Detection coverage

Detection coverage describes which specified adversary behaviors a team can detect in a defined environment, supported by evidence about the required telemetry, detection logic, and validation results.

Coverage of which behavior, on which systems?

A useful coverage claim names the behavior, platform, systems, required logs, and expected detection result. MITRE describes building, testing, and refining behavioral analytics, with platform-specific logic in its ATT&CK analytics.

Mapping a rule to an ATT&CK technique helps organize the work. It does not show that the rule catches every version of that technique or even receives the events it needs. An ATT&CK Navigator view needs test results behind it.

Separate mapped, tested, and detected

A rule may be mapped to a technique but never tested in your environment. A test may run without the required logs arriving. Or the logs may arrive and the rule may miss the behavior. Recording those outcomes separately tells engineers what to fix.

Detection engineering covers building and maintaining the rule. Coverage describes how much of the agreed scope has supporting results.

Explain what a percentage includes

A coverage percentage needs a clear denominator: which behaviors and environments were tested? Keep the rule version, configuration, and test date so someone can repeat the check.

Retest after changing logs or detection logic. Include ordinary activity that might trigger the rule. Catching a behavior, producing useful alerts, avoiding false positives, and responding effectively are related but different goals.

Sources

See it in practice