Day 20: Create an HR Metrics Dashboard Narrative

The concept

HR dashboards should help leaders make decisions.

Many dashboards do not. They list metrics, show trends, and include traffic-light colours, but they leave the interpretation to the reader. Leaders scan the numbers, ask one or two questions, and move on.

AI can help HR turn metrics into a narrative. The value is in connecting numbers to business meaning and leadership action.

Plain English

A dashboard says what happened. A narrative says why it matters and what should happen next.

Do not report every metric equally

If every metric receives equal attention, leaders cannot tell what matters.

Some metrics are background. Some are signals. Some require a decision.

Before writing the narrative, decide:

• what is stable
• what moved significantly
• what is above or below target
• what has business consequence
• what requires leadership action

The narrative should focus on signal, not completeness.

Add context

AI cannot know why a number moved unless you tell it.

For example:

• turnover rose because two departments delayed pay review
• time to hire increased because three specialist roles skewed the average
• absence rose in one site after shift changes
• headcount is below plan because approvals were paused
• offer acceptance improved after salary bands changed

Context turns a metric into insight.

Translate metrics into business language

Instead of:

Time to hire is 47 days against a target of 35.

Write:

Critical vacancies are staying open nearly two weeks longer than planned, delaying project capacity and increasing manager time spent recruiting.

The number is still there, but now the implication is visible.

Frame asks as decisions

A weak ask says:

Leadership should note the turnover trend.

A stronger ask says:

Approve a targeted retention review for operations and customer service before the next pay cycle.

Leaders can act on decisions. They cannot act on general concern.

Use risk flags

Every narrative should identify the metric that most needs attention if the trend continues.

The risk flag should explain:

• the metric
• the trend
• the business consequence
• the time horizon
• the decision needed

This helps leaders prioritise.

Use exception reporting

Monthly reports should not repeat the same full dashboard if little has changed.

An exception report shows:

• what moved materially
• why it moved
• whether action is needed
• who owns the follow-up
• what is within expected variation

This trains leaders to pay attention to signal rather than scan a familiar pack.

Separate diagnosis from recommendation

Leaders need to understand both what HR believes is happening and what HR recommends doing.

Diagnosis answers:

• What changed?
• Where is it concentrated?
• What is driving it?
• How confident are we?

Recommendation answers:

• What decision is needed?
• Who should own it?
• What is the expected impact?
• What happens if we do nothing?

AI can help write both, but do not let a polished recommendation outrun the evidence.

Add confidence levels

Not every metric supports the same level of certainty.

Use confidence language:

• high confidence: multiple sources point to the same cause
• medium confidence: plausible cause with partial evidence
• low confidence: early signal requiring investigation

This makes the narrative more credible. Executives can handle uncertainty when it is named clearly.

Prepare for questions

Before presenting, ask AI to generate likely leadership questions.

Common questions include:

• What is driving this?
• Is it concentrated in one area?
• What will this cost?
• What happens if we wait?
• Is this within normal variation?
• What decision do you need from us?

Prepare concise answers. A good narrative anticipates challenge.

Do not hide uncertainty

Some people data is messy. Headcount definitions vary. Absence coding may be inconsistent. Turnover may be skewed by one small team. Time-to-hire can be distorted by a few outliers.

Name these limitations clearly.

For example:

Time to hire is above target, but the average is heavily affected by three specialist roles. The median is closer to target, so the issue appears to be specialist hiring rather than the whole process.

This kind of nuance improves credibility. It also prevents leaders from overcorrecting.

Link metrics to workforce planning

HR metrics become more strategic when connected to future capacity.

Ask:

• Does turnover affect critical roles?
• Does hiring speed affect growth plans?
• Does absence affect service delivery?
• Does internal mobility support succession?
• Does engagement predict retention risk?

This moves the dashboard from historical reporting to workforce planning.

End with the decision

Every executive narrative should end with the decision or action needed.

If there is no decision, say that the metric is being monitored. If there is a decision, make it explicit.

Leaders should not have to infer what HR wants from them.

Choose the right visual support

The narrative does not replace visuals. It tells you which visuals matter.

Use charts only where they support the story:

• trend line for a meaningful movement
• segment view for a concentrated issue
• funnel view for hiring friction
• comparison view for target versus actual
• heat map for risk concentration

Do not include a chart because it is available. Include it because it helps the decision.

Create a follow-up rhythm

After leadership agrees an action, track it.

For each ask, define:

• owner
• first step
• due date
• metric or signal
• next leadership update

This prevents HR dashboards from becoming recurring commentary with no accountability.

Use AI as a rehearsal partner

Before presenting, ask AI to challenge the narrative from the viewpoint of the CEO, CFO, COO, or divisional leader.

This can reveal weak assumptions, unclear asks, or cost questions you have not prepared for. Treat the challenge as rehearsal, not criticism.

It also helps you tighten the story before the room does it for you. The best HR data conversations feel prepared, focused, and decision-ready.

That discipline makes people data harder to ignore.

Today's practice

Use your real HR dashboard. Run the prompt with context.

Ask:

1. What is the headline story?
2. Which metric is a true signal?
3. What context explains the movement?
4. What decision do leaders need to make?
5. What should stay out of the narrative?

By the end, you should have a sharper HR data story and a clearer leadership ask.