01
HOW IT WORKS
The method is developed through use.
We begin with the intended outcome, reason through what the problem requires, build the logic, test it on real data and refine it until it is defensible and useful.
Iterate and refine based on evidence
New insights inform the next iteration
STARTING POINT
Begin with what the data should make possible.
The outcome can be clear even when the method is not.
02
Prioritise or rank options.
03
Forecast what may happen.
04
Make recurring reasoning repeatable.
DEVELOPMENT
The analytical path is specific to the question.
SELECTED PATH / FORECASTING
Selecting another question changes the path itself, not just the labels.
REAL-DATA TESTING
The logic has to survive contact with the data.
01
Observe behaviour
Test the logic on real data and examine what actually happens.
02
Find where it breaks
Look for edge cases, contradictions and unintended consequences.
03
Refine the reasoning
Adjust the approach based on what the data shows, not what was expected.
CONDITIONAL GATE
Is it ready for the intended use?
Validation is a decision point, not a ceremonial final step.
VALIDATE
Use it
It meets the bar for the intended use.
Iterate again
More work is needed to be confident.
Stop where the evidence says to stop
The data does not support further use right now.
ENDPOINTS
The work can end in two equally valid places.
A working internal solution
A purpose-built solution for your team, used in your environment, that solves the problem at hand.
Validated logic ready for integration
The analytical logic, tested and documented, ready to be integrated into your systems.