A dashboard can report every agreed measure and leave leadership unable to decide what to do next. Accuracy does not settle whether the company should act, wait or investigate. The first task is to identify which judgement the evidence supports.
The table below separates four conditions that a single red or green status can obscure. Use it to interrogate a material measure before approving the response.
Read the condition before choosing the response
Illustrative prompts for interpreting a material measure. The target, time horizon and reliability of the evidence must be explicit.
| Condition | Question for leadership | Evidence that matters |
|---|---|---|
| Improving; target still missed | Will the current response close the gap in time? | Remaining gap, credible trajectory and constraints |
| Target met; performance deteriorating | How much room remains before intervention is necessary? | Rate of deterioration and time needed to respond |
| Reported result stable; coverage incomplete | Could the missing observations change our conclusion? | Extent and pattern of missing data |
| Several limits under pressure | Do these movements reflect a common exposure? | Dependencies, combined consequence and response options |
A threshold prompts review. Management and directors remain responsible for judging the response.
Treat confidence in the data as a separate question
A result can improve and still require intervention. It can meet the current target while deteriorating quickly enough to threaten the next commitment. A stable average may hide incomplete coverage. A single status colour should not imply that these judgements are interchangeable.
Put the reporting basis close to the measure: period, business boundary, unit and definition. An unavailable value is not zero. An estimate is not a reconciled observation. If missing information could change the proposed decision, explain the uncertainty on the page where the decision is being discussed.
Keep detailed calculations accessible behind the summary. Brevity becomes useful when it preserves the qualifications that matter and keeps them within the reader’s attention.
Put the response and the decision date beside the exception
An explanation of why performance deteriorated is incomplete without management’s proposed response. Show the alternatives, the consequence of waiting and who can authorise action. The person who verifies a measure may not control the operating decision; both responsibilities should be visible.
The latest useful decision date may precede the next reporting date. If temporary capacity must be secured within two weeks, a quarterly review cannot provide the necessary authority in time. A dashboard should make that conflict apparent before it becomes an explanation for inaction.
Thresholds can prompt examination without determining the answer. One isolated breach may be harmless; several indicators approaching their limits may reveal a common exposure. The management judgement lies in the connection between the measures and the business consequence.
Carry the decision into the next review
The next pack should show what happened to the previous decision: whether the action was completed, whether the expected effect emerged and whether the measure improved for a different reason. Without that continuity, leadership can confuse completed activity with an effective response.
Not every indicator needs to produce an immediate decision. Some support routine oversight or reveal an emerging pattern. Separate those roles from the few matters requiring scarce board attention now.
Take one page from the current pack. Ask a director and the accountable executive, separately, what it allows them to decide and what evidence would change their view. A difference in their answers is a useful place to begin the redesign.
The boardroom test is whether a page helps leadership make a better-informed decision. A dashboard earns its place when management can identify that decision and explain how the evidence improves it.
