HR Analytics Glossary: Key Terms for People Science & Recruiting

HR and recruiting teams that cannot speak the language of analytics are at a structural disadvantage. When the CFO asks about the ROI of a hiring initiative or the CEO wants a workforce scenario model, vocabulary gaps become credibility gaps. This glossary defines the 12 terms HR and recruiting professionals encounter most — grounded in the practical, automation-first context that makes them actionable. It is part of the broader HR digital transformation strategy this site covers in depth.

Jump to a term:

What is HR analytics and how is it different from traditional HR reporting?

HR analytics is the systematic collection, analysis, and interpretation of workforce data to improve organizational performance. Traditional reporting describes what happened; analytics explains why and forecasts what will happen next.

A report tells an HR director that voluntary turnover rose 8% in Q2. Analytics identifies which teams, tenure bands, and managers are driving that number — and predicts which employees are most likely to leave next quarter. The practical gap between those two capabilities is the gap between reactive and proactive HR.

Automation is the enabler. Platforms that pull data automatically from your ATS, HRIS, performance management system, and engagement surveys eliminate the manual aggregation work that makes analytics unrealistic for lean teams. Without automated data collection, HR analytics is a weekend project that never stays current. For a structured approach to assessing where your team stands, see this digital HR readiness assessment.

Why it matters: SHRM research documents that organizations using workforce analytics report higher revenue per employee and faster time-to-fill than those that do not. The gap is not technology — it is whether the data plumbing is in place to make analytics continuous rather than episodic.

What is people science and how does it relate to HR analytics?

People science is an interdisciplinary field applying psychology, sociology, economics, and statistics to understand and improve human behavior at work. HR analytics is the data infrastructure; people science is the research discipline that interprets it.

Where analytics asks “what does the data show,” people science asks “why does this pattern exist and what evidence-based intervention will change it?” In recruiting, people science informs the design of structured interviews with validated scoring rubrics, unbiased assessment tools grounded in cognitive psychology, and onboarding sequences built on motivation and belonging theory.

Automation supports people science by providing clean, longitudinal datasets for research and by deploying interventions — personalized learning pathways, automated feedback nudges, structured check-in sequences — at scale without proportional headcount increases.

What are the four levels of HR analytics — and which level should HR teams prioritize first?

The four levels form a maturity hierarchy. Most HR teams stall at level one and never reach level four.

  1. Descriptive analytics — Summarizes historical data. “What happened?” (Turnover was 14% last year.)
  2. Diagnostic analytics — Explains causes. “Why did it happen?” (Turnover concentrated in managers with low 360-feedback scores.)
  3. Predictive analytics — Forecasts outcomes. “What will happen?” (These 23 employees have a 78% flight-risk probability in the next 90 days.)
  4. Prescriptive analytics — Recommends actions. “What should we do?” (Reassign these employees to high-autonomy projects and schedule manager coaching for these five leaders.)

The sequence matters: build descriptive accuracy first by automating data collection and standardizing definitions across systems, then layer diagnostic and predictive capabilities on that clean foundation. Jumping to AI-powered prediction before data quality is solid produces faster wrong answers — not better decisions. This progression connects directly to predictive HR analytics and workforce strategy covered in the sibling satellite on that topic.

Jeff’s Take: Most HR Teams Are Solving the Wrong Analytics Problem

When HR leaders tell me they want predictive analytics, I ask one question first: can you pull a clean, consistent headcount report without touching a spreadsheet? Ninety percent of the time, the answer is no. That is the real problem. Predictive models trained on dirty, manually assembled data do not produce better decisions — they produce confident wrong answers. The sequence that works is: automate data collection, standardize definitions across systems, build reliable descriptive reporting, then layer diagnostics and prediction. Skip steps and you are building a skyscraper on sand.

What is predictive analytics in HR, and what are the most common use cases?

Predictive analytics uses statistical algorithms and machine learning models trained on historical data to forecast future workforce outcomes. It shifts HR from documenting problems after they occur to intervening before they become costly.

The most common HR applications:

  • Employee turnover prediction — Identifies employees at elevated flight risk based on engagement signals, tenure, performance trajectory, and manager relationship indicators.
  • Candidate success scoring — Ranks applicants by predicted on-the-job performance using resume signals, assessment results, and historical hire-performance correlations.
  • Skill-gap forecasting — Projects future capability requirements based on business growth plans and models the internal development needed to close gaps before external hiring becomes necessary.
  • Offer acceptance modeling — Predicts which candidates are likely to decline offers based on compensation, location, role characteristics, and pipeline timing, enabling proactive countermeasures.

Harvard Business Review research confirms that data-driven talent decisions consistently outperform intuition-based approaches. The critical caveat: biased historical hiring decisions produce biased predictions. A model trained on a decade of homogeneous hiring outcomes will replicate that homogeneity unless bias auditing is built into the pipeline. See the section on algorithmic bias below for specifics. The sibling satellite on using predictive analytics to retain strategic talent covers retention modeling in depth.

What is workforce planning and why does it require automation to be effective?

Workforce planning is the strategic process of forecasting an organization’s future talent needs, identifying gaps between current capabilities and future requirements, and building a roadmap to close those gaps through hiring, development, or restructuring.

Done manually, workforce planning is a backward-looking exercise completed once a year, instantly obsolete, and disconnected from live business conditions. Automated workforce planning integrates HRIS, financial forecasting, and external labor-market signals in real time — enabling scenario modeling that static spreadsheets cannot support. For example: “What happens to our product engineering pipeline if we open a new facility in 18 months while simultaneously losing two senior engineers to a competitor?” That question requires dynamic data, not a Q1 headcount report.

Automation removes the data-assembly work that prevents HR leaders from exercising strategic judgment. It does not replace the judgment itself. The decisions — which gaps to fill via hiring versus development, how to sequence a restructuring — remain human calls informed by clean, current data.

What is an HRIS, and what is the difference between an HRIS, an ATS, and an HCM?

Three systems, three scopes — and the integration gaps between them are where HR data quality breaks down.

  • HRIS (Human Resource Information System) — System of record for core employee data: demographics, compensation, benefits, employment history, and compliance documentation.
  • ATS (Applicant Tracking System) — Manages the recruiting pipeline: job postings, candidate profiles, interview stages, scorecards, and offer letters.
  • HCM (Human Capital Management) — A superset platform that typically includes HRIS functions plus talent management, performance, learning, and often payroll in a unified suite.

The practical problem: data lives in separate systems and must flow between them reliably. When it does not — when ATS offer data must be re-entered manually into the HRIS — transcription errors occur. A documented failure mode: a $103K offer letter becomes $130K in payroll because someone misread a number during manual entry, costing $27K before the employee resigned over the mismanaged situation. Automated system integrations between ATS and HRIS eliminate that entire failure category.

What is employee lifetime value (ELTV) and how do HR teams measure it?

Employee lifetime value is the total net contribution an employee delivers across their tenure, relative to the total cost of acquiring, developing, and retaining them.

ELTV is not yet a standardized metric with a universal formula — organizations define and weight components differently. The core inputs are consistent: output or revenue attributed to the employee’s role, minus total employment costs including recruiting and onboarding investment, salary, benefits, training, and manager time. Parseur research estimates manual data entry alone costs organizations $28,500 per employee per year in lost productivity — a cost that factors directly into ELTV calculations for roles where administrative burden is high.

ELTV thinking shifts HR conversations from “what does this hire cost” to “what does this hire generate.” It is most actionable when tied to performance data and tenure cohort analysis, allowing HR to identify which sourcing channels, role categories, or manager assignments produce the highest long-term value — and to allocate acquisition spend accordingly.

What is attrition rate and how is it different from turnover rate?

Two terms that sound interchangeable but drive different strategic responses.

  • Attrition rate — Percentage of employees who leave and are not replaced. Headcount permanently reduces. Often intentional: planned workforce reduction, natural retirement, role elimination.
  • Turnover rate — Percentage of employees who depart regardless of whether the role is backfilled. Includes voluntary resignation, involuntary termination, and retirement.

Formula for both: (Departures ÷ Average Headcount) × 100, calculated over a defined period.

The distinction matters for workforce planning. High attrition shrinks organizational capacity and requires a structural response — headcount model revision, workload redistribution, or a deliberate decision that the roles are no longer needed. High voluntary turnover signals culture, compensation, manager-effectiveness, or role-fit issues that predictive analytics can identify before they compound. SHRM research on the cost of unfilled positions documents that each open role costs $4,129 in direct and indirect costs while it sits vacant — making early voluntary turnover detection a direct financial priority, not a soft HR concern.

What is organizational network analysis (ONA) and why are HR leaders paying attention to it?

Organizational network analysis maps the informal relationships, communication flows, and collaboration patterns that actually drive work — as opposed to the formal org chart, which reflects reporting structure but rarely reflects how decisions get made or knowledge flows.

ONA reveals: influence nodes (people who broker information across teams), collaboration bottlenecks (roles or individuals that create workflow dependencies), and employees at burnout risk due to network overload — being copied on too many communications, consulted across too many workstreams. It is conducted using email and calendar metadata, collaboration platform data, or validated survey instruments.

McKinsey research on organizational health demonstrates that informal networks frequently predict performance outcomes better than formal hierarchy. HR leaders use ONA to inform restructuring decisions, identify hidden key persons whose departure would cascade across multiple teams, and target engagement or workload interventions where they will have maximum structural impact — rather than defaulting to blanket programs that reach everyone and move no one.

What is algorithmic bias in HR analytics and what does ‘bias auditing’ mean in practice?

Algorithmic bias occurs when an automated decision-support model produces systematically different outcomes for groups defined by protected characteristics — gender, race, age, disability status — due to biased training data, flawed feature selection, or model design choices that proxy for protected attributes.

In HR, the highest-risk applications are resume screening, candidate ranking, performance scoring, and promotion recommendation systems. Models optimizing for “fit with successful past hires” will replicate historical hiring patterns, which in many organizations means replicating historical underrepresentation.

Bias auditing means regularly testing model outputs for disparate impact: comparing selection rates, score distributions, and outcome distributions across demographic groups, then tracing discrepancies back to model inputs and training data. A responsible audit examines both historical training data and live model outputs on a defined schedule — not once at deployment and never again. Gartner research consistently positions algorithmic accountability as a top governance risk for HR technology investments. See the companion satellite on ethical AI frameworks for HR leaders for implementation guidance.

What We’ve Seen: Bias Auditing Gaps in Automated Screening

Gartner consistently flags algorithmic accountability as a top HR technology governance risk — and we see why. Organizations deploy AI-assisted resume screening, run it for 12–18 months, and never look at disparate impact data. The model optimizes for signals correlated with past hires, replicating historical bias at scale. A basic bias audit — comparing selection rates by gender and race against applicant pool composition — takes less than a day once data is accessible. Not running it is a compliance and reputational risk, not just an ethics concern.

What is a talent dashboard and what metrics should it display?

A talent dashboard is a real-time visual interface consolidating key workforce metrics into a single view for HR leaders and business partners. It converts raw system data into decision-relevant signals.

The most actionable dashboards organize metrics into five clusters:

Cluster Key Metrics
Acquisition Time-to-fill, cost-per-hire, offer acceptance rate, sourcing channel mix
Retention Voluntary turnover rate, attrition by tenure band, flight-risk scores
Performance Goal completion rate, performance rating distribution by manager and department
Development Learning completion rates, internal mobility rate, skills inventory coverage
Composition Headcount by department, span of control, DEI representation metrics

The dashboard is only as valuable as the data feeding it. Automating data ingestion from HRIS, ATS, and performance systems is the prerequisite. A dashboard built on manually exported spreadsheets is always stale — and stale data drives stale decisions.

What does ‘data governance’ mean in the context of HR analytics?

HR data governance is the set of policies, standards, roles, and processes that determine how employee data is collected, stored, accessed, used, and retired. It answers four questions: who owns each data element, who is authorized to access it, how is its accuracy maintained, and how long is it retained.

Poor governance produces cascading quality failures. A single definition inconsistency — “department” coded four different ways across three systems — corrupts every downstream report and model that aggregates by department. The 1-10-100 rule, documented by Labovitz and Chang and cited in MarTech research, captures the cost structure precisely: it costs $1 to prevent a data error at the source, $10 to correct it after the fact, and $100 to do nothing and absorb the downstream consequences in flawed decisions.

For HR teams building analytics programs, data governance is not bureaucracy — it is the load-bearing foundation that makes every metric trustworthy. Without it, analytics investment produces expensive dashboards that HR leaders learn not to trust. With it, the data becomes an asset that compounds over time. The sibling satellite on building a robust HR data governance framework covers implementation specifics.

In Practice: The Data Governance Tax Nobody Budgets For

Every HR analytics project I have seen underestimates the same thing: data governance. Teams budget for the dashboard tool and the data scientist, but not for the policy work — defining who owns each data element, resolving the four different ways “department” is coded across three systems, or deciding how long terminated employee records are retained. That governance gap is why the 1-10-100 rule hits HR so hard. The $1 prevention cost is skipped, and teams absorb the $100 consequence downstream in corrupted reports and flawed model outputs.

Putting These Terms to Work

Vocabulary is the starting point, not the destination. These definitions become strategic assets when HR leaders use them to evaluate vendor claims accurately, challenge data assumptions in leadership conversations, and sequence analytics investments in the order that builds on itself — clean data first, predictive models second, prescriptive automation third.

The broader sequence is the same one this site’s HR digital transformation strategy covers in full: automate the repetitive data layer, govern the data that flows through it, then use analytics and AI at the specific decision points where human judgment needs data support rather than data replacement.

For teams ready to act on these concepts, two next steps matter most: map your current automation gaps using a structured process like shifting HR from manual processes to strategic workflows, and build the cultural infrastructure those tools require by working through building a data-driven HR culture. The terms in this glossary only produce ROI when the systems and practices behind them are operational.