Governance

BCM KPIs and KRIs: Performance vs Risk Indicators

Distinguish BCM KPIs from KRIs and use leading indicators to detect deteriorating continuity exposure before recovery capability fails.

Why BCM needs both KPIs and KRIs

Business continuity programs often report what has been completed but fail to show whether exposure is increasing. A Key Performance Indicator (KPI) measures whether a process or capability is performing against an agreed objective. A Key Risk Indicator (KRI) signals that the likelihood or impact of continuity failure may be increasing, sometimes before a recovery test actually fails. The distinction matters because management action can be different.

KPI, KRI and control indicator

IndicatorDecision questionExample
KPIAre we meeting an approved capability or process objective?Percentage of critical services demonstrated within approved RTO using current evidence.
KRIIs continuity exposure deteriorating?Critical services dependent on unsupported technology or a single recovery specialist.
Control indicatorIs a required preventive or recovery control operating?Generator load test completed for the required duration and load.

The labels are less important than the decision they support. A measure should not be called a KRI simply because a red threshold exists. It should indicate changing exposure that warrants attention.

Start with the management decision

Define what action management should consider when the indicator changes. If no plausible decision follows, the measure may be reporting noise. For each indicator record the owner, population, formula, source, observation window, threshold, escalation path and required response. Avoid vanity metrics such as raw counts of plans created unless those counts demonstrate meaningful capability.

Examples of useful KPIs

  • Critical services with current exercise evidence demonstrating approved recovery objectives.
  • High-priority corrective actions closed by due date and independently verified.
  • Critical plans reviewed after material business or technology change.
  • Recovery participants completing required role-based competence assessment.
  • Priority suppliers with current continuity assurance evidence accepted by the accountable owner.

Examples of useful KRIs

  • Growth in untested shared dependencies supporting multiple critical services.
  • Critical applications approaching end-of-support without an approved treatment.
  • Recovery actions repeatedly deferred beyond accepted due dates.
  • Supplier recovery commitments weakening or assurance evidence becoming stale.
  • Increasing key-person concentration in incident and recovery roles.
  • Material site, technology or outsourcing change without corresponding BIA and strategy validation.
  • Repeated exercise findings with the same root cause.

Set thresholds around tolerance, not aesthetics

A threshold should represent a meaningful change in performance or exposure. Define the observation period and the response expected at each threshold. A KRI breach may trigger investigation, treatment or risk acceptance even though current recovery capability still passes. A KPI breach may demonstrate that an approved objective has already not been achieved.

Where possible, use multiple threshold bands only when they produce different actions. For example, a workforce-concentration KRI might move amber when one critical role lacks a trained deputy and red when a service cannot sustain the required shift pattern with currently available competent staff.

Worked example: capability green, exposure red

A critical payment service completed a recovery test within its two-hour RTO, so the recovery KPI is green. However, two of the three specialists who performed the recovery are leaving within six weeks, cross-training is overdue and the alternate administrator does not yet have production recovery access. A workforce-concentration KRI becomes red. The program should not wait for the next annual exercise to discover the capability has degraded.

Management can assign accelerated cross-training, provision access, schedule a targeted recovery test and track the risk until evidence shows the dependency has been reduced. This is the value of pairing performance and risk indicators.

Worked example: avoid denominator manipulation

Suppose 30 critical services require annual exercise evidence. Twenty-four have current evidence, 22 passed and two failed. Six are overdue. Reporting 91.7% passed based only on the 24 tested services hides the overdue population. A better KPI definition makes the denominator explicit and separately reports overdue evidence. Depending on governance, the principal readiness KPI might therefore be 22 of 30, or 73.3%, with the six overdue services clearly visible.

Combine indicators into a decision view

Do not average unrelated red and green measures into a single score that masks critical failure. An executive view can show capability achievement, emerging exposure, overdue actions and evidence freshness as separate dimensions. A single severe dependency affecting a highly critical service may require escalation even if the overall portfolio percentage remains high.

Governance and evidence

Each indicator needs a named owner responsible for data quality and interpretation. Define who can accept a threshold breach, who must investigate it and how long an exception may remain open. Retain source evidence and calculation history so management reports can be reproduced. When definitions change, version them and explain breaks in trend.

Review and retire weak indicators

At least periodically, ask whether an indicator predicted a problem, demonstrated capability or changed a decision. Retire measures that remain permanently green because they are too easy, duplicate other reporting or cannot be supported with reliable data. Add new KRIs when material changes create new concentration, supplier, technology, workforce or regulatory exposure.

Acceptance checklist

  • The indicator answers a specific management question.
  • The formula and denominator are unambiguous.
  • The source and refresh period are defined.
  • Missing data cannot silently improve the result.
  • Thresholds map to explicit actions or escalation.
  • An accountable owner can explain and evidence the result.
  • The measure is reviewed when business conditions change.

Use trends with context

A trend is useful only when the underlying population and definition are stable or the change is disclosed. A falling percentage may reflect newly added critical services rather than deteriorating capability; an improving result may come from removing overdue records. Annotate material scope and definition changes so executives can distinguish real improvement from reporting effects.

Related guidance

Use the BCM KPI Framework for detailed metric specification, the Program Health Dashboard for executive portfolio design and the BCM Metrics Dashboard Guide for lineage, drill-down and reporting controls.