The four levels of leverage

Part one

The four levels

Level 1 · produces delivery

Is the work actually being done, reliably, in real conditions?

Level 2 · produces speed and scale

Are activities organised into a repeatable system, or carried by individuals?

Level 3 · produces sustainability

Will this survive the next budget, transfer and election?

Level 4 · produces transformation

Do people want this, own it, and believe it is possible?

Part two

Law one — the levels multiply, they do not add.
A result is the product of what each level contributes. That is why ImpactZ scores your initiative with a weighted geometric index rather than an average: an average hides a zero, a product cannot.
Law two — below a threshold, returns are not small, they are zero.
Any level scoring below 35% of its range is flagged as a zero. Spending at other levels multiplies against it.
Law three — visibility runs opposite to power.
The countable level is the weakest lever; the level that decides what is even considered possible is the hardest to photograph.
Law four — leverage equals size times movability.
A huge constraint you cannot move this year is not your first move. ImpactZ ranks priorities by gap × level weight × movability, which is how it names the binding constraint.

Part three

Step	Requires	Fails as
Capacity → Capability	Opportunity, freedom to act, resources, working support systems	Trained people with no chance, permission or tools to use it
Capability → Performance	Motivation, agency, commitment, persistence, accountability	People who can act but do not, because nothing makes acting rational
Performance → Results	Supply chains, referral paths, infrastructure, demand, coordination	Good individual performance inside a broken ecosystem
Results → Durable value	Institutional home, financing continuity, ownership, identity	A result that disappears with the project that produced it

Part four

01
Training as the default answer
A skills response to a problem that was never a skills problem.
02
Form over function
Adopting the visible shape of a good system without its working mechanism.
03
Piloting with no institutional home
A successful pilot nobody is mandated or funded to continue.
04
Measuring the easiest level
Counting activities because they are countable, not because they matter.
05
Working above the binding constraint
Improving a level that is not what is stopping the result.
06
Ignoring friction
Adding capability to a system that punishes using it.
07
Treating attendance as commitment
Presence recorded, belief unchanged.
08
Funding activity, not conversion
Budget buys inputs and nothing buys the step from capability to performance.
09
One clock for four levels
Judging a paradigm shift on a quarterly reporting cycle.
10
Individual memory instead of institutional memory
Learning that leaves when the person leaves.
11
Replacing ownership with compliance
Imposed change protected by nobody.
12
Declaring success at output level
Closing the report before the result was ever produced.

Part five