Match review frequency to volatility: weekly for fast-changing, quarterly for stable, instant for surprises
Schedule schema reviews at cadences matched to environmental volatility: weekly/biweekly for high-change contexts, monthly/quarterly for stable contexts, plus triggered reviews when surprises occur.
Why This Is a Rule
Schemas drift from reality at the rate of environmental change. In fast-changing contexts (early-stage startups, rapidly evolving markets, new technologies), schemas that were accurate last month may be obsolete this week. In stable contexts (established processes, mature markets, fundamental principles), schemas remain valid for months or years. A single review cadence for all schemas either over-reviews stable ones (wasting time) or under-reviews volatile ones (allowing drift).
The three-tier review system matches frequency to volatility: High-change (weekly/biweekly): technology stack decisions, market assumptions, team dynamics, competitive landscape. Stable (monthly/quarterly): core values, fundamental principles, well-established processes. Triggered (immediate): any schema whose prediction was surprised by reality — surprise is the real-time signal that a schema has drifted.
The triggered tier is the most important because it's event-driven rather than calendar-driven. A schema can be reviewed quarterly for years and never need updating — but a single surprise means it needs immediate review regardless of when it was last checked.
When This Fires
- Setting up a schema review practice
- When some schemas feel over-reviewed while others are stale
- After a surprise reveals that a schema drifted since its last review
- During any planning of recurring epistemic maintenance
Common Failure Mode
Applying a single cadence to all schemas: "I review everything monthly." This under-reviews volatile schemas (a month is too long for startup-stage assumptions) and over-reviews stable ones (monthly reviews of fundamental principles produce no insights). Match cadence to volatility.
The Protocol
(1) Classify your schemas by environmental volatility: which domains change weekly? Monthly? Rarely? (2) Set review cadences: high-volatility → weekly/biweekly, low-volatility → monthly/quarterly. (3) Add triggered reviews: any surprise (Surprise = automatic schema review — do not wait for the next scheduled cycle) triggers immediate schema review regardless of scheduled cadence. (4) Adjust cadences based on experience: if scheduled reviews consistently produce no updates, lengthen the cadence. If surprises keep appearing between scheduled reviews, shorten it.