Execution playbook
Founder OS
Move uncertainty out of your head and into the world. Produce evidence, protect the system, and restart quickly.
Habits stabilize the system. Tasks serve the week.
Operating doctrine
Founder OS measures progress by evidence, not effort. These commitments are enforcement mechanisms, not aspirations.
Commitments
- Optimize for external signal, not internal clarity.
- Plan at two levels only: quarter and week.
- Do not judge a week by hours, mood, or task count.
- Protect the system before you push output.
- Explicitly end goals, buckets, and experiments.
- Allow ambition without pressure through a parking lot.
- Never confuse movement with progress.
Non-negotiables
- Maximum three weekly objectives.
- Maximum one outreach-heavy objective each week.
- At least one stabilizing habit lands four or more days per week.
- The weekly review happens even during bad weeks.
Default question
Not evidence: planning, reading, thinking, feeling productive.
Disconfirming evidence still counts. If something fails fast, the system is working.
Core vocabulary
Each term below has a strict meaning. Do not let the words drift.
Q1 2026 goals
Three goals. Each ends in a decision. No goal continues without evidence.
Venture exploration
Milestones
- Weeks 1-3: lock ICP and problem framing, then run 10-15 conversations.
- Weeks 4-8: get a narrow product into the hands of 10 or more real users.
- Weeks 9-13: synthesize the signal and write the decision memo.
Anti-goals
- No scope creep to make the thing look impressive.
- No continuing without evidence.
- No redefining success after the results come in.
Valid evidence
- Three to five real discovery conversations completed.
- One synthesized insight document exists.
- An MVP is used by at least one external user.
- A clear keep, reframe, or kill decision is written down.
Not evidence
- Feature planning or refactoring.
- Reading about product-market fit.
- Internal confidence.
Sustainable daily operating rhythm
Milestones
- Weeks 1-2: define the rhythm and minimum viable day.
- Weeks 3-6: stabilize the rhythm with only minor weekly adjustments.
- Weeks 7-13: complete a 30-day streak without burnout.
Anti-goals
- No all-or-nothing thinking.
- No punishing low-output days.
- No endless system redesign.
Valid evidence
- Thirty consecutive days using the system.
- 80 percent or better tier-one habit adherence.
- Weekly reviews completed, even during dips.
- Fewer system-avoidance days over time.
Not evidence
- Feeling motivated.
- Perfect routines or ideal mornings.
High-leverage income asset
Milestones
- Weeks 1-3: define asset type, buyer, and core promise.
- Weeks 4-8: build it, price it, and make it public.
- Weeks 9-13: evaluate demand and decide what continues.
Anti-goals
- No synchronous delivery or disguised hourly work.
- No expanding scope to justify weak demand.
- No continuing if it consumes more than 10 hours per week.
Valid evidence
- A scoped offer exists publicly.
- At least one pitch is sent.
- At least one payment attempt happens.
- A reusable artifact is created.
Not evidence
- Polishing or waiting for inbound.
- Over-customization.
The bucket list
The bucket list is not a productivity tool. It is a psychological load balancer that reduces ambient anxiety and allows non-optimized human experiences during intense execution cycles.
Remove invisible mental tabs. These are releases, not improvements.
- Close one unresolved financial uncertainty.
- End one ambiguous obligation with explicit closure.
- Make one final decision on a deferred option.
Restore nervous system capacity. These are non-instrumental experiences.
- A full day where nothing is optimized or recorded.
- Time in a space without an agenda.
- Sleep and meals without structure.
Rebuild self-trust through one irreversible signal.
- Kill a project decisively without reopening the loop.
- Deliver something small without over-polishing it.
- Hold a boundary that costs short-term optionality.
Preserve curiosity without pressure and keep an anti-burnout buffer.
- Sample something without declaring intent.
- Learn just enough to satisfy interest.
- Create without publishing it.
Bucket misclassification detector
Run this test on every addition. If an item fails twice, delete it instead of reclassifying it.
Day types
Three to five focused hours. One objective only. No context switching. Produce one externally legible artifact.
Iterate, test assumptions, send follow-ups, refine docs, or review feedback. Progress matters more than polish.
Stabilize momentum and make tomorrow easier. Journaling, review, cleanup, and integration are fine. No new big initiative starts here.
Prevent spirals. Preserve identity. Walk, shower, eat, clean one small surface, or write one grounding sentence. No catching up.
Admin, scheduling, notes cleanup, and prep for the next week. No heavy thinking and no new complexity.
Learn, write, read, or explore with no deliverable pressure. Feed curiosity without forcing monetization.
System-level reset. Rest, presence, and non-instrumental time. No planning and no justification required.
Daily execution sheet
This is how a day gets planned and closed. Every field has a rule. Every rule has a purpose.
Examples: "Ship over polish." "Evidence over narrative." "Energy protection over output."
Rule: if you have two focus lines, you have none.
If no evidence is produced, the day did not move forward. If targets do not causally produce the goal, they are noise.
Outcome: _________________________
Why it matters: _________________________
Link to weekly objective or 13-week goal.
Done when: _________________________
Binary completion condition.
Constraint: _________________________
Time, scope, or energy guardrail.
Rules
- One goal only.
- Must be externally verifiable.
- Must produce evidence, not effort.
- If completed, the day is a win.
Litmus test
If no, the goal is too weak or unclear.
2. ___________________________________________
3. ___________________________________________
5. ___________________________________________
If not, your targets are invalid. Rewrite them.
No outcome, no evidence, no causal chain. Easy to complete without progress.
Clear outcome, clear evidence, each target directly causes the goal.
| __:__ - __:__ | Primary outcome work |
| __:__ - __:__ | Secondary / support |
| __:__ - __:__ | Admin / closure |
I believe: ________________________________
If this happens, I will reconsider: ________________________________
What moved forward: ________________________________
Energy change: + / 0 / -
Carryover (max 1): ________________________________
Execution quality matters less than system integrity.
Eisenhower -> Founder OS
The classic matrix assumes stable roles, external deadlines, and clear task ownership. Founder reality needs a different translation.
| Old label | Founder interpretation | Where it routes |
|---|---|---|
| Important and urgent Stabilizers |
Anything that stops bleed or unblocks execution: admin confirmations, health coverage, payment rails. | Bucket -> closure or This week's top 3 |
| Important and not urgent Foundations |
Systems, strategy, identity, and capability. These become goals, habits, or decisions rather than loose tasks. | Goal or Habit |
| Not important and urgent Noise that feels productive |
The main danger zone. It only survives if it clearly creates evidence for a goal this week. | Weekly objective or Delete |
| Not important and not urgent Parking lot |
Someday items and soft curiosity. Healthy by design, but irrelevant to weekly self-evaluation. | Parking lot |
Task -> bucket -> goal -> habit
Bad: "Reach out to 50 engineers."
Good: "Generate 3 qualified discovery conversations without exceeding 5 hours total."
Decides what gets done in the short term: triage, noise removal, and daily prioritization.
Decides who you are becoming: quarterly direction, identity alignment, and energy protection.
System diagnosis
The Q1 failure pattern was structural, not moral. The system worked when energy was high and collapsed everywhere else.
Over-indexed on ideal execution
Miss one or two days, backlog appears, shame appears, and disengagement follows. The system was too fragile.
No degraded mode
There was no clear answer to the question: "I fell off for five days. How do I re-enter in ten minutes?"
Journal debt
Catch-up pages turned the journal from a tool into an obligation. Obligation became avoidance.
New: "I am designing a system that survives 13 weeks of inconsistency."
Inconsistency is not failure. It is the environment your system must handle.
Minimum viable day and system modes
Minimum viable day
A day is a win if all three are completed. Anything extra is a bonus.
System modes
You are always in one mode. If you are unsure, default to recovery first.
Full journaling, deep work, time blocks, and focused artifact output.
MVD plus one or two additional actions. Keep moving without pushing.
MVD only. No journaling required. No planning required. No guilt allowed.
Hard rules and execution log
The three hard rules
If you miss a day, it does not exist. No catch-up and no reopen.
Only operate inside today and this week. The backlog is outside the system.
Execution comes first. Writing is secondary and optional.
Execution log
One line per day. No narrative, no catch-up, no pressure.
"Shipped feature Z"
"No progress, but walked and reset"
Recovery days can simply say: "MVD only (reset day)"
Weekly system
Q2 goals, reframed
Same three goals, but cognitively lighter. Measurement is evidence-only and perfection is not required.
Product validation
Ship something and get real user signal.
Measure: weekly evidence from conversations, usage, or value exchange.
Operating rhythm
Maintain an MVD streak instead of chasing perfection.
Target: 70 percent or better of days with MVD complete.
Income asset
Build one low-maintenance income path.
Constraint: must stay under 10 hours per week and show real demand.
Anti-shame protocol
This is not motivation advice. It is an operating protocol for the moment you fall off.
- Reflect deeply on why you fell off.
- Analyze the failure in place of action.
- Catch up journals or missed days.
- Redesign the system mid-emotion.
- Complete the smallest possible MVD right now.
- Write one execution-log line.
- Continue forward with no backfill.
New identity: "I am someone who restarts quickly."
Missing days is expected. Restarting is the only requirement.
Comparison framework
Comparing your worst internal moments to someone else's mythologized peak identity produces net-negative execution. The comparison model is broken.
The core error
Tier separation model
Do not compare daily
Use famous operators for long-horizon inspiration or system extraction only.
Correct comparison
Your benchmark is MVD consistency, weekly evidence, and re-entry speed.
The only thing that matters
Did I complete MVD today? Did I generate evidence this week?
Comparison interruption protocol
- I am in a system-building phase.
- They are in a different game or timeline.
- Comparison is not a useful signal by itself.
Ask: what specific behavior or system can I observe, and what is the smallest version I can apply today?
- Do MVD now.
- Log one line.
- Continue.
Peer comparison filter
New: "I am building a system that compounds like theirs over time."
Daily sheet add-ons
Two additions take under a minute and keep the system from collapsing.
Did I complete my MVD today?
The smallest valid version still counts.
Did comparison show up today? Mark yes or no, then force it into an action.
Extract: ____________________________
Action today: _______________________
- Primary KPI -> re-entry speed instead of consistency.
- Journaling -> execution log.
- Full planning -> three weekly questions.
- Day types -> explicit system modes.
- Goal metrics -> MVD percentage and restart quality.
- Anti-shame protocol.
- No journal debt rule.
- Tier separation for comparison.
- Comparison interruption protocol.
- Peer comparison filter.
Buckets remove pressure. Goals create evidence. Habits stabilize the system. Tasks serve the week.
Missing days is expected. Restarting is the only requirement.
Get the systems notes when new playbooks ship
Join the newsletter for new operating systems, experiments, and product notes from carrington.cool.
Useful founder notes only.