The Executive Summary
Creators at $60–$150K/year repeating a single $10K mistake annually lose $100K over a decade — the Decision Journal Protocol closes that loop with four fields and one 15-minute quarterly review.
Who this is for: Solo creators and operators at $60–$150K/year making recurring consequential decisions — pricing, hiring, platform bets, launches — with no system capturing the reasoning behind them
The decision learning problem: Memory rewrites reasoning within 30 days of an outcome; one $5K–$15K mistake repeated annually compounds to $50K–$150K over 10 years; post-mortems only capture the story outcome knowledge creates, not the logic that drove the decision
What you’ll learn: The Decision Journal Protocol, the Four-Field Format (Field 1–4), the 90-Day Review Cycle, the Quarterly Review Prompts, and the Pattern Identification Guide
What changes if you apply it: You shift from running on selectively optimistic memory to operating with a documented record of your prediction accuracy — named biases replace invisible patterns
Time to implement: 30 minutes to set up; 10 minutes per entry; one 15-minute quarterly review session; first named bias pattern visible at the 90-day review
Written by Nour Boustani for solo creators and operators at $60–$150K/year who want measurable improvement in decision calibration without adding a daily journaling practice.
› Library Navigation: Quick Navigation · Internet Solos and Creators
Decision Journal Protocol: Closing the Feedback Loop on Costly Mistakes
Repeating expensive business mistakes isn’t a character flaw. It’s a data problem.
Creators in the Scaling band ($60K–$150K/year) who have no system for reviewing past decisions are running their businesses on memory. Memory is selectively optimistic, chronologically distorted, and structurally blind to the patterns it keeps producing.
The Decision Journal Protocol is a four-field logging system with a 90-day review cycle. It installs a feedback loop between decisions and outcomes, transforming each mistake from a sunk cost into a recoverable asset.
A creator who repeats one $5K–$15K mistake annually loses $50K–$150K over 10 years from the same pattern. The journal closes that leak at its source.
Where are you with this right now?
“I keep making the same category of mistake — wrong hires, failed launches, wasted spend, and I can’t figure out why the pattern persists.” You’re inside this constraint. The Decision Journal Protocol below installs the feedback architecture. Start at Field 1 and don’t skip the 90-day review cycle.
“I’m not yet making decisions at the scale where this matters — I’m still trying to get to consistent revenue.” The journal compounds with time and requires a decision history to review. Build a basic operating cadence first. See Quarterly Review Template for Solo Creators: Diagnosing What Actually Broke for the diagnostic foundation, then return here once decisions are recurring enough to pattern-match.
“I’ve tried journaling before and stopped after two weeks.” The Decision Journal Protocol is not a journaling practice. It’s a triggered logging system — you log only when a decision meets a specific threshold ($1K financial impact or 10 hours time impact), and you review only once per quarter for 15 minutes. It’s designed to run without discipline.
Try This Now
Pull your last three business decisions with either a financial outcome above $1,000 or a time commitment above 10 hours.
For each decision, answer this question:
“Did the actual outcome match what I predicted when I made the decision?”
If two out of three outcomes diverged significantly from your prediction, you have a calibration gap. Your internal model of how your business behaves does not match how it actually behaves. That gap is what the Decision Journal Protocol is designed to close.
Every uncaptured decision is a lesson that dissolves the moment the outcome arrives.
Why Business Decisions Become Difficult to Learn From
A creator business in the Scaling band generates a relentless stream of consequential decisions:
Pricing adjustments.
Launch timing.
Offer positioning.
Contractor hires.
Platform bets.
Content pivots.
Each decision carries real financial and time consequences. Most are made in the moment, using whatever reasoning feels most available.
The decision gets made. Time passes. The outcome arrives. Within days, the creator’s memory has already begun revising the story of why they made the choice.
This is not a weakness. It is how human memory works. Outcome knowledge rewrites the memory of reasoning.
A decision that failed gets remembered as “obviously wrong in retrospect.” A decision that succeeded gets remembered as “clearly the right call.”
The creator who made both decisions with identical confidence and reasoning quality comes to believe they were decisive and strategic on the winner, but careless or naive on the loser. Neither story is accurate. Both stories are useless for learning.
What Is Actually Happening
The failure mechanism is identical across creator types at this revenue stage. The surface varies. The engine underneath does not.
A newsletter operator at $85K/year with 4,200 subscribers runs three paid launches per year. Each launch involves a sequence of consequential decisions:
Offer positioning.
Price point.
Launch timing.
Email sequence length.
Early-bird structure.
Urgency mechanism.
After each launch, they review revenue and declare the launch a success or failure.
They attribute the results to the decisions that were most visible, such as the offer, price, or email count. The decisions that actually drove the outcome are never isolated and reviewed:
Timing relative to the audience’s buying cycle.
The specific framing of the urgency mechanism.
The gap between the warm-up content and the launch opening.
The next launch repeats the same invisible mistakes with different surface-level variables.
A high-ticket coach at $95K/year makes a second contractor hire after the first one failed at three months. They hire a different person, in a different role, at a slightly different rate.
The hire fails again at four months for a structurally identical reason. The role was not documented before hiring, and the quality standard could not be communicated.
The coach attributes the second failure to “bad luck with contractors.” The pattern of hiring before documentation is never isolated because the decision was never logged with its reasoning. There is no record of what they predicted when they made it.
A course creator at $70K/year makes a platform bet by moving their primary course to a new hosting platform after a competitor endorses it. The migration takes three weeks of unplanned time. Three months later, they move back.
The total cost is $8,400 in unplanned hours plus $2,100 in delayed launch revenue during the migration window.
When asked why they made the platform move, they cannot accurately reconstruct the reasoning. The decision was influenced by a combination of FOMO, a single peer recommendation, and optimism about migration complexity that they no longer remember holding.
THE DECISION LEARNING GAP
Decision made
-> Reasoning exists (clearly, in your head)
-> Prediction exists (implicitly)
Time passes
Outcome arrives
-> Memory rewrites reasoning
-> Prediction is forgotten
-> Learning is impossible
Pattern repeats
-> Same mistake, new context
-> $5K-$15K per cycle
-> $50K-$150K over 10 yearsAll three examples share the same architecture failure: decisions are made, outcomes are observed, but the loop between them is never closed. The reasoning at the time of the decision is never captured, so it cannot be compared with what actually happened.
Without that comparison, there is no learning. Without learning, the pattern repeats.
The Advice That Made It Worse
The most common prescription for this constraint is: “Do a post-mortem after every launch or major project.”
This fails because post-mortems happen after the outcome is known. Once you know what happened, your memory of why you made the decision has already been revised to match the result.
A failed-launch post-mortem produces explanations for why the launch was always going to fail. A successful-launch post-mortem produces explanations for why it was always going to succeed.
Neither explanation is necessarily accurate. Both feel accurate.
The post-mortem captures the story created by outcome knowledge, not the reasoning that actually drove the decision. It becomes a narrative exercise rather than a learning system.
The reasoning that needs to be captured is the reasoning at the moment of decision, before the outcome is known and while the actual logic is still accessible.
The Real Cost of Repeated Decision Mistakes
The compounding cost of repeated decision mistakes in the Scaling band is straightforward:
One $5K mistake repeated annually: $50K over 10 years.
One $10K mistake repeated annually: $100K over 10 years.
One $15K mistake repeated annually: $150K over 10 years.
Common Scaling-band decision failure categories and their typical cost per incident include:
Wrong hire or premature hire: $5K–$15K per incident, including contractor fees, lost time, and transition costs.
Failed launch from a preventable positioning error: $5K–$12K per incident, including opportunity cost and production time.
Wasted ad spend from repeated targeting failure: $3K–$8K per incident.
Platform over-investment before validation: $4K–$10K per incident, including migration time and delayed revenue.
Premature pivot away from a working offer: $8K–$20K per incident, including the revenue gap during the transition.
Cost calculator preview:
- Your estimated annual repeat-mistake cost: $__
- Multiply by 10: $__
That’s the 10-year cost of running without a feedback system.Who Should Use the Decision Journal Protocol
This constraint is specific to the Scaling band ($60K–$150K/year) for two compounding reasons.
First, decision frequency and stakes both increase at this band. A creator at $15K/year makes fewer consequential decisions per quarter, and the stakes per decision are lower.
A creator at $80K/year makes pricing, hiring, platform, and launch decisions regularly. Each carries enough financial weight that a single repeated mistake costs real money.
Second, the Scaling band is where systematic bias becomes visible if you are looking for it. A creator who has operated for two or more years at this band has enough decision history to identify patterns. They simply do not have the infrastructure to surface them.
The Decision Journal Protocol installs that infrastructure.
Creators below $30K/year who are still finding their first consistent revenue source will extract limited value from this system. The journal requires a decision history to review. Build operating consistency first.
If the Damage Is Already Done
If you have already identified a recurring mistake pattern and are currently paying for it, the timeline determines your recovery options.
Within 30 days of identifying the pattern
The pattern is still fresh. You can reconstruct the reasoning behind recent decisions with reasonable accuracy, even without logged entries.
Spend two hours writing retroactive entries for the last three to five instances of the pattern you can recall:
- Date
- Decision
- What you were thinking
- What you predicted
- What happenedThis retroactive log will not be as accurate as a prospective one, but it gives you a starting dataset.
Cost to reset: $200–$500 in time. The pattern’s future cost can be eliminated or significantly reduced once the bias is named.
30–90 days since identifying the pattern
Memory has degraded further. Retroactive entries are possible for decisions connected to financial records, such as invoices, ad-spend reports, or launch revenue numbers. These records anchor the reconstruction.
Decisions without financial records are harder to reconstruct accurately.
Protocol:
- Use financial records to anchor the last two to three instances of the pattern.
- Accept that the reconstruction is approximate.
- Start prospective logging immediately.Cost: $500–$1,500 in time. The pattern continues at roughly 50% of its previous cost while the prospective journal builds enough history for the first quarterly review.
90+ days since identifying the pattern
The pattern has repeated enough times to become structural. It will continue without a system to interrupt it.
Retroactive reconstruction is no longer worth the effort. The pattern is knowable from outcomes alone, including financial records, launch results, and contractor tenure.
Protocol:
- Skip retroactive reconstruction.
- Start prospective logging immediately.
- Plan the first 90-day review for the entries accumulated by that point.The pattern will begin to resolve during the first review once the reasoning has been captured prospectively.
Cost: The pattern has already generated its full repeated cost. Going forward, logging has a $0 marginal cost. Pattern interruption should become visible within six months.
The pattern costs exactly what it costs until the reasoning is captured in writing at the moment of decision. That capture is the only intervention that works.
The failure mechanism is diagnosed. Install the four-field Decision Journal Protocol to close the feedback loop and capture what actually matters, without adding unnecessary detail.
The Decision Journal Protocol: Four Fields and a 15-Minute Quarterly Review
Decision quality doesn’t improve from experience alone. It improves through structured feedback on the gap between what you predicted and what actually happened.
Annie Duke documents this mechanism in Thinking in Bets and How to Decide: operators who create explicit feedback loops between decisions and outcomes become more calibrated over time.
The mechanism is not journaling as reflection. It is logging the reasoning at the moment of decision, then systematically comparing it with what actually happened. The comparison, not the writing, is where the learning occurs.
The Decision Journal Protocol installs this mechanism in the lightest possible form:
Four fields.
A 90-day review delay.
One quarterly 15-minute session.
No daily practice.
No elaborate system.
The protocol runs only when a decision meets a specific threshold.
Field 1: The Decision, One Sentence
What it captures: The decision itself, stated as a single sentence. Not the context or rationale.
Why one sentence: Forcing the decision into one sentence tests whether it is clear. A decision that cannot be stated in one sentence is usually several decisions bundled together, or a decision that has not been fully committed to.
The one-sentence discipline acts as a clarity gate before the rest of the entry is written.
Format:
I decided to [specific action] on [date].Worked example: A course creator at $72K/year is considering moving their flagship course from a self-hosted platform to a major marketplace. After weighing the options, they decide to proceed.
Field 1 entry:
I decided to migrate my flagship course from my self-hosted platform to [Marketplace] on March 14.Decision threshold for logging: Log a decision only when it has either:
Greater than $1K financial impact in either direction, including cost, revenue, or opportunity cost.
Greater than 10 hours of time impact, based on implementation time rather than ongoing operations.
Below these thresholds, the decision does not carry enough consequence to generate useful pattern data.
Edge case 1: Recurring operational decisions
Do not log recurring decisions that repeat weekly, such as what to post or which email subject line to use.
Log one instance if a recurring decision type produces a recurring mistake pattern. The goal is to examine the pattern, not record every instance.
Edge case 2: Delegated decisions
Log decisions you make about delegating, such as “I decided to delegate X to contractor Y.” Do not log decisions made by the contractor.
Your decision log tracks your decision patterns, not operational outcomes you do not control.
Field 2: The Reasoning, Logic, and Evidence at the Time of Decision
What it captures: Why you made the decision, including the logic, evidence, assumptions, alternatives, and influences involved.
Why this field is the most important: Memory destroys this information fastest. Within 30 days of an outcome, your memory of the reasoning has already begun adapting to fit what happened.
Within 90 days, the original reasoning is largely inaccessible. Writing it at the moment of decision is the only reliable way to preserve it for review.
What to include:
The specific evidence or data supporting the decision.
The assumptions you were making about timeline, cost, response, or market behavior.
The alternative you considered and why you rejected it.
Any person, piece of content, or event that influenced the decision.
Worked example, continuing the course creator example:
The marketplace has 2M active buyers. My self-hosted platform generates zero organic discovery. All traffic comes from my email list. The migration is estimated at 2 weeks.
I’m assuming marketplace discovery revenue will exceed the 30% platform fee within 3 months. I considered staying self-hosted but concluded the discovery opportunity outweighed the fee. A peer creator reported a 40% revenue increase after their migration.What not to include in Field 2: Do not write the outcome you hope to achieve. That belongs in Field 3.
Field 2 should contain only the reasoning that led to the decision, not the reasoning for why the desired outcome would be beneficial.
Quick signal: Write Field 2 for your last major decision from memory, as accurately as possible.
Then check whether your memory includes what you predicted, which belongs in Field 3, mixed with why you made the decision, which belongs in Field 2.
If those two elements are already blurring, that is precisely the problem the four-field format prevents when you log decisions prospectively.
Field 3: The Expected Outcome, With a Timeframe
What it captures: What you predicted would happen as a result of the decision, including a specific timeframe.
Why a timeframe is non-negotiable: A prediction without a timeframe cannot be evaluated.
“The migration will increase revenue” is not a prediction. It is a hope.
“The marketplace discovery revenue will exceed the platform fee within three months” is a prediction because it includes:
A specific outcome: Marketplace discovery revenue exceeds the platform fee.
A specific direction: Exceeds.
A specific timeframe: Three months.
On day 90, the prediction can be evaluated as true or false.
Format:
I predict [specific outcome] by [specific date or timeframe].Worked example:
I predict marketplace discovery will generate at least $800/month in net new revenue after the 30% fee within 90 days of migration completion.
I predict the migration will be complete within 2 weeks.
I predict my existing audience will not significantly churn because of the platform change.Edge case: Multiple predictions
Log all significant predictions, not just one. A decision typically involves several implicit predictions.
Surfacing each prediction creates a richer dataset for the 90-day review and reveals which predictions were wrong more precisely.
Field 4: The Actual Outcome Review, Filled 90 Days Later
What it captures: What actually happened, compared directly with the predictions in Field 3.
Why 90 days: The 90-day delay serves two functions:
It gives the actual outcome enough time to become observable. Most business decisions do not produce their full effect within a week or two.
It creates enough distance to keep the prediction and review cognitively separate.
You wrote the prediction without knowing the outcome. You review the outcome without being able to revise the prediction. That separation makes the comparison honest.
Format:
- Prediction: [what you predicted]
- Actual: [what happened]
- Accurate?: [Yes, No, or Approximately]Worked example, completed 90 days later:
- Prediction: Marketplace discovery generates $800+/month net within 90 days.
- Actual: $140/month net.
- Accurate: No. Off by 83%.
- Prediction: Migration completes within 2 weeks.
- Actual: 3.5 weeks.
- Accurate: No. Underestimated by 75%.
- Prediction: Existing audience does not churn significantly.
- Actual: 4% churn on the email list during the transition.
- Accurate: Approximately. Churn was real but not catastrophic.Pattern this reveals: The creator systematically underestimates migration complexity and overestimates marketplace discovery revenue for their niche.
This is a specific, actionable bias. It is not “I make bad decisions.” It is “I consistently underestimate implementation time and overestimate platform discovery in my vertical.”
Decision Journal Four-Field Format
Field 1: Decision
“I decided to [X] on [date].”
Field 2: Reasoning
Evidence + assumptions + alternatives rejected + influences
Field 3: Prediction
“I predict [specific outcome] by [specific date].”
Field 4: Review, 90 days later
Prediction / Actual / Accurate? → Pattern namedWhat This Framework Is Really Teaching You
The Decision Journal Protocol is not a journal. It is a calibration system.
The goal is not self-knowledge in the reflective sense. The goal is measurable improvement in prediction accuracy over time. That is also measurable improvement in decision quality because decisions are predictions about the future.
A creator who runs this system for six months does not become a better decision-maker simply because they have reflected more. They become better because they have accumulated documented evidence showing where their mental model of the business is systematically wrong, then adjusted that model.
The six most common decision failure patterns at the Scaling band become visible through this system:
Optimism bias about launch revenue.
Anchoring on the first price quoted.
Inaction on pricing because of fear.
Hiring too late, then hiring the wrong person.
Platform over-investment.
Underestimation of implementation time.
Each pattern produces a consistent divergence between Field 3, the prediction, and Field 4, the actual outcome.
Once visible, these patterns are addressable. Until visible, they compound invisibly for years.
What AI-Assisted Decision Journaling Looks Like
The highest-friction point in the Decision Journal Protocol is Field 2. You need to capture your full reasoning at the moment of decision, when you are usually immersed in the decision rather than stepping back to analyze it.
AI can reduce that friction.
Manual Field 2 completion: A creator working through Field 2 manually typically spends 15–20 minutes articulating reasoning that is partly intuitive and partly evidence-based. The output is often incomplete because the creator does not know which parts of the reasoning to surface.
AI-assisted Field 2 completion: Using Claude at claude.ai, a creator can complete Field 2 in 5–8 minutes by describing the decision and having AI ask structured questions that surface reasoning components they might otherwise omit.
Prompt to use:
I’m logging a business decision in my decision journal.
Decision: [Field 1 entry]
Ask me 5 questions that will help me surface all of the reasoning behind this decision, including:
- The evidence I used.
- The assumptions I’m making.
- The alternatives I rejected.
- Any outside influences on the decision.
After I answer, compile Field 2 from my responses. Preserve my meaning and use first-person language.What AI catches that manual logging can miss: AI can surface the assumption layer, including things you are taking for granted but have not explicitly articulated.
For example:
“You said the migration will take two weeks. What is that estimate based on? Have you migrated platforms before?”
That question helps reveal whether the estimate is evidence-based or optimistic. That distinction is often critical to prediction accuracy.
Voice preservation note: Write Field 2 in your own language, not AI’s. Use AI to surface the reasoning components, then write the entry yourself in the first person.
Manual timeline: 15–20 minutes per entry, 1–3 entries per week.
AI-assisted timeline: 5–8 minutes per entry.
The speed gap matters because Field 2 friction is the main reason creators stop logging. A five-minute entry is sustainable. A 20-minute entry requires discipline.
The protocol is designed to run without relying on discipline. AI removes the friction that makes discipline necessary.
The post-mortem captures the story created by outcome knowledge. The decision journal captures the reasoning that actually drove the decision. Only one produces learning.
The same $8K–$12K mistake pattern can continue for three to five years when the operator never builds the infrastructure to surface it. The mistake feels different each time because the surface variables change:
A different contractor.
A different platform.
A different launch format.
The underlying reasoning pattern does not change.
A six-month decision journal makes that pattern impossible to ignore.
Log the reasoning before the outcome. Review the gap 90 days later. The pattern names itself.
Premium Toolkit available for members
The Decision Journal System includes:
Decision Journal Template — four-field format with space for 20 decisions per quarter and completed example showing 3 logged decisions with 90-day reviews
Quarterly Review Prompts — 5 structured questions surfacing patterns across a quarter’s decisions in 15 minutes
Pattern Identification Guide — 6 most common creator decision failure patterns with specific Field 3-to-Field 4 divergence signatures
Calibration Scorecard — scoring instrument rating prediction accuracy per decision, tracking calibration trend quarter-over-quarter
Plug-and-play AI diagnosis sessions — drop into Claude, Gemini or ChatGPT, answer a few questions, save hours of guessing, get your exact next move
Audio key points — concentrated frameworks you can absorb in minutes, implement while you move
Unlock 750+ ready-to-use constraint toolkits — built to solve every business problem operators actually face.
A creator repeating a single $10K mistake annually for 10 years loses $100K from the same pattern; the Decision Journal System closes that loop in the first quarter.
Cancel anytime. Every download you’ve accessed stays with you.
This toolkit is for creators who are making recurring consequential decisions — pricing, hiring, launching, platform bets, and can see a pattern in their outcomes but can’t isolate the reasoning that drives it.
If you’re not yet at consistent Scaling-band revenue with recurring decision cycles, start with Quarterly Review Template for Solo Consultants: Diagnosing What Actually Broke first.
The Decision Journal System gives you the feedback loop that turns every mistake into data rather than just cost.
One thing from this section:
Decision quality improves only when the reasoning at the moment of decision is compared to the actual outcome 90 days later — everything else is narrative, not learning.
The framework is defined. The next section installs it in a specific sequence, with time benchmarks and named outputs at every step.
Installing the Decision Journal Protocol in 30 Days
A decision journal that is not running produces no data. The implementation below gets your first entries logged and your first quarterly review scheduled before the month is over.
Each step includes a named output, time estimate, tool, and failure mode. If you take longer than the estimate, the failure mode tells you what to adjust.
Step 1: Set Up the Template, Day 1, 30 Minutes
Action: Create your decision journal using the Decision Journal Template from the toolkit. Set up the four-field format and identify your first two or three entries from recent decisions.
How to execute:
Open the Decision Journal Template PDF.
Copy the four-field structure into the medium you will actually use: a dedicated notes app document, a simple text file, or paper.
Keep the journal in the first place you reach after making a decision, not in a folder you open once a month.
Identify the last three decisions that met either threshold: $1K financial impact or 10 hours of time impact.
Write retroactive entries for those decisions.
Label them “Retroactive” so the 90-day review accounts for the reconstruction caveat.
Field 2 and Field 3 will be reconstructed from memory and will be less accurate than prospective entries.
Tool: Any text editor, notes app, or printed PDF. No software is required.
Cost: Free.
Time: 30 minutes.
Output: The decision journal is created, three retroactive entries are logged, and the format is familiar.
What correct output looks like: You can open the journal and complete a new four-field entry in under 10 minutes without referring to the template.
If it takes longer than 30 minutes: You are debating the medium instead of using the first available one. The journal can be moved later. Use what is immediately available now.
Step 2: Log Your First Prospective Entry, Week 1, 10 Minutes
Action: The next time you make a decision that meets the threshold, log it in real time, before the outcome is known.
How to execute:
When a qualifying decision is made, open the journal immediately.
Complete Field 1, the decision, in one sentence.
Complete Field 2, the reasoning. Use the AI prompt from What AI-Assisted Decision Journaling Looks Like if Field 2 is difficult to articulate.
Complete Field 3, the prediction, with a specific timeframe.
Set a calendar reminder for 90 days from today labeled “[Decision] - review Field 4.”
Do not wait until you have more decisions to log. Log the first qualifying decision after setup, even if it feels small. Logging the first prospective entry matters more than the significance of the decision itself.
Tool: Your decision journal from Step 1 and a calendar app for the 90-day reminder.
Cost: Free.
Time: 10 minutes per entry.
Output: The first prospective entry is logged, and a 90-day review reminder is set.
What correct output looks like: Field 3 contains at least one prediction with a specific date attached. The 90-day reminder is in your calendar.
If it takes longer than 10 minutes: Field 2 is the bottleneck. Use the AI prompt from What AI-Assisted Decision Journaling Looks Like to reduce it to 5–8 minutes.
Do not let Field 2 become a writing exercise. It is a capture exercise.
Step 3: Build the Logging Habit, Weeks 2–4, 10 Minutes per Entry
Action: Log every qualifying decision as it occurs throughout the month. Target three to five entries by the end of Week 4.
How to execute: The logging habit is trigger-based, not schedule-based. You are not sitting down every day to write. You are logging whenever a decision meets the threshold.
Add this sentence to your post-decision routine:
“Does this meet the threshold? If yes, log it before moving on.”
Log decisions that feel uncomfortable, such as an uncertain hire or a price you are not confident about. The discomfort signals unclear reasoning, which is exactly where the journal generates the most value.
Tool: Your decision journal and a calendar app for 90-day reminders.
Cost: Free.
Time: 10 minutes per entry.
Output: Three to five logged entries by the end of Week 4, each with a 90-day review reminder.
What correct output looks like:
Each entry has a specific prediction in Field 3.
Each prediction has a specific date.
Each entry has a 90-day calendar reminder.
The entries feel like data capture, not writing.
If you are logging fewer than two entries per week, one of two problems is likely:
Your threshold is too high. Temporarily adjust it to $500 in financial impact or five hours in time impact to build the habit.
You are making decisions without recognizing them as decisions. Review your week and identify one decision that met the threshold but was not logged.
Step 4: Schedule the Quarterly Review, Day 30, 15 Minutes
Action: Schedule your first quarterly review session and familiarize yourself with the Quarterly Review Prompts from the toolkit.
How to execute:
Open your calendar.
Add a recurring quarterly event on the first business day of each quarter.
Label the event “Decision Journal Review - 15 minutes.”
Read the five Quarterly Review Prompts before the first review session.
This event is non-negotiable. It is when the Field 4 reviews happen and patterns are identified.
The prompts are designed to surface patterns in your data, not in your general thinking. They ask questions about the entries, not your beliefs about the business.
Tool: Calendar app and Quarterly Review Prompts PDF.
Cost: Free.
Time: 15 minutes to schedule the event and read the prompts.
Output: A recurring quarterly review event is on your calendar, and the Quarterly Review Prompts have been read and understood.
What correct output looks like: January 1, or the nearest business day, April 1, July 1, and October 1 have “Decision Journal Review - 15 minutes” on your calendar as a recurring annual event.
This Framework Across Three Creator Situations
The Decision Journal Protocol applies the same four-field mechanism across creator types. The decisions being logged differ. The patterns it surfaces differ. The calibration it produces is the same.
Newsletter operator at $85K/year, 4,200 subscribers, three launches per year
Decisions most commonly logged:
Launch timing.
Offer positioning.
Price point.
Email sequence length.
Urgency mechanism.
Post-launch offer extension.
Pattern most commonly revealed by the 90-day review: Optimism bias about launch revenue. Field 3 predictions consistently run 30%–50% above Field 4 actuals.
Once quantified, this bias helps the operator adjust launch revenue forecasts, cash flow planning, and launch investment decisions.
Calibration outcome: Launch revenue predictions reach within 15% of actuals by quarter 3 of journaling, compared with the initial 30%–50% overestimate.
High-ticket coach at $95K/year, 18 active clients, 1:1 and group formats
Decisions most commonly logged:
Client acceptance decisions.
Offer structure changes.
Rate increases.
Contractor hires for delivery support.
Pattern most commonly revealed by the 90-day review: Anchoring on the first price quoted in a client conversation.
When the first price discussed during a prospect call is below the coach’s actual rate, the coach consistently closes at or below that anchor instead of charging the standard rate. This happens even when the stated rate is higher.
Field 3 predictions in rate conversations are consistently more optimistic than Field 4 actuals.
Calibration outcome: Rate anchor discipline. The coach learns never to name a number below their floor in an early conversation because the journal has made the cost of that anchor visible and specific.
Course creator at $70K/year, 1,800 subscribers, two flagship courses
Decisions most commonly logged:
Platform decisions.
Marketing channel investments.
New offer development decisions.
Contractor hires for course production.
Pattern most commonly revealed by the 90-day review: Platform over-investment before validation.
Each platform bet predicts a traffic or revenue outcome that the platform fails to deliver within the predicted timeline. The bias is consistent. The platform’s potential is evaluated against its best-case scenario rather than the creator’s specific niche performance.
Calibration outcome: Platform validation gate. The creator installs a personal rule that no platform bet exceeds $2,000 in cost before achieving one validation metric:
50 organic leads.
$1,000 in platform-attributed revenue.
The rule comes directly from the journal, not from general advice.
Checkpoint
The Decision Journal Protocol is installed when:
At least three prospective entries are logged with Field 3 predictions and 90-day review reminders.
A recurring quarterly review is on the calendar.
The Quarterly Review Prompts have been read.
If any of these three conditions do not exist, the system is not installed. It is only intended.
An intended decision journal produces no data and no patterns.
Readiness Check: Decision Journal Installation
Criteria:
Decision journal created and accessible in under 30 seconds.
At least three prospective entries logged with Field 3 predictions included.
A 90-day calendar reminder set for each logged entry.
Recurring quarterly review event on the calendar.
Quarterly Review Prompts read.
Pass: All five criteria are met.
Fail: Any criterion is missing.
If the result is fail, stop. Do not proceed to the quarterly review cycle.
An uninstalled journal produces no pattern data and no calibration. Proceeding without installation means the $5K–$15K annual mistake pattern continues uninterrupted.
Three prospective entries and a scheduled quarterly review are the minimum viable installation. Everything else the system produces depends on these two outputs existing.
The protocol is installed. Calibration Calculator and 90-Day Milestones validates it with your specific numbers, two possible futures 90 days out, and milestones that show whether the system is producing learning.
Validate Your Decision Journal Before Installation
Your Decision Calibration Calculator
Run these numbers to estimate what improved calibration could be worth in your business.
Pre-filled example: Course creator at $72K/year, making three major decisions per quarter with an average impact of $8,000 per decision.
- Major decisions per quarter: 3
- Average financial impact per decision: $8,000
- Estimated current prediction accuracy: 50% (half of predictions significantly diverge from actuals)
- Decisions where a more accurate prediction would have changed the decision: 1 in 3 (estimated)
- Revenue or cost impact of those changed decisions: $8,000 x 1 = $8,000 per quarter
- Annual value of improved calibration: $8,000 x 4 = $32,000This calculator estimates the value of decisions you would have made differently with better calibration. It does not project a specific improvement percentage. The mechanism is documented, but the magnitude is specific to each operator’s decision patterns.
Your numbers:
- Major decisions per quarter: __
- Average financial impact per decision: $__
- Decisions per quarter where better calibration would likely change the outcome: __
- Average financial impact of those decisions: $__
- Annual value estimate: $__ x 4 quarters = $__Run the Simulation Before You Build
Before installing the Decision Journal Protocol, run this scenario with Claude at claude.ai to identify where your version of the system will face the most friction.
Prompt to run:
I’m installing a decision journal for my creator business.
I make roughly [X] major business decisions per quarter with an average financial impact of $[Y]. The decisions I make most frequently are [list your top 3–4 decision types, such as hiring, pricing, platform, or launching].
Walk me through:
- Which of my decision types is most likely to reveal a systematic bias when reviewed at 90 days.
- What Field 2 looks like for a [specific decision type] in my business.
- What the most common prediction failure looks like for operators at $[your revenue] making these types of decisions.
Show specific examples.What AI catches that you miss: AI can identify which decision types have the most predictable bias patterns based on what you describe about your business.
A course creator who mentions platform decisions may receive a flag about platform over-investment bias. A coach who mentions rate conversations may receive a flag about anchor risk.
Use these flags to pay special attention to those entry types during your first quarter of logging.
Two Futures
Without the Decision Journal Protocol, 90 days later
A creator at $80K/year continues making decisions from intuition without a feedback loop. They hire a contractor for a new role, predicting that the role will be filled successfully and save $3,000 per month in time.
The hire fails at eight weeks for a preventable reason that a documented reasoning review could have exposed. The role was not documented before hiring, repeating the same pattern as the previous failed hire.
Cost:
- Contractor fees: $4,800
- Time cost managing the failed transition: $6,200
- Total: $11,000The pattern is recognized only in retrospect. The next hire faces the same risk.
With the Decision Journal Protocol, 90 days later
The same creator logs the hire decision in real time.
Field 2 captures:
Hiring for a new role. The role is partially documented. I’m assuming the contractor will adapt to the unclear scope.Field 3 predicts:
The role will be clear within 2 weeks. The contractor will perform at the required standard by week 4.At the 90-day review, the creator compares this entry with a prior failed-hire entry from the retroactive log.
The pattern becomes visible:
Both failed hires had “partially documented” in Field 2.
Both included optimistic clarity predictions in Field 3.
Both failed for a structurally similar reason.
The creator installs a personal rule: no hire until the role documentation passes a specific completeness standard.
The next hire succeeds. The pattern is broken.
The difference over four decision cycles per year is $11,000 per cycle avoided:
$11,000 x 4 = $44,000 in annual decision cost recoveredThat recovery can compound as calibration improves.
What Good Looks Like at Each Stage
Day 14
The decision journal is created and accessible in under 30 seconds.
At least two prospective entries are logged with Field 3 predictions and 90-day calendar reminders.
Retroactive entries are labeled as retroactive.
If you are below this threshold at Day 14, the journal exists but logging has not started. The barrier is usually the medium, because the journal is in a friction-heavy location, or the threshold, because you are waiting for a “big enough” decision.
Temporarily adjust the threshold to $500 or five hours, then log the next qualifying decision regardless of size.
Week 4
Three to five entries are logged prospectively.
A quarterly review event is on the calendar.
The Quarterly Review Prompts have been read.
At least one Field 2 entry shows that the AI prompt surfaced reasoning you would not have written manually.
If you are below this threshold at Week 4, Field 2 is taking too long and the logging habit has not formed. Use the AI-assisted prompt for every Field 2 entry until the habit is stable.
The goal at this stage is logged entries, not Field 2 depth.
Week 8
The first 90-day review has occurred if retroactive entries were logged from 90 days earlier, or is scheduled within the next 30 days.
Logging has become trigger-based. You log decisions without consciously deciding to.
At least one Field 4 review has been completed and revealed a prediction-to-actual divergence.
You can identify at least one decision pattern from your entries, even with limited data.
If you have not completed a Field 4 review by Week 8, the 90-day reminders were set but the review was skipped when the reminder fired.
The review is the mechanism. Without it, the entries are data without analysis. Make Week 8 the week you complete the first Field 4 review, even if the entry is retroactive.
If It Does Not Work: Roll Back and Retest
Failure Mode 1: Logging stops after the first two or three entries
Early signal: No new entries for two or more weeks despite qualifying decisions occurring.
Recovery:
Lower the threshold to $500 or five hours.
Move the journal to the first app you open after making any business decision.
Retest for two weeks.
Timeline: Identify the issue within one week of noticing the gap and adjust immediately.
Failure Mode 2: Field 2 entries are too thin to be useful
Early signal: Field 2 entries contain one or two sentences describing the decision rather than the reasoning.
Recovery: Use the AI prompt from What AI-Assisted Decision Journaling Looks Like for every Field 2 entry during the next four weeks. The structured questions surface reasoning components that a self-directed entry can miss.
After four weeks, attempt a manual Field 2 entry and compare its depth with the AI-assisted entries.
Timeline: Identify the issue during the first quarterly review, when Field 4 reviews show that Field 2 entries do not contain enough information to explain the outcome divergence.
Failure Mode 3: Field 3 predictions are too vague to evaluate
Early signal: At the 90-day review, Field 3 entries cannot be scored as accurate or inaccurate because they lack specific outcomes or timeframes.
Recovery: Rewrite the prediction convention. Every Field 3 entry must answer three questions:
What specifically will happen?
By what date?
By what measure?
Add these prompts to the top of every Field 3 section as writing aids.
Timeline: Identify the issue during the first quarterly review and fix the template before logging the next entry.
Failure Mode 4: The quarterly review runs but no pattern is named
Early signal: The quarterly review session is completed and entries are reviewed, but the output is “interesting observations” rather than a named bias and decision rule.
Recovery: The Quarterly Review Prompts are being used as reflection questions rather than pattern-detection tools.
Rerun the review with one constraint. The session does not end until this sentence exists:
I consistently [overestimate / underestimate / avoid] [specific variable] in [specific decision type].That sentence is the output. Without it, the review has not produced a calibration.
Timeline: Identify the issue at the end of each quarterly review session. Pattern-naming may take up to 30 additional minutes when it is resisted. That 30 minutes is the most valuable part of the system.
One-variable adjustment rule: Change one element of the system at a time and run it for four weeks before evaluating.
Multiple simultaneous changes make it impossible to identify what produced the improvement.
The Three Single Points of Failure in This System
The Decision Journal Protocol has three structural points where one failure can collapse the entire learning cycle.
SPOF 1: Field 2 is skipped or reduced under time pressure
When a decision is made during a high-stress or fast-moving moment, Field 2 may be compressed to one or two sentences or skipped entirely.
A thin Field 2 makes Field 4 difficult to interpret. You can see that the prediction was wrong, but you cannot identify which assumption caused the divergence.
Redundancy protocol: Use the AI-assisted Field 2 prompt.
When time pressure hits, the prompt replaces the manual process. It takes 5–8 minutes instead of 15–20, without sacrificing reasoning depth.
Field 2 is never skipped. It is compressed using a tool.
SPOF 2: The 90-day review reminder fires and is dismissed
The review reminder fires, gets swiped away, and is never rescheduled. The entry remains complete through Fields 1–3, but Field 4 is blank.
Without Field 4, the learning loop remains open. The entry is data without analysis.
Redundancy protocol:
Set two reminders for each entry.
Set the first reminder at 85 days.
Set the second reminder at 90 days.
The 85-day reminder primes the review. The 90-day reminder is the execution trigger.
If both reminders are dismissed, the quarterly review session catches the entry with this flag:
Field 4 incomplete. Complete before reviewing patterns.SPOF 3: The quarterly review is deprioritized because no crisis is forcing it
The quarterly review feels optional when the business is stable. There is no acute crisis or obvious pattern causing immediate pain, so the review gets pushed and forgotten.
The data accumulates without analysis.
Redundancy protocol: Treat the quarterly review as a non-negotiable calendar block, not a conditional task. Run it on the first business day of each quarter, regardless of whether a pattern feels urgent.
The 15-minute block is short enough that there is no legitimate time objection. If the review is consistently skipped, the block is in the wrong calendar position. Move it to a time that actually holds.
What This Framework Trains You to See
Early signal 1: Field 3 optimism cluster
What to watch: Multiple Field 3 entries predict revenue or time outcomes that are consistently higher or faster than the Field 4 actuals.
Action: You are running an optimism bias on a specific decision type.
Name it:
I consistently overestimate [launch revenue / contractor productivity / platform discovery] by approximately X%.That named bias becomes a calibration input for every future decision of that type.
Early signal 2: Field 2 assumption cluster
What to watch: Multiple Field 2 entries contain the same unexamined assumption:
“Assuming the platform will deliver organic traffic.”
“Assuming the contractor will adapt to unclear scope.”
“Assuming the list is warm enough for a premium offer.”
Action: That assumption is your active blind spot. Design a validation step before the next decision of that type. Use one piece of evidence to confirm or disconfirm the assumption before committing.
Early signal 3: High divergence in one decision type
What to watch: Prediction-to-actual divergence is consistently larger in one decision category than in others. For example, platform decisions may be consistently less accurate than pricing decisions.
Action: That category has a systematic bias specific to your mental model. Use the Pattern Identification Guide from the toolkit to identify which of the six common patterns is driving the divergence.
The calibration gap between what you predict and what actually happens is measurable, nameable, and correctable once it is captured in writing. When it is not captured, it simply costs money.
The system validates. Calibration Improvement Trajectory covers what the journal teaches you about your prediction accuracy during the first six months and why the first 90-day review usually reveals the most important pattern.
The Calibration Improvement Trajectory
The journal does not improve your decisions by itself. Your first quarterly review does. Everything before that is data collection.
The calibration improvement trajectory follows a consistent pattern for creators in the Scaling band with two or more years of operating history.
Month 1
Logging is new and slightly awkward. Field 2 entries are often too brief because the habit of capturing reasoning has not formed yet.
Field 3 predictions are made, but their specificity varies. When measurable, prediction accuracy is in the range of 40%–50%. Most predictions diverge significantly from actual outcomes, which is normal for an uncalibrated operator.
Month 3: First quarterly review
The first review is where the most important work happens.
With eight to 15 entries logged, the review reveals two or three recurring patterns in the divergence between Field 3 predictions and Field 4 actuals.
For most Scaling-band creators, the first review reliably surfaces:
Optimism bias on launch revenue: Field 3 launch-revenue predictions consistently run significantly above actuals.
Underestimation of implementation time: Field 3 predictions for migrations, hires, and new systems are consistently 50%–100% shorter than the actual timelines.
Overconfidence in new offers: Field 3 predictions for new-offer performance are consistently more optimistic than predictions for established offers.
These are not unusual findings. They are recurring biases in Scaling-band creator businesses and are responsible for many of the $5K–$15K annual repeat mistakes at this stage.
The improvement mechanism is simple: naming the bias changes the behavior.
A creator who knows they have a 40% launch-revenue optimism bias stops making cash-flow and investment decisions based on optimistic launch projections. Applied to three launches per year, that single calibration typically saves $6K–$15K annually in over-committed costs.
Month 6: Second quarterly review
By the second quarterly review, the first-generation biases have been named and partially corrected. A second layer of patterns becomes visible, including more subtle biases in specific subcategories.
Prediction accuracy improves as the operator’s mental model begins to match actual business behavior.
The operator starts making decisions with explicit acknowledgment of known biases:
I’m predicting $12K in launch revenue. My optimism bias typically runs 30%–40% high, so I’m planning cash flow on $8K.The percentage improvement cannot be quantified in advance. It depends on the operator’s prior decision patterns and how directly they act on the patterns the review reveals.
The mechanism is documented and reliable. The magnitude is personal.
The Decision Journal Protocol compounds in two directions:
Forward: Each new entry expands the dataset.
Backward: Past entries become more informative as patterns emerge from later entries.
A journal running for three years contains more than three years of decision data. It also contains three years of pattern data, making each historical entry more interpretable.
This is why the Decision Journal Protocol is most valuable for creators who have operated for two or more years in the Scaling band. Below that point, the decision history may be too thin to produce reliable patterns during the first quarterly review.
After two years at the Scaling band, the compounding effect accelerates. The cost of not having started the journal two years earlier becomes visible in the patterns revealed by the first review.
The first quarterly review reveals the specific decision biases that have been costing money for years. Naming the bias is the mechanism of improvement, and it takes 15 minutes once the data exists.
Running This System in Your Current Condition
Contraction: Revenue Declining or Unstable
The specific risk the Decision Journal Protocol creates during contraction is discomfort with logging decisions when outcomes are bad.
A creator with declining revenue is making decisions under stress. Logging those decisions, including the reasoning behind decisions that failed, can feel like documenting failure.
The failure mode to avoid is stopping the logging process during contraction. This is when the journal’s value is highest.
Contraction is driven by decisions. If you do not log those decisions, you lose the data needed to understand what is driving the decline.
Minimum viable version under contraction:
Log only decisions above the $2K financial impact threshold.
Accept fewer entries in exchange for capturing the most consequential decisions.
Replace the quarterly review with a simplified monthly check.
Ask: “Which predictions from the last 30 days were significantly wrong?”
Ask: “Which assumption drove the miss?”
Signal that the system is making contraction worse: Reviewing Field 4 outcomes creates paralysis rather than calibration.
If you are reviewing losses and feeling stuck instead of identifying adjustments, temporarily step back from the review. The journal is an analytical tool, not a performance review.
If the review produces anxiety rather than pattern recognition, adjust the review format before stopping the logging process.
Stability: Revenue Consistent but Not Growing
The specific blind spot this framework addresses during stability is confirmation bias around decisions that appear to be working.
A creator with stable revenue is making decisions that seem successful because revenue remains stable. The journal may reveal that this stability comes from fewer decisions than the creator assumes.
Several decision categories may be producing flat or negative outcomes that are masked by one or two strong categories.
Use stable periods to run the full quarterly review and identify which decision categories have the best prediction accuracy. Those categories represent your actual areas of business competence, where your mental model is most accurate.
Stable periods provide the clearest signal.
The drift number to watch is the ratio of high-divergence entries to total entries. High-divergence entries are those where the Field 3 prediction was significantly wrong.
A rising ratio during a stable period means decisions that appear stable are producing increasingly unpredictable outcomes. This is an early signal that the underlying model is under stress before the revenue signal shows it.
Expansion: Revenue Growing and Complexity Increasing
The first thing that breaks during expansion is usually logging frequency. Decision volume increases.
A creator growing from $70K to $120K is making more decisions each month. The 10-minute logging habit that was sustainable at three decisions per week may feel overwhelmed at six to eight decisions per week.
The operator may also over-rely on the quarterly review as the only learning mechanism. A creator with 25 or more entries per quarter cannot conduct a meaningful 15-minute review without a longer session or a prioritization filter.
The required guardrail is a threshold filter for the quarterly review:
Review the five highest-impact entries by financial impact.
Review the three highest-divergence entries, where Field 3 was most wrong.
Review eight entries deeply instead of 25 entries superficially.
The capacity signal that triggers an adjustment is a growing logging backlog. If you have qualifying decisions you know you have not logged, temporarily increase selectivity by focusing on decisions above $2K or 15 hours until the backlog clears.
The Decision Journal Protocol in the Creator Operating System
I Keep Making the Same Expensive Mistakes - The Decision Pattern Audit — identifies which decision failure pattern is active before enough journal data exists. Use this before installing the decision journal.
Should I Trust My Gut or Am I Being Stupid - The Signal Authority Tracker — provides structured criteria for distinguishing genuine pattern signals from one-off outcomes. Use this when decision outcomes are difficult to evaluate objectively.
I Keep Making Bad Decisions and I Don’t Know Why - The Decision Diagnosis System — provides prompt library for diagnosing complex multi-variable failures. Use this when Field 4 reveals significant divergence but cause isn’t obvious.
Quarterly Review Template for Solo Creators: Diagnosing What Actually Broke — covers operational and financial review while Decision Journal covers decision-quality review. Use this for 60-minute combined quarterly review sessions.
Solo CEO Weekly Review: How to Stop Drifting and Stay on Strategy — covers updating operating rules based on new evidence from decision journal findings. Use this when translating findings into forward commitments.
Of the four systems listed above, which one is currently blocking your decision-learning cycle?
Decision pattern diagnosis.
Signal evaluation.
Decision diagnosis.
Quarterly review.
If the Decision Journal is running but patterns are not emerging, the bottleneck is usually one of two systems:
Signal evaluation: You are not sure which outcomes represent real patterns rather than noise.
Pattern diagnosis: You can see that something is wrong but cannot identify the category.
Both bottlenecks have specific tools linked above.
Your Decision Learning Starts Now
What you’ll be able to say at Week 8:
“I have at least 3 prospective entries logged with specific predictions and 90-day review reminders set.”
“I’ve completed at least one Field 4 review that revealed a divergence between what I predicted and what happened.”
“I can name one decision bias that’s been costing me money — and I know which future decision type it affects.”
Three time-boxed actions:
Next 30 minutes: Create your decision journal using the template. Log one retroactive entry from a recent qualifying decision. Set a 90-day review reminder.
This week: Log the next qualifying decision in real time. Complete all four fields. Set the 90-day reminder before closing the entry.
Before next month: Add the quarterly review recurring event to your calendar — January, April, July, October. Block 15 minutes. It doesn’t move.
Decision Journal Protocol Progress Milestones:
Milestone 1: Journal created and accessible in under 30 seconds from any device used for business decisions.
Milestone 2: First three prospective entries logged with specific Field 3 predictions and 90-day calendar reminders set for each.
Milestone 3: First Field 4 review completed, with at least one entry containing a 90-day prediction-to-actual comparison.
Milestone 4: First quarterly review completed using the Quarterly Review Prompts, with at least one recurring pattern identified and named from the quarter’s entries.
Milestone 5: Named pattern translated into a personal decision rule, using a specific if/then protocol that prevents the identified bias from producing the same outcome in the next decision of that type.
If you take one thing from each section:
The pattern repeats because the reasoning at the moment of decision is never captured. Memory rewrites it to fit the outcome before comparison can happen.
Decision quality improves only when the reasoning at the time of decision is compared with what actually happened 90 days later. Everything else is narrative.
Three prospective entries and a scheduled quarterly review are the minimum viable installation. Everything else the system produces depends on those two outputs existing.
The calibration gap between what you predict and what actually happens is measurable, nameable, and correctable once it is captured in writing.
The first quarterly review reveals the specific decision biases that have been costing money for years. Naming the bias is the mechanism, and it happens in 15 minutes once the data exists.
But if you remember only one thing:
You don’t keep making the same mistakes because you lack judgment. You keep making them because you have no record of the reasoning that drove them — and without that record, every mistake feels like a new one.
Decision Journal Protocol Checklist
Pull this before your next qualifying decision crosses the $1K or 10-hour threshold.
☐ Log Field 1 as one sentence starting with “I decided to [action] on [date]”
☐ Complete Field 2 with evidence, assumptions, alternatives rejected, and outside influences
☐ Write at least one Field 3 prediction with a specific outcome and a specific date
☐ Set a 90-day calendar reminder labeled with the decision before closing the entry
☐ Add a recurring quarterly review event on the first business day of each quarter
When all five items are checked, your first feedback loop is installed and running.
FAQ: Decision Journal Protocol
Q: What qualifies as a decision worth logging?
A: Any decision with more than $1,000 in financial impact or more than 10 hours in time impact. Recurring operational choices like subject lines or post formats don’t qualify unless a recurring mistake pattern has emerged. The threshold keeps the journal lean and ensures every entry carries enough consequence to generate useful pattern data.
Q: How is this different from a regular journal or post-mortem?
A: A post-mortem happens after the outcome is known, so memory has already revised your reasoning to match what happened. The Decision Journal Protocol captures Field 2 and Field 3 at the moment of decision, before any outcome is observable. That separation is what makes the 90-day comparison honest and the learning structural rather than narrative.
Q: What if I can’t remember my reasoning clearly when I sit down to log?
A: Use the AI-assisted Field 2 prompt from the article — paste your Field 1 entry into Claude or ChatGPT and ask it to ask you five questions that surface your evidence, assumptions, rejected alternatives, and outside influences.
Q: How many entries do I need before the quarterly review is useful?
A: A minimum of three prospective entries with specific Field 3 predictions. Eight to fifteen entries gives the first review enough data to surface two or three recurring patterns. Below three entries, the review will reveal individual outcomes but not the decision patterns that drive repeated mistakes.
Q: What if I miss the 90-day review reminder?
A: Set two reminders per entry — one at 85 days and one at 90. The 85-day reminder primes you; the 90-day reminder is the execution trigger. If both are dismissed, the quarterly review session flags every entry with Field 4 blank and you complete those reviews before analyzing patterns.
Q: Can I run this system retroactively on past decisions?
A: Yes, with caveats. Within 30 days of identifying a pattern, you can reconstruct entries from memory with reasonable accuracy — label them “Retroactive.” Beyond 90 days, use financial records to anchor reconstruction rather than memory. Retroactive entries are less accurate than prospective ones but still useful as a starting dataset for the first quarterly review.
Q: What are the six decision failure patterns this system surfaces at the $60–$150K/year band?
A: Optimism bias on launch revenue, anchoring on the first price quoted in a client conversation, inaction on pricing due to fear, hiring too late then hiring wrong, platform over-investment before validation, and underestimation of implementation time.
Q: What does the quarterly review actually look like in practice?
A: Fifteen minutes on the first business day of each quarter. You read every Field 4 entry completed since the last review, compare each prediction to its actual outcome, and use the five Quarterly Review Prompts to surface patterns across entries.
Q: What happens when Field 2 is too thin to interpret a Field 4 divergence?
A: This is Failure Mode 2 from the article. Use the AI-assisted Field 2 prompt for every entry for the next four weeks. Structured questioning surfaces reasoning components a self-directed entry misses. After four weeks, attempt a manual entry and compare depth.
Q: At what revenue stage does this system produce the most value?
A: The $60–$150K/year band, where decision frequency and financial stakes are both high enough to generate pattern data within a single quarter. Creators below $30K/year have too few recurring consequential decisions for the quarterly review to surface reliable patterns.
⚑ Found a Mistake or Broken Flow?
Spotted a math error, unclear framework, or broken link? Use this form to flag it — helps me keep the articles accurate and useful. Report a problem →
› More to Explore: Quick Navigation · Internet Solos and Creators
➜ Help Another Founder, Earn a Free Month
If the Decision Journal Protocol just showed you where your prediction accuracy is leaking money, share it with one founder stuck in the same loop of repeated expensive mistakes.
When you refer 2 people using your personal link, you’ll automatically get 1 free month of premium as a thank-you.
Get your personal referral link and see your progress here: Referrals
Get The Decision Journal Protocol Toolkit
You’ve read the system. Now implement it.
Premium gives you:
Ready-to-use PDF toolkit—every template, diagnostic, and formula pre-filled, zero setup, immediate use
Plug-and-play AI diagnosis sessions—drop into Claude, Gemini or ChatGPT, answer a few questions, save hours of guessing, get your exact next move
Audio key points—concentrated frameworks you can absorb in minutes, implement while you move
Unrestricted access to the complete library—every system, every update
What this prevents: Repeating a $10K mistake annually for 10 years at $60–$150K/year.
What this costs: $49/month.
Download everything today. Implement this week. Cancel anytime, keep the downloads.
Already upgraded? Scroll down to download the PDF, audio, and your AI session.
