Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Security metrics measure whether controls work and whether the team is improving. A good set is small, has denominators, and leads to a decision somebody can actually take.

What Are Security Metrics and KPIs?
“We blocked 4.2 million attacks this month” tells a board absolutely nothing. It does not say whether the organisation is safer than last month, where the remaining risk sits, or what anyone should do about it. Security metrics earn their place only when they answer a question someone is going to act on, and most published metric sets fail that test.
A metric is a measurement. A key performance indicator is one of the small number of measurements the organisation has agreed to steer by. The distinction matters because teams that treat every measurement as a KPI produce dashboards with forty tiles that nobody reads.
Detection and response. Mean time to detect, mean time to triage, and mean time to contain, measured from real incidents rather than exercises. Alongside them, the proportion of incidents detected by your own controls rather than reported by a third party, which is the single most honest indicator of monitoring effectiveness.
Coverage. Percentage of assets with endpoint detection installed and reporting. Percentage of privileged accounts with phishing-resistant authentication. Percentage of expected log sources that sent data in the last 24 hours. Coverage metrics are unglamorous and they find more real exposure than any other category, because the gaps are where incidents start.
Vulnerability management. Not the raw count of open findings, which tracks scanner coverage more than risk, but the age distribution of unremediated critical issues on internet-facing assets and the proportion remediated within the agreed service level.
Detection quality. True positive rate per rule, escalation rate, and the share of alerts closed without investigation. A rule firing 300 times a week with a 1% true positive rate is doing harm, and this is the measure that proves it.
Resilience. Backup restore tests completed successfully, against the number attempted. Recovery time achieved in the last exercise, against the stated objective.
People. Phishing simulation reporting rate, which measures whether staff tell you, rather than click rate alone, which mostly measures how convincing the template was.
Goodhart’s observation applies with unusual force here. Make mean time to close a target, and cases will be closed faster rather than investigated better. Pair every efficiency measure with a quality measure, and review both together.
A board needs five to seven measures, each with a direction of travel, a target, and one sentence of interpretation. The useful framing is risk rather than activity: which of our material risks moved this quarter, why, and what decision is being asked for. A trend line over four quarters communicates more than any absolute figure, because the audience has no intuition for whether 14 days to patch is good.
Say what the measure does not cover. A 97% endpoint coverage figure is reassuring until someone asks what the missing 3% consists of, and the answer turns out to be the manufacturing systems. Reporting that caveat proactively is what earns the figures credibility the next time they are presented. This is covered further in reporting security to the board.
Measurement is an explicit requirement rather than a nicety. ISO/IEC 27001 clause 9.1 obliges organisations to determine what needs to be monitored and measured, the methods used, and who evaluates the results — and auditors increasingly ask to see the evidence rather than the policy. The NCSC’s Cyber Assessment Framework expects organisations in UK essential services to understand and evidence the effectiveness of their security monitoring, which is a measurement question in all but name.
Build the set from what you can measure reliably today, retire anything nobody has acted on in two quarters, and derive new measures from real incidents through your post-incident reports. At Privia Security, we help teams define and benchmark operational measures of this kind through our SOC maturation services.
Track as many as you need operationally, but report far fewer. A SOC may monitor dozens internally; a board pack should hold five to seven. The test for inclusion is whether a change in the number would cause someone to decide something differently. Measures that fail that test belong in an appendix or nowhere at all.
A metric is any measurement you take. A KPI is a metric the organisation has agreed to steer by, with an owner, a target, and consequences when it moves. Most security measurements should stay as metrics. Promoting too many to KPI status dilutes attention and encourages teams to optimise numbers rather than outcomes.
By output rather than effort. Count hunts completed against the plan, detections created or improved as a result, telemetry gaps identified, and incidents found that no alert would have surfaced. Hours spent hunting is an input and says nothing about value. A hunting programme that produces no new detection content over a quarter needs a different conversation.
Carefully. External benchmarks rarely compare like with like, because definitions, scope, and estate complexity differ, and they can be used to argue that poor performance is normal. Your own trend over time is more actionable. Where benchmarks help is in challenging a figure that looks acceptable internally but sits far outside what comparable organisations achieve.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.