After the storm

ISO 22301, Clause 10: improvement.

Every disruption, near-miss, and exercise leaves debris worth examining. Clause 10 asks the organisation to treat nonconformities as input: correct them, find their causes, and adjust the continuity system so the same water does not come through the same crack twice. This is the stage that keeps the rest of the playbook from ossifying.

Three levels of fixing

Correction addresses the surface condition: the failover is repaired, the contact list updated, the missing spare ordered. Corrective action addresses the cause: why the contact list was stale, why the spare was never stocked.

There is a third level. When a gap found in one exercise reappears in the next after a corrective action was applied, the corrective action addressed the surface condition but left an assumption intact. The question worth asking is not why the corrective action failed but what the organisation believed about this control’s operating conditions that made the recurrence seem impossible. That belief, once named, can be checked against current reality and corrected at the design level rather than the procedural level.

Feeding the lessons back

Findings flow uphill through the playbook, not just into a tracking sheet:

  • A recurring single-person dependency revises the storm charts, not just the rota.

  • A failover that missed its recovery objective revises the emergency systems, or the objective.

  • A procedure the team quietly worked around revises the drill book; the workaround is information about where the procedure’s model stopped fitting.

  • A scenario that surprised everyone earns a place in the next drill schedule.

Near-misses deserve the same treatment as incidents. A disruption avoided by luck carries the same lesson as one that landed, at a considerable discount.

Executive gap-spotting

  • Closure honesty: Are corrective actions verified against a repeated test, or marked complete when the document is updated?

  • Recurrence tracking: Would anyone notice if this year’s finding matched one from two cycles ago?

  • Upstream flow: Do lessons reach the impact analysis and strategy choices, or stop at procedure edits?

  • Review cadence: Does management review act on trends, or receive them?

A continuity system that has not changed since the last audit is not stable. It is unread.

Output

By the end of this stage, the organisation has a corrective action log with verification status, updated plans and registers reflecting what incidents and exercises taught, and a visible trail from finding to change. That trail is itself audit evidence: it shows a system that is lived in, which is what the dossier is trying to prove.

Last updated: 4 July 2026