HR metrics that matter: What to track and why

The HR metrics worth tracking fall into four groups: workforce, talent, capability, and cost. Twelve numbers cover almost every decision an executive team needs from HR. Most dashboards carry forty and get acted on less.

The problem isn’t measurement. It’s that HR reports what’s easy to count rather than what changes a decision.

They report activity instead of consequence.

Training hours delivered, requisitions filled, tickets closed. All true, all countable, none of it answers a question an executive was asking. The dashboard gets built, circulated, and quietly stops being opened.

The second problem is aggregation. Company-wide averages hide everything useful. Turnover of 14% could be evenly spread or concentrated in three teams. Only the segmented version tells you what to do.

Careerminds research on AI-led layoffs found that 40.6% of companies say they would have benefited from more data on employee capabilities. That’s the gap: organizations had headcount data and lacked capability data, then made irreversible decisions with the wrong information.

The base layer. Everything else is read against these.

Headcount and full-time equivalent. Segmented by function, level, and location. Track the trend, not just the point.

Turnover rate. Total, but always paired with the regretted figure below.

Regretted turnover rate. Voluntary exits of people you wanted to keep, divided by average headcount. This is the number that matters. Total turnover includes exits you wanted, and treating those as failures pushes bad decisions.

Span of control. Reports per manager. Where this drifts above eight, manager quality drops regardless of individual capability.

How well people move into and through the organization.

Time to fill and time to hire. Time to fill runs from approval to acceptance and measures the whole process. Time to hire runs from application to acceptance and measures the recruiting team. Executives usually want the first.

Internal fill rate. The share of roles filled by existing employees. Rising means development is producing usable capability. Flat means you’re developing people for the external market. Internal mobility is the mechanism behind this number.

First-year turnover. Exits within twelve months of joining. High figures point at hiring or onboarding, not retention, and the fix is upstream.

Offer acceptance rate. A fast indicator of pay position and employer reputation. It moves before turnover does.

The group most often missing, and the one that predicts the others.

Skills coverage on critical capabilities. How many people can do the things you can’t operate without. Where one person holds a capability, that’s a risk with a name.

Ready-now successors per leadership role. Bench depth. Below 1.5 you’re relying on external hiring. Our guide to the leadership pipeline covers how to size this properly.

Manager effectiveness. Team outcomes, observed behaviors, and self-rated preparedness, cut by manager. The detail sits in our guide to manager effectiveness.

Development participation, distributed. Not total hours. Who gets development and who doesn’t. Careerminds research found that only 17% of remote workers get access to leadership development programs, which is the kind of gap a total-hours figure conceals completely.

Two numbers, both frequently misused.

Cost per hire. Total recruiting spend divided by hires. Useful as a trend, misleading as a benchmark, because organizations count wildly different things.

Total labor cost as a share of revenue. The figure your CFO already tracks. Knowing it lets HR enter the conversation in the language finance is using.

Five rules.

Twelve metrics maximum. Anything longer gets skimmed.

Segment everything. By function, level, and manager. Averages hide the finding.

Pair every metric with a trend. A single number invites debate about whether it’s good. A direction invites a decision.

Lead with the two or three that changed. Put movement at the top and the rest below.

Say what you want done. A metric with no recommendation attached is trivia. State the decision you’re asking for.

Group	Metrics
Workforce	Headcount, turnover, regretted turnover, span of control
Talent	Time to fill, internal fill rate, first-year turnover, offer acceptance
Capability	Skills coverage, bench depth, manager effectiveness, development distribution
Cost	Cost per hire, labor cost as share of revenue

Slow-moving models sell reporting tools and call it analytics. A tool that produces forty charts nobody reads is a more expensive version of the problem.

Regretted turnover, internal fill rate, time to fill, first-year turnover, skills coverage on critical capabilities, bench depth for leadership roles, and manager effectiveness. Regretted turnover and internal fill rate carry the most weight, because together they show whether you’re keeping the people you need and building capability rather than buying it.

HR metrics are measurements. HR KPIs are the small number of metrics tied to a target you’re accountable for. Every KPI is a metric, but most metrics aren’t KPIs. Organizations run into trouble when they treat the full metric set as targets, which spreads attention across figures that were only ever meant as context.

Divide the number of exits in a period by the average headcount for that period, then multiply by 100. Average headcount means the start figure plus the end figure divided by two. Calculate voluntary and involuntary exits separately, and split voluntary into regretted and non-regretted, because a single combined figure hides which direction the problem runs.

Around twelve for executive reporting, covering workforce, talent, capability, and cost. Operational teams need more detail underneath, but the reported set should stay small enough that every number gets discussed. Dashboards with forty metrics tend to be reviewed less often than dashboards with ten, because nobody knows where to start.

Careerminds Workforce Intelligence gives HR skills, role, and capability data for planning decisions, built on 20M people supported through workforce change. We help you measure what changes a decision.

Talk to our team about your workforce data.