Findings arrive ranked, graded and already checked against your limits.

Every metric rendered at equal weight

Charts render every metric at equal weight and leave the ranking to you. The movement that matters is usually small, inside one segment, on a cut nobody filtered to that morning. By the time it is large enough to show at account level, weeks of spend have already passed through it.

Written first, charted second

Each line states what changed against that metric's own baseline range, how much spend sits behind it and which service returned the figure. Prose can rank findings and name the next step; a column of numbers can only line them up. The chart is there when you want to check the shape of a claim, not as the place you start.

Decisions, with their receipts

The shortlist gives each item the change it would make, the budget engine's expected value for it and the Constitution limit that governs it. Reading the brief executes nothing; approval is the only path from a finding to a change. Approve one and the same evidence bundle follows it into the action log.