The Decision Learning Loop
Log → Revisit → Record → Learn
By the end of this module
Why This Loop Matters
Most founders make decisions and forget them. When the same situation comes up again, they re-derive the same reasoning (or worse, reach a different conclusion with no memory of why they changed their mind).
The decision learning loop closes the gap: make → record → revisit → learn.
Step 1: Log Immediately
Log decisions as close to the moment of decision as possible. Don't wait until you "have time to write it up properly." A rough decision log is 10 times more valuable than a polished one that never gets written.
Minimum viable log:
- Title: what you decided
- Decision Made: the actual choice
- Assumptions: what has to be true for this to work
Add reasoning, alternatives, and risks as time allows.
Step 2: Set a Revisit Date
Every decision worth logging is worth revisiting. Set a date based on when you'll know if the decision was correct:
- Pricing decision: 90 days (enough time to see customer response)
- Technology choice: 6 months (enough time for pain to surface)
- Positioning decision: 3 months (enough time for marketing data)
- Hiring/partnership: 3-6 months
Step 3: Honor the Revisit
When the date arrives, it appears on Today. Don't dismiss it without actually reviewing the decision.
Navigate to the decision. Read the original context. Ask: Was I right? What happened? What do I now know that I didn't then?
Step 4: Record the Verdict
Use the DecisionOutcome panel to record:
- VALIDATED. The decision proved correct
- PROVED_WRONG. The decision was wrong; explain why
- SUPERSEDED. Replaced by a newer decision (link to it)
- STILL_ACTIVE. Too early to call; reschedule revisit
Write the outcome note. This is where the learning lives. Even a single sentence is valuable.
Practice Checkpoints
Sign in to track your progress and mark checkpoints complete.
Next in Foundations Path
The Weekly Operating Rhythm
Three nested cadences, daily, weekly, monthly, that keep the system from rotting.