Incident Severity Matrix¶
Template — instantiates Tiered Escalation
Combines impact, urgency, scope, and reversibility to route an incident into a response regime.
The Incident Severity Matrix is a pre-agreed grid that turns an incident's observable dimensions — how many are affected, how urgent, how wide the blast radius, how reversible — into a discrete severity grade, and each grade carries a fixed response regime and time target. Its defining move is that it is a routing artifact, not a ladder and not a switch: it does not hold the tiers, and it does not merely flip one command mode on. It reads several signals at once and lands the incident in a cell, and the cell already answers "who responds at what intensity, and how fast." Because the mapping is written down before the incident, the severity call is a lookup rather than an argument, and every grade routes to a different response level with its own service-level target.
Example¶
A payments platform runs a four-grade matrix. The axes are customer impact (a few merchants vs. all) and urgency (money-movement blocked now vs. cosmetic), cross-cut by reversibility. At 02:00 the settlement API starts returning errors for a single regional processor: impact "partial," urgency "funds delayed but not lost," reversibility "high." The on-call engineer reads the grid — that cell is SEV-3: primary on-call owns it, a fifteen-minute acknowledgement target, no bridge call. An hour later the failure spreads to all processors and money stops moving: the cell shifts to impact "total," urgency "funds blocked," reversibility "degrading" — SEV-1, which the matrix routes to an incident lead, a five-minute page, an executive comms bridge, and a status-page update. Nobody debated severity; the grade moved because the signals moved, and each grade came pre-loaded with a response regime and a clock.
How it works¶
- Fix the axes. Choose a small set of orthogonal signals (impact, urgency, scope, reversibility) that genuinely change how hard a response should be — not a single "how bad" guess.
- Bin into grades. Map cells to a short ordinal scale (SEV-1…SEV-4); coarse enough to be memorable, fine enough to separate an all-hands crisis from a nuisance.
- Attach a regime and a clock per grade. Each grade names the required response level and its acknowledgement/resolution target, so classification and escalation-threshold and SLA are read off in one motion.
It does not carry the incident between hands or define who is on which tier — it decides which regime an incident belongs in and when the current one is no longer enough.
Tuning parameters¶
- Axis count — more axes discriminate finer but slow the read under pressure; too few collapse distinct incidents into one grade.
- Grade granularity — more grades route more precisely but invite boundary disputes; fewer are faster but blunter.
- Threshold placement — where each cell tips into a higher grade; set low and you over-escalate nuisances, set high and serious incidents linger under-resourced.
- SLA tightness per grade — aggressive targets on high grades protect stakes but consume responder capacity; loose targets risk slow response on real harm.
When it helps, and when it misleads¶
Its strength is speed and consistency: it removes argument from the worst possible moment and makes escalation a legible, auditable lookup that similar incidents share. Its failure mode is the coarse-bin trap — a small grid can drop a rare catastrophe and a frequent nuisance into the same cell, so the grade flatters or understates the real risk; this is the well-known limitation of any risk matrix that compresses a continuous hazard into a few boxes.[n1] The classic misuse is severity inflation: teams grade upward to guarantee attention, the top grade loses meaning, and the matrix becomes a signal-jamming device. The guarding discipline is to review graded incidents after the fact against what response they actually needed and recalibrate the cell boundaries — the matrix is a hypothesis about routing, not a verdict.
How it implements the components¶
issue_classification— the axes and cells are the classifier: they interpret the incident into a grade.escalation_criteria— each cell boundary is a criterion; crossing it is what declares the current response level insufficient.service_level_target— each grade carries an acknowledgement and resolution clock, so severity and response speed are set together.
It does not implement tier_structure or lowest_competent_level_rule — the standing ladder is Support Tier Model's; nor handoff_protocol or escalation_path, which On-Call Runbook Escalation owns. Its nearest matrix-twin, Approval Matrix, routes decision classes to authority_boundary, whereas this grid routes incident signals to a response regime and a clock.
Related¶
- Instantiates: Tiered Escalation — supplies the classification-and-threshold artifact that decides when an incident must move up.
- Sibling mechanisms: Support Tier Model · Clinical Referral Pathway · Appeal Process · Management Escalation Review · Emergency Command Levels · On-Call Runbook Escalation · Approval Matrix
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: Incident Severity Matrix operates as a case-specific gate, selection, routing, prioritization, or resource disposition because it combines impact, urgency, scope, and reversibility to route an incident into a response regime
Independent corroboration: The frozen evidence defines Incident Severity Matrix as 'Combines impact, urgency, scope, and reversibility to route an incident into a response regime', so its operative form is Decision, Gate & Allocation.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Disaster Management & Risk Reduction
Origin pattern: Convergent development
Present-day reach: Multi-domain
Rationale: Categorical impact-urgency grading to select a response regime is standard emergency and incident management.
Related originating lineages:
- Computer Science & Software Engineering — Technology operations materially standardized severity levels tied to service response.
- Engineering & Design — Risk matrices independently developed severity-likelihood bins in safety engineering.
Review resolution: Both reviewers independently assign disaster_management as the primary originating domain, so that shared primary is retained. Alternate domains are the union of reviewer-identified formative or independently originating lineages; later application settings alone are excluded. The record preserves independently developed forms rather than treating every alternate as mere application. It has established independent use across several domains, but that does not make it domain-free. The encyclopedia entry generalizes the established mechanism without creating a new composite lineage.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] A risk matrix bins a hazard's likelihood and impact into a small grid of categorical cells. Its recognized limitation is that coarse bins can assign the same cell to risks that differ by orders of magnitude, so the grade can rank incidents in a way the underlying numbers would not — a reason to treat cell boundaries as tunable and to audit grades against outcomes. ↩