Anti-patterns: what can go wrong

Phase 5 has six recurring operational anti-patterns. Unlike previous phases, these anti-patterns don't manifest immediately — they accumulate silently over weeks or months until the gap between the market and the Context Document is too large to ignore.

Quick reference table

Anti-pattern	Primary signal	Severity	Typical detection
Launch as arrival	Empty Signal Log 3 months post go-live	High	Quarterly
Cosmetic iteration	Same problems in consecutive cycles	High	Monthly
KPIs as target	KPIs rise but users don't improve	Medium-high	Monthly
Signal without cause hypothesis	Full Signal Log, empty Context Update Record	Medium	Weekly
Update without traceability	Context Document changes without signal references	Medium	Monthly
Frozen Context Document	60+ days without updates during active operation	Critical	Monthly

---

1

Launch as arrival

›

2

Systematic cosmetic iteration

›

3

KPIs as target instead of input

›

4

Signal without cause hypothesis

›

5

Update without traceability

›

6

The frozen Context Document

›

---

Detection pattern: the monthly anti-pattern review

At the close of each monthly synthesis session, the team spends 15 minutes reviewing this table:

Question	Healthy answer
Does the Signal Log have entries from the last 2 weeks?	Yes
Does the Context Update Record have entries from this cycle?	Yes
Are the same problems appearing for the third consecutive time?	No
Do KPIs match what users are saying?	Yes
Does each Context Document update reference a signal?	Yes
Has a return to Phase 1 or 2 been considered in the last quarter?	At least once