Decision architecture

Most decision rights were never designed. They were inherited from whoever happened to own the first case, and a growing company inherits the wrong owner as often as the right one.

Consensus is not the problem. Running it as the decision rule instead of the check is. The method is three lines settled before the meeting, not a better meeting.

In fintech the two are named in one breath and handed one seat. One of them is answering a question of law and the other is quoting a price, and a single seat gets both of them wrong.

A bank can staff a payment product decision correctly and still get it wrong, because having an owner is not the same question as having the right one.

Every checkout roadmap item that removes friction to raise conversion is, from another desk, a change to the fraud-loss curve. Nobody assigned who wins.

A payment product touches product, risk, compliance, and technology. The decision belongs, by charter, to none of them.

A meeting that keeps not-quite-closing is not short of information. It is running on the wrong decision rule.

A decision can be final in the room and still unsettled everywhere the room cannot reach.

A strategy can be fully agreed and still decide nothing, because it names several good things without saying which one gives way when they collide.

Discipline does not keep a decision closed. A place to check it, and a rule for when it may return, does.

Teams do not reopen decisions because they lack discipline. They reopen them because nothing about the original decision was ever closed.

Reversible and irreversible decisions carry different risk. Most organisations run them through one process, and pay for it at both ends.

The important work is rarely the decision itself. It is the preparation that makes a decision clear enough to carry.