Begin with one decision that cannot wait

Choose the decision the meeting needs to improve this week. It might be where to put limited delivery capacity, which sales opportunities need attention, or whether a cash concern needs action. The decision gives every number a purpose.

If the team cannot name a decision, do not start by building a new report. Ask what has been difficult to see or agree in the last working week. A real disagreement is often a much better starting point than a request for ‘more visibility’.

Ask five questions of every metric

Use the same short challenge for the few measures on the agenda. It makes the discussion more disciplined without requiring a data team.

  • What decision will this number change?
  • What exactly does it include and exclude?
  • How current is it, and what could have changed since it was updated?
  • Who owns the underlying definition and can explain an exception?
  • What action follows if the number is surprising?

Look for the exception, not just the average

An average can make a busy week look calm. A service team may have an acceptable average turnaround while two important jobs have stalled. A sales conversion rate may look steady while a valuable customer enquiry has had no reply.

Show the detail that helps somebody act: the overdue job, the unallocated task, the customer waiting for a decision or the order with an unusual margin. The metric is the prompt; the exception is often where the conversation becomes useful.

Separate a data issue from a performance issue

When figures do not match, resist treating the discrepancy as bad performance. One report may use a different cut-off, status or source. Put the purpose, source, timing, rule and owner next to both figures before deciding which one matters for the meeting.

The Government Data Quality Framework describes quality as fitness for purpose. That is a helpful practical standard for a small business: a number does not need to be perfect to be useful, but the team needs to understand whether it is fit for the decision they are about to make.

Assign a small repair when the definition is unclear: make a field required, agree a status boundary or name the system of record. Do not paper over the difference by adding another chart.

Finish with a short action record

Capture the action, owner and due date before the meeting moves on. Keep it beside the numbers for the next review. Over a few weeks, this tells you whether the measure gave an early enough signal and whether the response actually helped.

Review the view itself every month. Remove a metric that no longer changes anything. Add a breakdown only when the team keeps asking the same follow-up question. This keeps the dashboard close to how the business is really run, rather than turning it into a permanent collection of charts.

The aim is not more reporting. It is a calmer weekly rhythm: a small number of trusted measures, visible exceptions and decisions that have an owner.

A useful dashboard earns its place in the action taken after the meeting, not in the number of charts it contains.

Sources and further guidance

This article is practical operational guidance. The Government Data Quality Framework provides a fuller reference for thinking about whether data is fit for purpose.

Make the next step useful.

Bring us the workaround, the repeated report or the AI idea that has not found an owner yet.

Discuss a challenge