Value against ease
Both axes run 1 to 5. The lines sit at 3.00, the cut the framework uses. Nothing is rounded before it is placed.
What your weights changed
Where the forty sit
Crossed into the quick-win box
Dropped out of it
Biggest rank moves
Ranked candidates
Every 1–5 cell is editable. Click a row for the case, the conditions and the scoring anchors. Column headers sort.
The framework does not produce a single ranked list on purpose. A composite would sit zone pricing — high value, not our work, a multi-quarter programme — next to a permissions change. The ranking here is a conversation aid; the quadrant, the evidence grade and the tier are the instrument.
Tier follows from the score, not the other way round. Above both lines with evidence A or B is Recommend now; above both lines on a single-source assertion is Recommend pending evidence, because nothing graded C enters Tier 1. High value and low ease is a bigger bet. Out of scope, client-owned, superseded and enabler override the box they land in.
Sixteen of the forty are graded C — a single source, our own inference, or a hypothesis with no number behind it. The instrument is not meant to survive contact with the exports unchanged. When the numbers land, the scores move, and this tool is where you will see what they move.