sethmblack Iterative Excellence Framework
Determine when to iterate further versus when work is complete, using Kubrick's insight that iteration is discovery, not repetition.
When to Use
- Uncertain whether to revise again or ship/publish/deliver
- Feeling tension between perfectionism and paralysis
- Need to justify continued iteration to stakeholders (or yourself)
- User asks "How do I know when it's done?" or "Am I being perfectionist or thorough?"
- Any creative or quality-sensitive work where "good enough" might mean "not yet there"
Inputs
| Input | Required | Description |
|---|---|---|
| work | Yes | The current state of the work in question |
| iterations | No | How many revision passes have occurred |
| concerns | No | What specifically feels unfinished or wrong |
| constraints | No | Time, resources, or external deadlines |
Kubrick's Iteration Philosophy
Kubrick was famous for demanding many takes—sometimes 70, 80, or more for a single shot. But his explanation reveals the method:
The mythology: "Stanley's such a perfectionist, he does a hundred takes on every scene."
The reality: "If I did a hundred takes on every scene, I'd never finish a film. Lee Ermey would spend every spare second with the dialogue coach and always knew his lines—he averaged eight or nine takes, sometimes three. Preparation determines how many takes you need."
The insight: Iteration is not repetition. Each pass reveals something the previous pass concealed. The question is not "how many iterations?" but "what am I learning with each one?"
The Framework
Step 1: Identify What You're Looking For
Kubrick didn't do take 50 hoping for magic. He knew what he wanted and recognized when he hadn't found it.
Before iterating again, articulate:
- What specifically is not yet right?
- What would "right" look like if you achieved it?
- How would you recognize success if you saw it?
If you cannot articulate what's wrong: You may be done, or you may need to step away and return with fresh eyes.
Step 2: Assess Whether Iteration Will Find It
Not all problems are solved by iteration:
| Problem Type | Solution |
|---|---|
| Execution not matching vision | More iterations |
| Vision unclear | Stop and clarify, then iterate |
| Fundamental structural issue | Restart at structure level, not iterate surface |
| Anxiety about quality | This is not a problem; anxiety indicates caring |
| Genuine discovery still occurring | Keep iterating |
| Same result each pass | Stop—you've found the ceiling |
The test: Is each iteration teaching you something new? If yes, continue. If you're getting the same result repeatedly, the answer isn't in iteration.
Step 3: Apply the Diminishing Returns Check
Plot your recent iterations:
| Iteration | What Changed | Magnitude of Improvement |
|---|---|---|
| N-2 | [What] | [Large / Medium / Small / None] |
| N-1 | [What] | [Large / Medium / Small / None] |
| N | [What] | [Large / Medium / Small / None] |
Pattern interpretation:
- Large → Medium → Small: Healthy convergence. One or two more passes may find completion.
- Large → Medium → Large: You discovered something new. Explore it.
- Small → Small → Small: Diminishing returns. Consider stopping.
- None → None → None: You're done or you're stuck. Determine which.
Step 4: Apply the Stakes Assessment
The appropriate level of iteration depends on what's at stake:
| Stakes | Iteration Approach |
|---|---|
| One-time, permanent work (book, film, building) | Iterate until excellent |
| High-visibility, long-lasting (major launch, keynote) | Iterate through diminishing returns |
| Regular production work | Iterate until good, not perfect |
| Throwaway/test work | Minimal iteration |
Kubrick principle: Barry Lyndon exists forever. A single film justifies years of work. Your work's expected lifespan should inform your iteration budget.
Step 5: The Completion Tests
Test 1: The Sleep Test If you stop now, will you regret it tomorrow? Next week? In a year?
Test 2: The Specific Objection Test Can you name something specific that's wrong? If you can only say "it doesn't feel done" without specifics, that feeling may be anxiety, not insight.
Test 3: The Discovery Test Did your last iteration reveal anything unexpected? If yes, you're still learning. If no, you may have found the ceiling.
Test 4: The Preparation Test Was this work sufficiently prepared? If you rushed the research or planning phase, iteration will not compensate. The problem is foundation, not finish.
Test 5: The Kubrick Test If you knew this would exist forever, unchanged—would you release it now?
Output Format
## Iteration Assessment
### Current State
[Brief description of the work and iteration count]
### What's Not Yet Right
[Specific articulation of remaining issues, or "Cannot articulate—may be complete"]
### Iteration Pattern
| Recent Iteration | Change | Improvement Magnitude |
|------------------|--------|----------------------|
| [N-2] | [What] | [Level] |
| [N-1] | [What] | [Level] |
| [N] | [What] | [Level] |
**Pattern:** [Converging / Discovering / Diminishing / Stuck]
### Stakes Assessment
**Work type:** [One-time/permanent, High-visibility, Regular, Throwaway]
**Appropriate iteration level:** [Exhaustive, Thorough, Adequate, Minimal]
### Completion Test Results
| Test | Result | Implication |
|------|--------|-------------|
| Sleep Test | [Pass/Fail] | [What this means] |
| Specific Objection Test | [Pass/Fail] | [What this means] |
| Discovery Test | [Pass/Fail] | [What this means] |
| Preparation Test | [Pass/Fail] | [What this means] |
| Kubrick Test | [Pass/Fail] | [What this means] |
### Recommendation
**[ITERATE FURTHER / STOP—YOU'RE DONE / STOP—WRONG PROBLEM]**
[Explanation and specific guidance]
### If Iterating Further
[Specific focus for next iteration]
### If Stopping
[What "done" means in this context]
Outputs
Primary Output: A structured analysis document that identifies and articulates patterns, insights, and actionable recommendations based on the input data.
Format:
## Analysis: [Topic]
### Key Findings
- [Finding 1]
- [Finding 2]
- [Finding 3]
### Recommendations
1. [Action 1]
2. [Action 2]
3. [Action 3]
Example output: See the Example section below for a complete demonstration.
Error Handling
| Situation | Response |
|---|---|
| Insufficient input data | Request specific additional information needed for analysis |
| Ambiguous requirements | Ask clarifying questions before proceeding |
| Conflicting constraints | Highlight the conflicts and ask for prioritization |
| Out of scope request | Explain the skill's boundaries and suggest alternatives |
| Incomplete analysis | Acknowledge limitations and indicate what additional inputs would help |
Constraints
- Do not use this analysis as the sole basis for critical decisions
- Do not apply this framework to situations outside its intended scope
- Acknowledge that analysis is based on available data, which may be incomplete
- Honor the complexity of real-world situations that resist simple categorization
- Present findings with appropriate confidence levels
- Recognize the limits of the methodology
Example
Input: "I've rewritten this blog post six times. It's about productivity systems. Each time I think I've made it better, but I'm not sure anymore. Am I done?"
Output:
Why this works:
This example demonstrates the key principles of the skill in action. The approach is effective because:
- It follows the systematic workflow outlined above
- It shows concrete application of the framework
- It produces actionable, specific outputs rather than vague generalizations
- The analysis is grounded in observable details
- The recommendations are prioritized and implementable
Alternative applications:
This same approach can be applied to:
- Different contexts within the same domain
- Related but distinct problem types
- Scaled up or down depending on scope
- Combined with complementary analytical frameworks
Iteration Assessment
Current State
Blog post on productivity systems, six revision passes completed. Author uncertain whether further revision will improve it.
What's Not Yet Right
Cannot articulate specific issues. Author says "I think I've made it better" but is uncertain. This suggests either completion or burnout, not an identified problem.
Iteration Pattern
| Recent Iteration | Change | Improvement Magnitude |
|---|---|---|
| 4 | [Ask user] | [Ask user] |
| 5 | [Ask user] | [Ask user] |
| 6 | [Ask user] | [Ask user] |
Assumed pattern based on description: Likely diminishing returns—"I think I've made it better" suggests small improvements, not breakthroughs.
Stakes Assessment
Work type: Regular production work (blog post) Appropriate iteration level: Adequate—iterate until good, not until perfect
Reality check: This is a blog post. It has a short lifespan. Six iterations is already thorough for this format. This is not Barry Lyndon; it does not need to be timeless.
Completion Test Results
| Test | Result | Implication |
|---|---|---|
| Sleep Test | Likely Pass | If you've revised 6x, you care; unlikely to regret |
| Specific Objection Test | Fail | No specific objection means no clear target for improvement |
| Discovery Test | Unclear | Are iterations still teaching you something? |
| Preparation Test | Unknown | Did you outline properly before drafting? |
| Kubrick Test | N/A | This work is not meant to exist forever |
Recommendation
STOP—YOU'RE DONE
Six iterations on a blog post is thorough. The absence of specific objections after six passes means either (a) you've found the ceiling of this draft, or (b) if something is wrong, it's structural and more surface iterations won't find it.
The harder question: Are you iterating because the work needs it, or because publishing feels vulnerable? Perfectionism about minor work is often fear disguised as standards. The post is ready; the question is whether you are.
If You Disagree
If you feel something is genuinely wrong:
- Step away for 24 hours
- Return and read aloud
- Write down exactly what bothers you
- If you can articulate it, do one focused revision
- If you cannot articulate it after fresh eyes, publish
The Kubrick Wisdom
"My 'excessive' takes are actually efficiency. Every take teaches you something."
Ask yourself: What did iteration six teach you that iteration five did not? If you can answer, you're still learning. If you cannot, you've arrived.
The Kubrick Principle
Iteration is discovery, not repetition. The filmmaker who shoots take 30 is not hoping to get lucky—they're learning what take 29 revealed about what they're actually making.
But there is a ceiling. When iterations stop teaching, you've either found excellence or found the limit of this approach. Continuing past that point is not perfectionism—it is avoidance.
The master knows when take 30 revealed the truth. The master also knows when to call "print."
Integration
This skill is part of the Stanley Kubrick expert persona. Use it when the question is not "is this good?" but "should I keep working on this, or is it time to finish?"