A dashboard that nobody acts on is a screensaver with extra steps. After a year of building KPI reporting for marketing and executive teams, I’ve become convinced that the difference between analysis that gets nodded at and analysis that changes a decision is almost never the maths. It’s the story.
Start from the decision, not the data
The most useful question I ask before building anything is: what decision should be different after someone sees this? If the answer is “none, it’s for awareness”, that’s a red flag. Awareness dashboards grow stale in a week. Decision dashboards get opened every Monday.
When we rebuilt lead KPI reporting at work, we didn’t begin with the data model. We began by listing the recurring meetings where campaign performance was discussed, and the specific calls made in each one: extend this offer, kill that segment, shift budget between channels. Every view on the final dashboard maps to one of those calls.
One number per sentence
Executives read dashboards like headlines. If a tile needs a paragraph of explanation, it belongs in the analyst view, not the executive view. My rule: each visual should support exactly one sentence a stakeholder could repeat in their next meeting. “Conversion in the north-east segment doubled after the location-trigger launch” travels. A 12-series line chart does not.
Anticipate the “why”
Every headline number provokes the same follow-up: why? Good data storytelling pre-answers it. That’s what drill-throughs are for. The headline says what changed, the next click says which segments, channels or offers drove it. If your audience has to email you to get the “why”, the story is unfinished.
Analysis only counts when it survives contact with a commercial meeting.
None of this is decoration. In telecom, a campaign insight that lands a day late is money not made. The story structure of decision first, one number per sentence, and why on the next click is what makes the numbers move at the speed of the business.