KPI tree
Decomposes an outcome you care about into the drivers that actually produce it, so people working on inputs can see how their number reaches the top one.
The printable canvas and the handoff into Claude Code are part of a paid plan. See what a plan includes. It needs the Claude desktop app on this machine, and your team and the prompt library already installed in that project — we cannot see your disk, so open it there.
What it is
A hierarchical decomposition of one outcome metric into the drivers that mathematically produce it, continuing down until each branch reaches something an individual team can influence. Its purpose is to connect daily work to results that are otherwise several removes away.
- Root metric
- The single outcome being decomposed. One per tree, or the structure stops being a tree.
- Decomposition
- Breaking each level into its component drivers, arithmetically wherever possible so the relationships are checkable rather than asserted.
- Leaf metrics
- The bottom of each branch: something a specific team can actually move within a quarter.
- Movability
- Which branches are genuinely influenceable and which are structural. Effort spent on immovable drivers is invisible waste.
- Leverage
- How much the root moves for a realistic change in each driver. Frequently identifies a neglected driver with more effect than the one everyone watches.
How you run it
- Start with one outcome metricRevenue, margin, contribution. One. Multiple roots produce a thicket rather than a tree.
- Decompose arithmetically where you canRevenue equals customers times frequency times basket. Arithmetic decomposition is checkable; thematic decomposition is opinion.
- Keep going until you reach something someone controlsThe bottom of a branch should be a number a specific team can move this quarter. Anything above that is a summary.
- Mark which branches are movableSome drivers are structural and effectively fixed. Marking them stops teams pushing on things that will not move.
- Find the leverageWhich driver produces the biggest change in the root for a realistic movement? That is usually not the one currently getting attention.
The prompt
Run this tool in your own Claude
The short prompt starts your partner against the library on your disk. The long one carries everything with it and needs nothing installed.
Your playbook
It lands in the earliest stage this tool suits. Move it on the Playbook page.
You’ll need
- One outcome metric that genuinely matters
- Understanding of the arithmetic behind it
- Data at each level, or an honest note where there is none
You’ll end up with
- The tree from outcome down to operational drivers
- Which branches are actually movable
- The two or three drivers with most leverage