Four-question output review: What results? What patterns? What assumption to question? What one change next cycle?

Why This Is a Rule

Unstructured output reviews produce two failure modes: rumination without insight ("I should have done better") and insight without action ("Interesting patterns, but what do I actually change?"). The four-question sequence prevents both by forcing a progression from observation to pattern recognition to assumption questioning to committed action.

The questions implement Chris Argyris's double-loop learning at the output level. Single-loop learning asks "Did I do it right?" (Question 1). Double-loop learning asks "Am I doing the right things?" (Questions 2-3). The fourth question converts learning into structural change (Exactly one improvement per execution cycle — not zero, not three — so you can attribute changes to effects's one-improvement-per-cycle). The sequence is deliberate: Q1 (observation) establishes facts without evaluation. Q2 (pattern analysis) looks across multiple outputs for recurring success/failure patterns rather than treating each output as isolated. Q3 (assumption questioning) challenges the beliefs driving your production decisions — the double-loop component. Q4 (committed action) converts the review into a single actionable change, preventing the paralysis of trying to fix everything at once.

When This Fires

• During any regular review of your output practice (monthly is typical)
• After completing a significant output or a batch of outputs
• When your output practice feels stuck or stagnant
• Complements Open weekly planning by reviewing plan vs. actuals — identify the single biggest gap without judgment, then make one structural fix (weekly plan-vs-actual review) with the output-specific retrospective protocol

Common Failure Mode

Stopping at Question 1: "I produced 4 articles this month, 2 performed well, 2 didn't." This is observation without learning. The valuable work happens in Q2 (why the pattern?), Q3 (what assumption is wrong?), and Q4 (what changes?). If your review stops at listing outputs and results, it's an inventory, not a learning exercise.

The Protocol

(1) Q1 — Results: List each output produced this cycle with its measured results (Track four output dimensions: reach, resonance, downstream action, personal growth — single metrics get gamed; balanced scorecards don't scorecard dimensions). Pure observation, no evaluation. (2) Q2 — Patterns: Compare successful outputs to unsuccessful ones. What structural differences exist? Topic? Format? Distribution channel? Timing? Production process? (3) Q3 — Assumptions: Based on the patterns, what belief or assumption should I question? "I assumed LinkedIn posts drive engagement, but my detailed guides produced more impact." (4) Q4 — One change: Name exactly one specific thing you will change in the next production cycle. Not three, not "try harder" — one structural change. Write it down and schedule it (Integrate the committed action from your review into the next cycle's plan before ending the session — insights must become operations, not observations). (5) Total review time: 15-30 minutes. The structure prevents both rushing (can't skip questions) and spiraling (four questions provide natural boundaries).