A key risk indicator (KRI) is a metric that signals a change in the level of a risk, ideally before the risk turns into an incident. Many organisations report "KRIs" that are really outcome counts: number of incidents, number of breaches. These tell you that something has already gone wrong. Good KRIs give warning.
KPIs, KRIs and KCIs
| Type | Measures | Example |
|---|---|---|
| KPI (performance) | How well a process meets its objectives | Percentage of tickets resolved within SLA |
| KRI (risk) | How exposure to a risk is changing | Number of internet-facing systems with critical vulnerabilities older than 14 days |
| KCI (control) | Whether a control is working | Percentage of privileged accounts reviewed this quarter |
The same number can be more than one type. What matters is that a KRI is linked to a specific risk in the register and has thresholds that trigger action.
Leading versus lagging
- Lagging indicators count outcomes: incidents, losses, audit findings.
- Leading indicators track conditions that make outcomes more likely: unpatched systems, excessive privileges, staff turnover in key roles, vendor concentration.
A useful KRI set has mostly leading indicators, with a few lagging ones to check that the leading ones predict correctly.
Examples for cyber and technology risk
| Risk | Leading KRI | Example thresholds (green / amber / red) |
|---|---|---|
| Ransomware via exploited vulnerabilities | Internet-facing critical vulnerabilities older than 14 days | 0 / 1–3 / more than 3 |
| Account takeover | Percentage of workforce accounts without phishing-resistant MFA | under 2% / 2–5% / over 5% |
| Insider misuse | Privileged accounts not reviewed in the last 90 days | 0 / 1–5 / more than 5 |
| Data loss | Sensitive files shared externally via public links | Trend-based: rising for 2 months = amber |
| Recovery failure | Days since the last successful restore test of critical systems | under 90 / 90–180 / over 180 |
| Vendor breach | Critical vendors with overdue security reviews | 0 / 1–2 / more than 2 |
| Key person risk | Critical systems with only one person able to administer them | 0 / 1–2 / more than 2 |
| Privacy non-compliance | Rights requests open beyond the target response time | 0 / 1–5 / more than 5 |
These thresholds are illustrations. Set yours from your risk tolerance.
Setting thresholds
- Start from tolerance. If your tolerance is "critical patches within 14 days", the red threshold follows from it.
- Use history. Look at the past 12 months of data to understand normal variation. A threshold breached every week will be ignored.
- Define the response. Amber means the owner investigates and reports at the next review. Red means escalation to a named executive within a set time.
- Revisit. As controls mature, tighten thresholds.
Qualities of a good KRI
- Measurable automatically from systems, not compiled by hand each month.
- Linked to a risk and an owner.
- Sensitive. It moves when risk changes.
- Comparable over time. Don't change the definition every quarter.
- Understandable to non-specialists.
Reporting so leaders act
- Show trend, not just the current value. A KRI at amber and worsening matters more than one at amber and improving.
- Report exceptions. Boards need the handful of indicators outside tolerance, with the plan to fix them.
- Pair each red KRI with a decision you need: budget, risk acceptance or a deadline.
Learn more
KRIs, risk monitoring and reporting are tested heavily in certification exams such as CRISC. Practise with our free CRISC track, and see how to write a risk register entry.
Written and reviewed by the DigiTrust Asia editorial team. How we research, review and correct articles.
General information, not legal advice. Laws and official guidance change; check the current text before acting.