A dashboard built backwards
Anatomy of a deck that answered nothing — and the decision-first rebuild that fixed it.
Most dashboards are built backwards. They start from “what metrics can we show?” instead of “what decision does someone need to make?” — and a dashboard that displays everything answers nothing.
We tore down a seven-panel weekly trading dashboard. Six panels were metrics-because-we-have-them: nice gauges, no decision attached. One panel — buried, small — actually drove the weekly reorder call.
The rebuild inverted it: lead with the decision (what to reorder, how much), show only the figures that move that call, and put the provenance one click away. Same data, a fraction of the panels, and for the first time the trading team opened it on a Monday.