A decision journal preserves the reasons you had before you knew what would happen. You record the choice, evidence, expected result and conditions for reconsidering it, then review those notes later. The template below is free to copy and use. You do not need an account, a paid app or a perfect forecasting system to begin.
The template: capture the decision, then review it
THE CHAIR — DECISION JOURNAL BEFORE YOU DECIDE Date and decision deadline: Decision to make, in one sentence: Who owns the decision: What a useful result would look like, and by when: Facts available now, with their sources and dates: Assumptions I am relying on: Realistic alternatives, including waiting: My choice and the strongest reason against it: What I expect to happen, with a time window: The downside I accept and the limit I will protect: What undoing this choice would cost, and to whom: First action and owner: Evidence that would make me reconsider: Review date: WHEN YOU REVIEW What actually happened, and what is still unknown: What new information arrived after the decision: Which assumption held up, and which did not: Was the original choice reasonable given what was known then: What I would change about the process next time: An original practical worksheet from The Chair. https://thechair.academy/blog/decision-journal-template
Copy this blank worksheet into your own notes and complete it there.
Use a document, notebook or notes app you already trust. Keep the original entry intact and add your later review beneath it. The worksheet captures what you knew, why you chose, and what you will check later.
- The decision and deadline: what must be chosen, by whom and by when.
- Evidence and assumptions: what you know now, where it came from and what you are taking on trust.
- Alternatives and choice: what you could do, your recommendation and the strongest objection to it.
- Expectation and downside: what you predict, over what time window and what loss you accept.
- Reversal and action: what undoing the choice would cost, the first step and its owner.
- Review: a date or signal that will bring you back, followed by what happened and what you learned.
A filled example: changing a workshop booking process
Imagine a fictional pottery studio considering online booking for a Saturday workshop. The owner currently answers booking requests manually. She wants to reduce that work while keeping places accessible to regular customers. The details below are invented to demonstrate the worksheet, not results from a real business.
Her first entry says: “For the next two Saturday workshops, offer online booking while retaining the phone option. I expect fewer manual confirmation messages without more booking errors. I will compare the number of messages and errors with our last two similar workshops. I will also ask the person handling calls which customers still need help.”
| Field | Example entry |
|---|---|
| Evidence | Staff say confirmation messages take time. Existing attendance records show capacity. Neither fact tells us how many people will use a new form. |
| Assumption | Most customers can complete the form; keeping phone booking prevents the test from making access harder for others. |
| Strongest objection | Two channels may create duplicate bookings and more reconciliation work. |
| Accepted downside | Time spent configuring the form and checking both booking lists during the trial. |
| Review trigger | Review after two workshops; pause the form sooner if confirmed reservations exceed available places. |
| First action | The owner tests a booking and cancellation with staff before making the form available. |
Make the prediction specific enough to inspect
“The new system will be better” leaves too much room to reinterpret success. Name what should change and when. Here, fewer confirmation messages is useful evidence, but only alongside booking accuracy and the extra work of maintaining two channels. Looking at one favorable measure could hide the cost moved elsewhere.
Avoid giving a precise confidence percentage just to make an entry look rigorous. If you use one, define the event it describes. You can also record a plausible range or the uncertainty that most threatens the plan. The aim is to make your belief understandable later, including where it was weak.
Include the strongest reason against your choice while you still have time to respond to it. In the studio example, duplicate bookings suggest a concrete preparation step: decide which record is authoritative and who checks it. A useful objection changes the plan; it is more than a ceremonial warning at the end.
Review the result separately from the reasoning
Suppose both workshops run smoothly, but staff quietly spend an extra hour reconciling the lists. Attendance alone would make the trial look successful. Compare that extra reconciliation time with the time saved on confirmations before deciding whether the process reduced work.
Now imagine one workshop is canceled by a power failure. The disappointing result says little about whether the booking approach was sensible. Record the interruption, identify which parts of the trial remain informative and decide whether another observation is needed. Do not erase the entry or force every surprise into the same explanation.
Review in two passes. First establish what happened and what new information arrived. Then examine the earlier choice using the information you actually had. You may find a reasonable decision that met an unusual event, a lucky result that concealed a weak assumption, or a correctable gap in preparation.
Keep the habit small enough to use
Start with one meaningful decision, not a record of every small action. A few specific sentences usually tell you more than a long entry full of abstract intentions. Set the review date when you decide; otherwise the note can become an archive you never reopen.
Keep sensitive business or personal details in a place appropriate for them. The blank download contains no private entries. You can complete it in your own document without submitting your reasoning to The Chair. When sharing a lesson, remove details another person has not agreed to make public.
For a low-stakes first exercise, write down a choice in a business simulation before seeing the consequence. The result is a model outcome, so it cannot validate your real-world ability. It can still give you a concrete opportunity to compare a prediction with feedback and name an assumption you would examine next time.
USE THE FRAMEWORK ON A REAL CASE