The strategic KPI on the bowler and the measures on a team’s daily board should be connected, but in most organizations they were designed by different people at different times. A KPI tree makes the connection explicit, and shows each team which branch it moves.
In this guide
Key Takeaways
- →No single team controls a strategic KPI; the tree shows which branch is whose.
- →It also lets the strategy see which branch is failing, and steer there.
- →One definition per measure, shared wherever it appears.
- →Leading drivers under lagging results, so something moves before the KPI does.
- →Three or four levels. Stop where a team can act on the measure.
- →Revisit annually when the strategy changes, not every month.
Why a tree
A strategic KPI such as on-time delivery is the result of many things: production schedule adherence, parts availability, changeover times, shipping cut-offs. No single team controls it. A KPI tree shows which drivers add up to it, so each team can see which branch is theirs and the strategy can see which branch is failing. It’s the same idea as the tree diagram among the seven management and planning tools, applied to measures.
Those seven tools, and what each one is for in a planning season, are covered in The Seven Management and Planning Tools in Hoshin Kanri.
A worked example: an on-time delivery tree
Read the hypothetical tree below as three levels, not seven rows: the result the plan names, the drivers that add up to it, and the measures a team can act on in a shift.
Read the tree below as three levels, not seven rows: the result the plan names, the drivers that add up to it, and the measures a team can act on in a shift.
The right-hand column is the one that makes it useful. Every measure has a home on a board somebody already stands in front of, which is what stops a tree becoming a diagram nobody consults.
| Level | Measure | Board |
|---|---|---|
| Strategic | On-time delivery to customer (%) | Hoshin plan and bowler |
| Driver | Schedule adherence (%) | Site board |
| Driver | Parts available at kitting (%) | Site board |
| Driver | Shipping cut-off met (%) | Site board |
| Team measure | Orders started on schedule today, line 3 | Tier 1 board |
| Team measure | Kits complete at shift start, line 3 | Tier 1 board |
| Team measure | Trucks loaded by cut-off today | Dispatch board |
If a team can’t say which branch of the tree its board moves, the strategy and the board aren’t connected.
Rules for a usable tree
- One definition per measure, shared wherever it appears.
- Leading drivers under lagging results. Parts availability moves before on-time delivery does.
- Stop where a team can act. Three or four levels is usually enough.
- Name an owner for each branch.
- Revisit annually, when the strategy changes, not monthly.
The third rule is the one most trees break. A five- or six-level tree is more accurate and completely unusable, because the bottom level is a measure nobody can influence in a shift. Accuracy below the level of action buys nothing.
Using the tree to steer
The tree makes steering concrete. If on-time delivery is behind and parts availability is the weakest driver, the sites with the worst parts availability make it their quarterly focus, and improvement capacity follows. See Steering Improvement Capacity Toward Breakthroughs, Without Killing Kaizen. Ratio KPIs especially should cascade as drivers rather than as copies of the ratio; see KPI Cascades: When a KPI Should Cascade, and When It Shouldn’t.
What a missing tree costs
Without a tree, improvement capacity is allocated by whoever is most persuasive about their own branch. That is how a site ends up running eighteen competent kaizens on changeovers while its delivery problem is parts: nobody was wrong, and nobody could see which driver was weakest. The tree does not create effort, it redirects effort that is already being spent, which is why it is one of the cheapest interventions in this series and one of the highest leverage.
The second cost is slower and harder to unwind. Where the same measure means slightly different things at tier 1 and on the bowler, every review spends its first ten minutes reconciling numbers instead of deciding anything, and the front line learns that the strategic KPI is head office’s number rather than theirs.
Among published alternatives for holding a shared definition across tiers, OnStrategy lists $200 per month for unlimited users, which removes the question of who gets to see a driver. i-nexus publishes a free Starter tier for up to five users and $28 per user per month at its Standard tier. Test the tree directly in either: can a driver be set up as a contributing measure with a weight, and does the tier 1 view read the same definition as the bowler without anyone re-entering it?
Profit.co publishes two plans, Essentials and Enterprise, quoting against modules and scale rather than a public seat rate, which fits a tree whose leaves are read by many teams and whose root is managed by a few. Hoshin Kanri, OKRs and Goals and Balanced Scorecard sit in Essentials, and the 30-day free trial is enough to build one real tree and put it on two tiers of boards.
One definition, every tier
In Profit.co, strategic KPIs sit on the X-matrix with their correlations, drivers can be set up as contributing key results with weights, and huddles at every cadence share the same definitions. See the lean strategy software page.
Holding the tree as records rather than as a diagram has three practical effects:
- The branches roll up by themselves. A driver set up as a contributing key result with a weight updates the result it feeds, so the tree cannot drift from the numbers.
- A weak branch is visible, not argued. The strategy can see which driver is behind without assembling a pack from three systems.
- The tier 1 board reads the root. One definition means a team can see the strategic KPI its measure feeds, which is the whole point of drawing the tree.
Where those measures are reviewed day to day is covered in Tiered Huddles and the Strategy Review Cadence, and the test of whether a driver genuinely leads the result is in Leading Indicators for a Lean Strategy.
Free Product Demo
Build the tree for your strategic KPIs
A 30-minute walkthrough, tailored to your team.
Book a free demoWhat happens in the call
How your teams set goals, track progress and review results.
Strategy, performance, projects and engagement on one platform, with AI.
A practical path to align teams, execute faster and drive results.
Questions this article answers
A breakdown of a strategic KPI into the drivers that move it, down to measures teams can influence on their daily boards.
Usually three or four levels: stop where a team can act on the measure.
It shows each team which driver of a strategic KPI it moves, and lets the strategy see which branch is failing and steer improvement there.
Annually, when the strategy changes. A tree redrawn monthly stops being a shared language, which is most of its value.
No. A ratio should cascade as its drivers rather than as copies of the ratio, or every level ends up accountable for a number it only partly controls.