Multi-Incident Prioritization When Things Cascade

Two incidents at once; which gets attention. The prioritisation rules.

The rule

Two simultaneous incidents force priority decisions under pressure. The rule is published, simple, and applied consistently: higher severity wins; customer impact breaks ties at the same severity. Without the rule, “both look urgent” paralysis costs MTTR on whichever incident gets the cold-start treatment.

• Higher severity wins. SEV1 always over SEV2. Strict ordering removes the tiebreak under pressure.
• Customer impact wins ties. Affected-customer count decides at the same severity. The number that matters most.
• Explicit ranking per incident. Incident commander documents the order. Catches “both look urgent” paralysis early.
• Published doctrine. Team-visible prioritisation rule. Consistent decisions across rotations.

Staff both

The trap is single-threading on the worst incident while the lower-severity one continues hurting customers unattended. Both incidents need response; allocation is explicit rather than drifted.

• Do not single-thread. Lower-severity incident is still real impact. Both get staffed.
• Explicit allocation. Named responder pool per incident. Resources are limited; allocation is decided, not assumed.
• Named IC per incident. Dedicated incident commander per incident. Prevents IC overload.
• Documented split. “You take A, I take B” in the channel. Mutual assumption costs minutes; explicit assignment does not.

Escalate

Multi-incident is itself an escalation trigger. Manager attention is required regardless of individual severity because the operational and reputational cost compounds when incidents stack.

• Multi-incident equals escalation. Auto-escalation rule per occurrence. Two concurrent incidents trigger manager engagement.
• Manager attention required. Named manager per pair. Drives staffing decisions and external comms.
• Executive notification. Leadership ping per pair. Business-side awareness arrives early, not after the recovery.
• Joint postmortem coupling. Per-pair the joint postmortem option. Shared root causes across concurrent incidents surface only when they are reviewed together.