When you make a choice, write down what you chose, why, and how sure you were. Later you can look back and learn from it.
A decision record is a short note made when the decision is made, not afterwards. It lists the options, what you knew, what you chose and how confident you were. It's the only fair way to judge a decision later.
Capture the decision at the time it's made, using a fixed schema: context, options, information, choice, confidence, owner, review date. Store records where they can be queried, so you can measure calibration and decision quality across many decisions.
The fields
| Field | What to write |
|---|---|
| Decision | One line, framed as a choice |
| Options | The options you considered, including "do nothing" |
| Goal | Save money, make money, or look after people, and the measure |
| What we knew | The key facts and their sources |
| Choice | The option picked |
| Confidence | A percentage: how likely is this to achieve the goal? |
| Expected result | What you expect to see, by when |
| Owner | Who made the call |
| Model involvement | Whether a model scored it, its suggestion, its score |
| Review date | When you'll look again |
Keep it short
A record that takes twenty minutes won't get written. Most fields take a line. If you're using a model, its suggestion and score can be filled in automatically.
Where to keep records
Keep them somewhere they can be searched and counted. A SharePoint list or a Dataverse table works. Free-text notes scattered across email don't. Once you have a few hundred records, you can measure calibration, spot repeated patterns, and see whether model-assisted decisions do better than unassisted ones.
Review
On the review date, add three fields: what happened, whether the decision was reasonable given what was known, and what you'd do differently. Keep the outcome and the judgement of the decision separate. That's the defence against outcome bias.