The HR pack goes to the board with forty numbers on it. Headcount by department, training hours delivered, applications received, diversity ratios, average tenure, three colour-coded engagement charts. The chief executive reads the first page and asks a question none of it answers.
That is not a data problem. It is a framing problem. HR reports what it can count. Executives want to know about capacity, cost and risk.
What a CEO is actually asking
Strip the phrasing away and executive questions about people reduce to three.
- Can we deliver? Do we have the people the plan requires, and can we get the ones we are missing in time
- What is it costing? Workforce cost against output, and whether that ratio is moving the right way
- What is about to break? The team, the manager or the skill where something is going wrong before it shows up in results
Every metric on the dashboard should answer one of those. Training hours delivered answers none of them. Training completion against a capability the plan depends on answers the first.
Nobody has ever changed a decision because average tenure moved from 3.1 to 3.3 years. Put numbers on the page that someone would act on.
The eight metrics that earn a place
| Metric | Formula | What it tells the CEO |
|---|---|---|
| Regretted attrition | Regretted leavers ÷ total leavers | Whether you are losing the people you wanted to keep, which total attrition hides |
| Attrition by manager | Leavers on a team ÷ average team headcount | Where the problem actually sits. Almost always a small number of teams |
| First-year attrition | Leavers within 12 months ÷ hires in that period | Whether hiring and onboarding are working, visible long before annual attrition moves |
| Time to fill, critical roles only | Days from requisition approved to offer accepted | Whether the plan is gated by hiring speed |
| Cost per hire | (Internal + external recruiting cost) ÷ hires | The real cost of growth, including the recruiter salaries most companies leave out |
| Revenue per employee | Revenue ÷ average full-time equivalent | Whether headcount growth is producing proportional output |
| Internal mobility rate | Internal moves ÷ average headcount | Whether people can grow without leaving, which predicts attrition a year out |
| Manager span and vacancy in key roles | Reports per manager, and days key roles sit empty | Where the organisation is structurally fragile right now |
Eight is not a magic number, but a page that holds more than about ten numbers stops being read and starts being filed.
The metrics people calculate wrong
Four of these are routinely computed in ways that make comparison meaningless.
Attrition rate. The formula is leavers in the period divided by average headcount for that period, multiplied by 100. The common error is dividing by opening or closing headcount, which distorts badly in a growing company. Average headcount means (opening + closing) ÷ 2 at minimum, or a monthly average if you have it.
Then split it. Voluntary and involuntary are different phenomena and mixing them makes the number useless. Within voluntary, mark regretted and non-regretted, because losing ten people you were managing out is a success and losing three you wanted to keep is not.
Time to hire and time to fill. These are different measures and are constantly used as synonyms. Time to fill runs from the requisition being approved to the offer being accepted, and it tells you how long the business waits. Time to hire runs from a candidate entering the process to that same candidate accepting, and it tells you how fast your process moves once someone is in it. A CEO wants time to fill. A recruiting lead needs both.
Cost per hire. Total internal and external recruiting costs divided by the number of hires in the period. External is easy: agency fees, job board spend, assessment tools, referral bonuses. Internal is the part that gets dropped: the loaded cost of the recruiting team, the ATS licence, and interviewer time. A cost per hire that excludes internal cost is not comparable to anyone else's, and usually understates the real figure by a wide margin.
Revenue per employee. Revenue divided by average full-time equivalent, not by headcount. If you use contractors heavily, decide whether they are in or out and stay consistent, because switching between definitions turns a flat trend into a rising one.
Leading and lagging indicators
Attrition is a lagging indicator. By the time it moves, the decisions that caused it were made six to nine months earlier. A dashboard built entirely on lagging measures reports history accurately and prevents nothing.
- Leading: internal mobility rate, share of employees with no manager one to one in 30 days, offer acceptance rate, first-year attrition, open critical roles, manager span
- Lagging: total attrition, revenue per employee, absence rate, annual engagement score
Offer acceptance rate deserves particular attention because it moves early and is easy to collect. When it falls, something has changed in your compensation position, your interview experience or your reputation, and all three are worth knowing about before the hiring plan slips.
Metrics that look important and are not
- Training hours delivered. Measures activity, not capability. Nobody acts on it
- Total applications received. Rises when the job ad is vague. Quality of shortlist is the number that matters
- Average tenure. Moves slowly, means different things in a growing and a shrinking company, and rarely changes a decision
- HR headcount ratio. Interesting for benchmarking once a year, not a dashboard metric
- Engagement score with no free text. A temperature reading with no diagnosis attached
- Any metric with no owner. If nobody is accountable for moving it, it is reporting rather than a metric
The test is simple. If the number moved twenty per cent next month, would anyone do something differently? If not, take it off the page.
How to build an HR dashboard people read
- One page. If it needs scrolling it needs cutting
- Trend, not snapshot. Twelve months of history against every number, because direction is what gets acted on
- Segment where the average hides the answer. Attrition by manager, tenure band and location almost always tells a different story from the company figure
- A threshold agreed in advance for each metric, set while everyone is calm rather than after a bad quarter
- The definition and the source system written next to each metric, so nobody recalculates it differently next quarter
- One sentence of commentary per number explaining what changed and what is being done. The commentary is what makes it a report rather than a chart
Pair metrics that can be gamed. Time to fill on its own encourages rushed hiring, so show it beside first-year attrition. Cost per hire alone encourages cutting sourcing spend, so show it beside offer acceptance rate.
Where to build it
Three practical options, in ascending order of effort.
- Your HRMS reporting module. Fastest to stand up and the data is already there. The limit is that it only knows what the HRMS knows, so revenue per employee needs manual entry. Most Indian platforms including greytHR, Keka, Darwinbox and Niyuk, our parent company, ship dashboards of this kind
- A BI tool such as Power BI, Looker Studio or Metabase reading from your HRMS and finance system. More work to set up, and the only option that can put people data and revenue on the same page automatically
- A spreadsheet. Genuinely fine below about two hundred employees, provided one named person owns it and the formulas are written down. Most companies over-engineer this step
The tool matters less than the definitions. A spreadsheet with agreed formulas and a named owner beats a live dashboard nobody trusts because two teams calculate attrition differently.
How to present it to the board
Lead with the answer. The first line is what changed and what you are doing about it, not a description of the chart. Then the numbers, then the detail if anyone asks for it.
Bring the bad number yourself. An executive who discovers a problem in the appendix trusts the next pack less. An HR lead who opens with "regretted attrition in engineering doubled this quarter, here is what we found and here is what changes" is doing the job.
And translate into money where you honestly can. "First-year attrition of 24 per cent means we replaced 31 people we had already paid to recruit and train" lands differently from a percentage on its own.
Frequently asked questions
What are the most important HR dashboard metrics?
Regretted attrition, attrition by manager, first-year attrition, time to fill for critical roles, cost per hire, revenue per employee, internal mobility rate, and vacancy in key roles. Each answers one of the three questions executives ask: can we deliver, what is it costing, and what is about to break.
How do you calculate attrition rate?
Divide the number of leavers in a period by the average headcount for that period and multiply by 100. Average headcount means opening plus closing divided by two, or a monthly average if available. Dividing by opening or closing headcount alone distorts the figure in a growing company. Split voluntary from involuntary, and regretted from non-regretted, or the number cannot be acted on.
What is the difference between time to hire and time to fill?
Time to fill measures from requisition approval to offer acceptance and shows how long the business waits for a role. Time to hire measures from a candidate entering the process to that candidate accepting, and shows how fast the process moves. Executives want time to fill; recruiting teams need both.
What is the cost per hire formula?
Total internal recruiting costs plus total external recruiting costs, divided by the number of hires in the period. External covers agency fees, job boards, assessments and referral bonuses. Internal covers the loaded cost of the recruiting team, the ATS licence and interviewer time. Excluding internal cost is the most common error and understates the real figure substantially.
How many metrics should an HR dashboard have?
Around eight, and rarely more than ten. Beyond that the page stops being read. Depth belongs in an appendix that answers questions raised by the front page rather than on the dashboard itself.
What is the difference between leading and lagging HR indicators?
Lagging indicators such as total attrition and revenue per employee report what has already happened. Leading indicators such as offer acceptance rate, internal mobility and first-year attrition move earlier and give you time to act. A dashboard built only on lagging measures records history accurately and prevents nothing.
Cut the pack to eight numbers that answer the three questions, write the formula next to each one, show twelve months of trend, and put a sentence of commentary beside every figure. That page will be read. The forty-metric version will not.
