Incident Blast Radius Mapping
What does this incident affect? Map it explicitly.
Affected services
Mapping affected services is the first impact dimension. The direct failures plus the dependency tree below them. Indirect impact is where most surprises hide during cleanup.
- Direct impact. The services where the symptoms are observable. Failed-request set per service.
- Indirect impact. Downstream dependents that degrade silently. Often discovered hours after the primary outage resolves.
- Live dependency graph. Per-incident reference to the service graph. Surfaces indirect impact while the IC can still act on it.
- Named owner per service. Each affected service has a named team. Parallel response only works when ownership is unambiguous.
Affected customers
Customer count is the second dimension. Pull from real data, not estimates; segment when customer tiers carry different stakes.
- Count, not estimate. The actual customer-impact count from real data. Guessing produces communications that fall apart in review.
- Per-segment when tiers differ. Enterprise versus free customers carry different stakes. The split matters for customer comms and for SLA credit math.
- Contact list per impacted customer. Named contact for proactive communications. Cold-starting the contact at incident time is too slow.
- Quarterly data-pipeline check. Impact-query freshness review. Stale CMDB or customer database produces wrong numbers.
Revenue impact
Revenue is the third dimension. Honest, tied to actual customer count and per-customer contract value. Inflated numbers fail finance review; under-reported numbers fail customer trust.
- Per-customer contracted value. Computed from billing data, not memory. The number must survive scrutiny.
- Honest reporting. Not inflated, not understated. The same number lands in postmortem and finance review.
- SLA-credit estimate. Per-customer contractual credit owed. Calculated proactively rather than after the customer asks.
- Quarterly trend chart. Revenue-impact trajectory across quarters. Surfaces recurring patterns that drive engineering investment.