Nearly every organisation we have worked with has an old dashboard that was built, watched for a few weeks, and then forgotten. It is usually one of three reasons.
It started from the data, not the decision
Build a dashboard from "what data do we have" and you get a set of charts that are correct and irrelevant. Start from "what decision is made weekly" and you usually end up with three or four numbers.
Its numbers are arguable
If a meeting debates whether a figure is right, the dashboard is dead. Every metric needs a written definition everyone has agreed to — especially at the edges: returns, cancellations, discounts.
Someone has to open it
Expecting people to check a dashboard daily is unrealistic. The system should message you when something leaves its normal range. A dashboard is for investigation, not for discovery.
- 01Write down three decisionsList the most important recurring weekly decisions.
- 02One number eachWhat is the smallest number that makes that decision possible?
- 03Write the definitionOne sentence, including the edge cases.
- 04Add an alertDefine the normal range and announce departures from it.
A dashboard with four numbers that gets looked at every morning is infinitely more useful than one with twenty charts opened once a month.