01
Why this matters commercially
Inconsistent decisions make outcomes difficult to interpret. When results improve or deteriorate, the business cannot tell whether the change came from the market, the mechanism or the individual.
That ambiguity is expensive: it slows learning, makes accountability feel arbitrary, and encourages leadership to centralize decisions again.
02
Why the visible explanation may be incomplete
The visible explanation is usually inconsistent discipline or training. Before accepting it, it is worth testing whether people are working from different information, different priorities or different definitions of an acceptable trade-off.
03
Explanations that remain plausible
Priorities are stated but not ranked
When speed, margin, quality and relationship are all important, each layer resolves the conflict using its own local logic.
The information available differs by layer
Frontline decisions are made without context that exists two levels up, and are then judged as if that context had been available.
No shared reasoning method exists
Teams have rules for routine cases and nothing for the exceptions, which is precisely where consequential decisions live.
Review is inconsistent
Some decisions are examined closely, others never, so the organization does not accumulate a common sense of what worked.
04
Questions that would distinguish between them
- 01Take one recurring trade-off. Ask three layers how they would resolve it, and compare the reasoning rather than the answer.
- 02Which decisions are reviewed after the fact, and does anything change as a result of that review?
- 03What information does a frontline decision-maker lack that would change their choice?
- 04Is there an agreed ranking of priorities when two of them conflict?
05
Where the work may sit
Architecture
Defining the decision categories, who owns each, and the trade-off ranking that should govern them.
Capability
Building a shared reasoning method so different people reach comparable conclusions from comparable evidence.
Leadership
Using review to teach: examining a small number of consequential decisions consistently rather than all decisions occasionally.
06
What cannot be concluded without evidence
Variation is not automatically a problem — some of it is appropriate local adaptation. Determining which variation is costly requires evidence from your own decisions.