Culture Metrics for the Board — Stop Reporting the Weather

Most boards do not skip the people section because they think people don't matter. They skip it because of how the numbers arrive: late in the agenda, in a different format from everything else, with no target, no owner and no decision attached. A revenue line that arrived like that would be skipped too. The fix is rarely a more persuasive HR director. It is making each people metric carry the same four properties that every financial metric in the same pack already has.

We have sat through enough board meetings to recognise the moment. The finance section runs for forty minutes and every number gets interrogated. Then the people section begins, usually after the coffee break, and the room changes register. Directors nod, someone remarks that the engagement score looks healthy, and the chair moves on to any other business. Nobody is being dismissive. The section simply gave them nothing to do.

"It was on slide nine"

The chair of a family-owned group in the Gulf told us about the quarter that changed how their board reads the people pack. Two regional general managers resigned within three weeks of each other, and with them went the relationships behind a large share of the group's recurring contracts. At the next meeting a director asked, with some heat, why nobody had warned the board.

Somebody had. Regretted attrition in that region had been climbing for three quarters, and the chart sat in the pack every time, on slide nine of fourteen, between a training-hours table and a photo from the annual family day. It had no target line, no owner, and no sentence explaining what it would cost if the trend continued. The chair's own conclusion was blunt: "We didn't miss the data. We were never asked to decide anything about it."

That is the whole pattern in one sentence. People data reaches most boards as information, while financial data reaches them as a decision.

The Board-Grade Test

We use a simple test with leadership teams preparing a board pack. A metric is board-grade when it has four properties, and they are exactly the properties any CFO would insist on for a revenue or margin line.

Run your current people pack through those four columns and most items fail on the last two. That is not a data problem. Those metrics were chosen because they were available, not because anybody planned to act on them.

Owner

A metric owned by a function is owned by nobody in particular. Regretted attrition in the commercial team belongs to the commercial head, with HR as the steward of how it is measured. The distinction matters at board level because directors ask questions of people, and a question about sales attrition directed at "HR" lands on the one executive who cannot fix it. If your executive leadership team shares the enterprise agenda, this is one of the first places it shows.

Target

Without a threshold a board cannot tell good from bad, so it defaults to "looks fine". Agreeing the range in advance is also a discipline for the executive team, because it forces a conversation about how much turnover the business can actually absorb in a critical role, which is almost never the same number it can absorb everywhere else.

Line to the outcome

This is where people metrics lose most of their weight. A board thinks in revenue, margin, risk and capacity. "Attrition rose" is abstract; "the two people who hold our largest accounts in the Eastern Province have left or are at risk" is not. Translate each people metric into the business quantity it protects, even roughly, and the conversation moves from HR housekeeping to commercial exposure.

Decision

The last property is the one boards respond to most. For each metric, write down what the board would do if it crossed the threshold: approve a retention budget, pause a regional expansion, ask for a succession plan on a named role, commission a culture review. A metric with no possible decision attached is context, and context belongs in an appendix.

Here's the contrarian part

The instinct, once a board starts taking people data seriously, is to give it more of it. Twenty metrics, a dedicated dashboard, a longer slot on the agenda. That usually makes things worse, because financial packs work precisely because they are ruthlessly edited and nobody shows a board every ledger account.

Present three to five people metrics, no more, each one passing the Board-Grade Test, and move everything else to an appendix directors can read if they choose. Then put those few inside the strategy discussion rather than in a section of their own at the end. If the strategy depends on opening a new market, the capability and hiring numbers for that market belong next to its revenue plan, since that is the only place they mean anything. It is the same logic as why strategy execution fails: the links only hold when they are read together.

The second contrarian point is about engagement scores. A single engagement index is the easiest metric for a board to nod past. It moves slowly, it has no obvious owner, and nobody knows what decision a two-point change calls for. The culture signal that does belong at board level is the gap between the culture you declared and the behaviour people actually report, measured on a cadence and tied to a named risk. Slogans on the wall describe exactly the gap a board should be watching, and organizational culture drives performance, which is the best argument for reporting it like a performance variable.

Why it works

None of this asks a board to become sentimental about people. It asks the people pack to meet the standard the rest of the pack already meets.

Kaplan and Norton made the underlying argument in the early 1990s with the balanced scorecard: financial measures report the past, and a board that reads only them is steering by the wake. Their learning-and-growth perspective was designed to sit alongside finance, not behind it. Human capital reporting has moved in the same direction since. ISO 30414 set out guidelines for internal and external human capital reporting in 2018, and in 2020 the US securities regulator began requiring listed companies to describe their human capital resources where material to the business. Neither tells a board which metrics to use. Both treat people data as governance material rather than HR housekeeping, and that shift is the one that matters here.

There is a warning in the same literature worth taking seriously. Goodhart's law, in its popular form, says that a measure which becomes a target stops being a good measure, and it applies to people metrics with particular force: tie a manager's bonus to an engagement score and you will get a better score, not a better team. That is why the Board-Grade Test asks for a decision rather than an incentive. The board uses the number to decide; it does not pay people to move it. It is also why this work belongs in corporate governance rather than in anyone's performance review.

A practical checklist

• List every people metric in your current pack and score each one against the four properties. Anything scoring below three moves to the appendix.
• Name one owner per metric, a line executive wherever possible, with HR stewarding the definition and the data.
• Agree thresholds with the executive team before the next period starts, with different ranges for critical roles and for everyone else.
• Translate each metric into the business quantity it protects: contracts, capacity, delivery dates, regulatory exposure or cost.
• Write the decision each threshold would trigger. If you cannot write one, you have found a metric for the appendix.
• Move the surviving metrics into the strategy section, next to the plan each one supports.
• Report the gap between declared and lived culture rather than a single engagement index, and show at least three readings of it.

Ask yourself

• Which people metric in our last board pack could the board have acted on, and what would that action have been?
• Who, by name, answers to the board if regretted attrition in our most critical team doubles?
• If our largest strategic bet fails for lack of people, which number would have warned us first, and is it in the pack today?
• Do we set people targets before the period, or explain the numbers after it?
• If we removed the people section entirely, would any director notice at the next meeting?

The takeaway

Boards take seriously whatever arrives in a form they can decide on. Give each people metric an owner, a target set in advance, a line to the outcome it protects and a decision it can trigger; cut the pack to the few that pass; and put them inside the strategy discussion. The people section stops being the weather report the moment it starts asking the board for a decision.