The HR board report: what to include and what to leave out
Boards do not want to know how busy HR was. They want to know whether the people plan is on track, what it costs, and what could go wrong. One pack does that; most packs do the other thing.
Report five numbers every time, one page of narrative, and a risk section - quarterly. The five: headcount and movement, regretted attrition, absence, time-to-hire, and total people cost. The narrative says what changed and what you are doing about it. The risk section - tribunal exposure, expiring checks, overdue training - is the part boards thank you for and most packs miss. Everything else is appendix.
What does the board actually want from HR?
Three things: evidence the people plan supports the business plan, early warning of anything that could cost money or reputation, and confidence that the basics are controlled. A board pack is not a performance review of the HR function - listing initiatives, training sessions run and policies updated tells the board how busy you were, not whether it worked. Every line in the pack should survive the question "what decision does this inform?"
The five numbers to report every time
1. Headcount and movement. Where you are against the plan the board approved: starters, leavers, open roles. Trend, not snapshot.
2. Regretted attrition. Total turnover mixes people you were glad to see go with people you fought to keep - the board needs the second number. Report both, benchmarked against your own trailing year (our turnover calculator does the arithmetic and the definitions).
3. Absence. Days per employee and the trend - and be honest about measurement: the ONS records 4.4 sickness days per worker from survey data while CIPD's employer-record figure is 9.4, and the gap is absence nobody wrote down. If your number looks better than the national figures, check your recording before you celebrate in front of the board.
4. Time-to-hire and cost-per-hire for the roles that matter. A board that understands hiring friction funds it properly (cost-per-hire calculator).
5. Total people cost. Salary bill, employer NI and pension, and the drift since last quarter - the number the CFO will reconcile against, so agree the definition with finance before the meeting, not during it (the true-cost calculator shows what sits on top of salary).
The risk section most packs miss
Boards carry legal responsibility, so give them the risk picture in plain numbers: open ER cases and their stage, tribunal exposure if any claim is live (what a tribunal actually costs), right-to-work and DBS checks expiring in the next quarter, mandatory training compliance, and evidence the harassment-prevention duty is being met - a live register beats an assurance. If these numbers come from systems rather than a hand-built spreadsheet, say so: provenance is part of the comfort.
Engagement: report the trend, not a trophy
An eNPS or engagement score belongs in the pack - as a trend line against your own history with the response rate beside it. Resist benchmark theatre: published eNPS benchmarks vary too much by survey method to compare against honestly, and a board sold a flattering benchmark once will discount the metric forever. A falling score with a rising response rate is a finding; present it as one, with what you are doing about it.
What to leave out
Activity metrics (courses run, policies updated) - appendix at most. Individual-level data - survey verbatims traceable to a person, mood scores, anything where a small team cut identifies someone; if the cut is small enough to identify a person it does not belong in a board pack, full stop. And manager league tables: ranking managers by team score teaches managers to farm the score. The same privacy line CoDash draws in the product - aggregates above a minimum group size, never individuals - is the right line for a boardroom too.
Cadence and format
Quarterly full pack; the five numbers refreshed monthly if the board meets monthly. One page of narrative up front - what changed, why, what you are doing - then the scorecard, then risk, then appendices nobody is obliged to read. The discipline that makes packs land is sameness: the same numbers, defined the same way, every quarter, so the trend is the story. If producing it takes days, the problem is the plumbing, not the pack - the numbers above should fall out of your HR system on demand.
Frequently asked questions
What should HR report to the board?
Five core numbers - headcount and movement, regretted attrition, absence, time-to-hire and total people cost - plus a short risk section (tribunal exposure, expiring right-to-work and DBS checks, overdue mandatory training) and one page of narrative. Decisions and risk, not activity.
How often should HR report to the board?
Quarterly as the full pack, with the same five numbers refreshed monthly if the board meets monthly. The discipline is reporting the same numbers every time so trends are visible - swapping metrics between packs hides more than it shows.
Should individual employees ever appear in a board report?
Only in genuine exception cases the board must own, such as a serious ER matter with material risk. Survey scores, mood data and feedback must stay aggregated - if a cut of the data is small enough to identify someone, it does not belong in the pack.
What is a good eNPS score to report?
Report the trend against your own history, not a benchmark from a vendor deck - published eNPS benchmarks vary too much by method to compare against honestly. A stable or rising score with a decent response rate says more than any absolute number.
Worth reading alongside
Keep reading: All 66 HR guides · 30 free templates · 24 calculators
A board pack that assembles itself
Headcount, turnover, absence, hiring funnel, engagement trend and the compliance registers - CoDash keeps them live all quarter, exports them board-ready, and can email the pack to you on schedule.