Iteration cycle anatomy
The Market Phase cycle is not a linear process executed once. It's a system of nested loops operating at different cadences. Understanding this anatomy is what allows the team to operate Phase 5 without it consuming all available time.
The three cycle loops
Inner loop: continuous observation
Cadence: Real time → weekly.
| Activity | Frequency | Automation |
|---|---|---|
| KPI dashboards updated | Real time | 100% automatic |
| Threshold crossing alerts | Real time | 100% automatic |
| Weekly KPI and behavior review | Weekly (45-60 min) | Human with automatic data |
| Signal Log entries from automatic layer | Continuous | 100% automatic |
Output: Raw signals captured and preliminarily classified. Alerts requiring immediate attention.
Middle loop: synthesis and update
Cadence: Monthly.
| Activity | Frequency | Automation |
|---|---|---|
| Qualitative feedback synthesis session | Monthly (2-3h) | Human |
| Ticket and support channel analysis | Weekly-monthly | Partial (AI synthesis) |
| Rule and Story File updates | As needed | Human (AI proposals at level 3+) |
| Context Update Record | With each update | Human |
Output: Refined Signal Log with cause hypotheses. Updated Context Document. Up-to-date Context Update Record.
Outer loop: strategic review
Cadence: Quarterly.
| Activity | Frequency | Automation |
|---|---|---|
| Strategic review with stakeholders | Quarterly (half day) | Human |
| Problem Statement validity review | Quarterly | Human |
| External and competitive signal analysis | Monthly-quarterly | Partial |
| Iteration Brief writing | At each cycle close | AI draft (level 3+) |
Output: Next cycle's Iteration Brief. Decision to return to earlier phases if applicable.
The signal-cause-context chain
The chain is the central mechanism that transforms market data into Context Document updates. Each significant signal must traverse the complete chain before producing a change.
┌─────────────────────┐
│ Observed signal │ ← Raw market data
│ (behavior, │
│ KPI, feedback) │
└────────┬────────────┘
│
▼
┌─────────────────────┐
│ Behavior behind │ ← What is the user doing?
│ the signal │
└────────┬────────────┘
│
▼
┌─────────────────────┐
│ Unmet need or │ ← What problem doesn't
│ unresolved problem │ the current solution solve?
└────────┬────────────┘
│
▼
┌─────────────────────┐
│ Affected Context │ ← Which part of the Context
│ Document element │ Document didn't capture
│ │ this reality?
└─────────────────────┘
If the team can't complete the chain to the Context Document element that needs updating, it doesn't have enough understanding of the signal to act on it. It needs more data or deeper analysis.
The update flow
Not all Context Document updates have the same process or urgency.
| Urgency | Change type | Response time | Approval |
|---|---|---|---|
| Immediate | Technical Rule affecting production operation | Same day | Context Engineer |
| Normal | Story File adjustment, new Rule from emerging pattern | 1-3 days, next iteration | Context Engineer + Product Lead |
| Strategic | Solution Brief or Problem Statement revision | Weeks. Formal return to Phase 1 or 2. | Full stakeholders |
Signs of a healthy cycle
- Each turn of the middle loop produces at least one Context Update Record entry
- The Iteration Brief can articulate concrete learnings linked to specific signals
- Phase 1-2 stakeholders participate in the outer loop
- Time spent on manual capture decreases progressively
- The same problems do NOT appear in consecutive cycles without producing changes
- The inner loop operates but the middle and outer loops don't execute
- The Signal Log grows but the Context Update Record doesn't
- Interpretation sessions produce discussions but not updates
- The Iteration Brief is a copy-paste of the previous one with changed dates