Define both good-enough criteria and over-investment…

When This Fires

• When creating quality standards for your common output types (Annotate each output type with its half-life…)
• When you notice spending 3x longer than needed on low-stakes outputs
• When high-stakes outputs are delayed because effort was consumed by low-stakes perfectionism
• Complements Annotate each output type with its half-life… (half-life annotation) with the specific quality boundaries for each lifespan category

Common Failure Mode

Floor-only standards: "Everything I produce should be high quality." Without a ceiling, "high quality" means "as good as I can possibly make it," which for a perfectionist means infinite effort on every output. The email takes 30 minutes instead of 5, the meeting notes take 45 minutes instead of 10, and the important report gets pushed to tomorrow because today's hours were consumed by over-polished ephemeral work.

The Protocol

(1) For each output type in your taxonomy (Annotate each output type with its half-life…), define two thresholds: Good enough (the minimum standard — what must be true for this output to be acceptable), and Over-investment (the maximum effort/time — the point beyond which additional polish produces no meaningful value improvement). (2) Good-enough criteria should be binary and checkable (Define completion as binary observables (draft exists)…): "Spell-checked, key points present, action items clear." Not "well-written." (3) Over-investment thresholds should be time-based: "Slack message: 2 minutes max. Email: 5 minutes max. Weekly report: 30 minutes max." (4) When you hit the over-investment threshold, stop. Ship it. The output is good enough, and the effort belongs elsewhere. (5) If you consistently exceed the over-investment threshold for a specific output type, either the threshold is too tight (adjust upward) or you're perfectionism-spiraling on that type (practice stopping).