If you have ever noticed that the same type of decision gets made differently depending on who is in the room, the problem is rarely that some people are making it wrong. It is that the logic for making it correctly is in someone's head rather than on paper. That is a documentation problem, and it is fixable.
The cost of undocumented decision logic ¶
When decision logic is undocumented, the organisation depends on specific people to make specific decisions correctly. Those people leave, get promoted, or go on holiday. The decisions that depended on them either stall or get made inconsistently by whoever is covering. Over time, inconsistent decisions create inconsistent outcomes, and the team spends energy managing the variation rather than the work itself.
How to surface the implicit logic ¶
The best way to surface implicit decision logic is to interview the people who make the decisions well and ask them to walk through real past cases. Not hypotheticals. Real decisions, with real outcomes. Ask them what information they looked at, what they weighed, and what would have changed their answer. Do this with three or four people and you will start to see the shared logic underneath the individual styles.
Test the draft framework against edge cases ¶
Once you have a draft framework, test it against the hardest cases: the ones that were genuinely difficult, the ones that were decided differently by different people, the ones where the outcome was bad. If the framework would have produced the wrong answer in a real case, it needs revision. A framework that only works on easy cases is not useful.
Write it in the format the team will actually use ¶
A decision framework that lives in a PDF nobody opens is not a framework. It is a document. The format matters. For some teams, a one-page flowchart is the right format. For others, a short checklist embedded in the tool they already use. The goal is that the framework is consulted at the moment of decision, not retrieved from a folder after the fact.
Revisit it after six months ¶
A decision framework is not permanent. The conditions that shaped it will change: new products, new regulations, new team members. Build in a six-month review. Ask the team which decisions the framework handled well and which it handled badly. Update it. A framework that is never revised is one that is slowly becoming wrong.
The goal is not to remove judgment from decisions. It is to make the shared logic visible so that judgment can be applied consistently. That is a small thing that makes a large difference over time.