine." Agents who miss targets get coaching. No one asks why a previously strong agent's performance is declining | An agent whose performance drops may have a personal issue, a schedule conflict, a compensation grievance, or a supervisor conflict. People analytics (or simply a conversation) reveals the cause. Workforce analytics only shows the symptom |
| Running people analytics without operational context | HR reports that 40% of agents are "disengaged" based on a survey. But the survey was administered during a period when the operation was chronically understaffed and agents were working mandatory overtime. The disengagement is a symptom of the staffing problem, not a separate HR issue | Combine: HR's engagement data + operations' occupancy and overtime data. The fix is staffing, not an engagement program |
Building both practices
| Step | People analytics | Workforce analytics |
|---|
| Start with | Segment attrition by tenure and reason. This single analysis reveals more than any other people analytics starting point | Compare forecasted vs. actual volume and calculate actual shrinkage. These two data points tell you whether your staffing model is working |
| Add next | Track 90-day new hire retention by source, trainer, and training class. Identifies which parts of the hiring and training pipeline produce successful agents | Add agent-level productivity profiles combining AHT, FCR, QA, and adherence. Identifies who needs coaching and what kind |
| Mature practice | Predictive models: which current agents are at risk of leaving based on tenure, satisfaction, compensation gap, and performance trend | Connected analytics: workload analysis linking staffing data, cost data, and quality data to find productivity improvements. See the workforce analytics implementation guide |
| Integration | Joint quarterly review where HR and operations share findings. HR presents attrition and engagement data. Operations presents productivity, cost, and service level data. Together they identify root causes and prioritize interventions | |
