Method
The method is simple: do not mix the levers.
Layout, OEE and capacity influence each other. If you do not separate them before calculating, the result cannot be attributed to any specific lever and the recommendation is not defensible.
1

Boundaries

What stays constant across scenarios is declared upfront. Without fixed boundaries, the comparison is not fair and any alternative can be made to look better than the others.

2

Separate levers

Layout, OEE and capacity are analysed in distinct modes. The software does not allow them to be mixed: isolation is structural, not an analyst’s choice.

3

Stress-test

Break-even and sensitivity: at which volume, mix or shift the choice reverses. The recommendation holds even when someone changes the assumptions.

Practical result
A defensible choice
Not “the best solution”, but the solution that holds when someone asks: “what if volumes drop?”
See a real case
What happens without the method
Inflated ROI
attribution to the wrong lever.
Fragile choice
optimal only at current volumes.
Endless meetings
numbers never converge: they measure different levers.