A technically correct answer can still go nowhere.
A risk is correctly identified, but the discussion becomes about whether the control is slowing the business. Credit capacity is created, but utilisation remains weak because the real constraint has moved. A review contains plenty of information, yet no consequential decision is made because ownership remains diffuse.
These are not necessarily failures of expertise.
Between knowing something and acting on it sits an organisational system: hierarchy, incentives, ownership, timing, authority, competing objectives and people who must carry the consequences.
The capability required to navigate that gap is what I think of as organisational judgement.
It is more than reading the room or managing stakeholders. Organisational judgement is the capability to diagnose what is preventing sound reasoning from becoming an owned, executable decision — and then intervene at the right level without distorting the truth.
1. Find the real constraint
One of the easiest mistakes in organisational life is to keep solving the problem that was originally assigned.
The problem may have moved.
Consider credit. An organisation can improve assessment, lender access and sanctioned limits. But if customers are not using that capacity, the constraint may now sit in documentation, transaction economics, sales behaviour, customer friction or lender process.
The credit function can be locally successful while the enterprise outcome remains weak.
The same pattern appears elsewhere. A product problem becomes an adoption problem. A data problem becomes an ownership problem. A technology problem becomes a governance problem.
Expertise naturally deepens around the problem it knows. Judgement asks a different question:
What is actually preventing movement now?
2. Change the evaluative basis
Sometimes an organisation is not arguing about the answer. It is evaluating the problem against the wrong question.
Consider a control failure where the eventual financial outcome is benign. The exposure is recovered. It is tempting to conclude that the system worked well enough.
But recovery and control quality are different questions.
If a transaction can cross an intended boundary without the right exception authority, the absence of loss does not make the decision path sound.
The same shift in evaluative basis can turn:
Sometimes judgement requires challenging the question before arguing about the answer.
3. Make ownership and decision rights explicit
Many apparently technical disagreements become easier to understand when we ask:
Who carries the consequence?
This is particularly visible in risk and governance.
A central Risk function can create the appearance of stronger control by taking ownership of every exception and operational gap. But doing so can weaken the first line, blur accountability and turn Risk into a bottleneck.
The difficult judgement is not whether Risk should intervene. It should. The question is how.
Sometimes the intervention is challenge. Sometimes it is a control requirement or escalation. Sometimes a disputed boundary must move to a forum where the decision right is clear.
The same issue appears elsewhere. Product can end up owning adoption problems that require business behaviour to change. Strategy can become owner of actions merely because it coordinates the review.
When ownership is unclear, organisations often compensate with more meetings and follow-ups.
Judgement distinguishes a coordination problem from an ownership problem.
4. Create movement without manufacturing agreement
Not every important problem requires consensus.
An operating-model change may have a compelling efficiency logic while also affecting customer continuity, commercial outcomes and relationship ownership. Debating until everyone agrees may not improve the decision.
A bounded pilot with explicit rules, metrics, exceptions and escalation paths can turn disagreement into a testable commitment without pretending uncertainty has disappeared.
Enterprise AI creates a similar challenge. A prototype can show that a system produces useful output. It does not establish that an organisation should rely on it.
That requires different questions: Who reviews the output? What can the system act on? Where does human judgement remain necessary? What happens when it is wrong? Who owns the workflow?
At that point, another demonstration may add less value than a well-designed pilot.
Judgement also recognises when movement should stop. If material risk is being hidden, evidence distorted or nobody with legitimate authority will own the consequence, further persuasion is not the answer. Escalation or explicit dissent may be.
Adapt the path, not the truth
This is the boundary I find most important.
Understanding hierarchy, incentives and timing can easily become accommodation to power. A person can become so good at sensing what the organisation will accept that they stop saying what it needs to hear.
That is not judgement.
But repeatedly stating a technically correct position in exactly the same form is not necessarily rigour either.
The harder standard is to preserve the substance while adapting the path.
Change the sequence. Change the evidence. Change the forum. Make the owner explicit. Narrow the commitment. Run the test. Escalate the unresolved risk.
Do not quietly change a material fact, risk or conclusion simply because doing so creates movement.
Judgement is built through decision exposure
Execution builds technical depth and reliability. Judgement develops when people are exposed to the decisions around the work.
Why was an attractive option rejected? Which risk was accepted, and why? Why was a technically weaker solution chosen because it was executable? When did more analysis stop being useful?
If people see only the task and not the decision around it, they can become very good at producing answers without learning how organisations use them.
AI makes this more important. As more first-draft analysis and routine execution becomes automatable, organisations will need to preserve exposure to trade-offs, exceptions, consequences and decision logic.
Otherwise we may automate some of the work through which judgement used to develop without replacing the learning mechanism.
A practical diagnostic
Organisations do not benefit from expertise merely because they employ experts.
They benefit when expertise can survive competing incentives and imperfect information, become an owned decision and translate into action.
That conversion requires organisational judgement.
Increasingly, I see it not simply as a personal leadership quality, but as part of an organisation’s decision architecture: the human capability operating where evidence, power, ownership, risk and action meet.