Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Every detection rule passes through requirement, development, testing, staged release, production monitoring, periodic review, and retirement. Skipping the last two is how estates accumulate thousands of useless rules.

What Is the Detection Rule Lifecycle?
Rules are not written once and left alone. They are built against an environment that changes, running on platforms that change, aimed at adversaries who change. The lifecycle is the agreed path a rule takes through those changes, and the reason to define it is that undefined lifecycles always end the same way: an estate of 4,000 rules of which nobody can say which 300 matter.
Requirement. A rule starts from a stated need: a technique on the threat model, a finding from a hunt, a lesson from an incident, a regulatory expectation, or a gap exposed by testing. Record which, because it determines who cares if the rule later fails.
Feasibility. Does the telemetry exist, on enough of the estate, with the fields the logic needs, retained long enough? Answering this before development prevents the common outcome of a beautifully written rule that silently covers 40% of endpoints.
Development. Write the logic, ideally in a portable format, with the alert fields an analyst needs, an ATT&CK mapping, and triage guidance written at the same time rather than later.
Testing. Two kinds. Does it fire when the technique is executed, and how much does it fire against historical data? A backtest over 90 days gives you the expected volume before anyone is paged.
Staged release. Deploy in audit mode so it generates records but no alerts. Two weeks is usually enough to see the real false positive profile, which never matches the predicted one.
Production. Promoted, owned, and monitored. From here the rule has a precision figure and a last-fired timestamp, both of which are watched.
Review. On a schedule, typically quarterly for high-value rules and annually for the rest, plus an unscheduled review whenever the environment changes in a way that affects it.
Retirement. Disabled, then removed, with the reason recorded and the coverage map updated so the gap is visible rather than forgotten.
Give every rule an explicit state rather than inferring it. Draft, in testing, audit, production, deprecated, retired. Attach an owner, a creation date, a last-validated date, a last-fired timestamp, the precision over the last 90 days, and the linked technique. Most of that can be generated automatically from the platform and the repository, and once it exists the awkward questions answer themselves: which production rules have never fired, which have not been validated this year, which have an owner who left.
Rule health monitoring addresses most of these. Alert when a rule that normally fires weekly has produced nothing for a month, and alert when a rule’s volume changes by an order of magnitude in either direction. Both conditions are far more informative than they sound.
Teams resist deleting rules because of the fear of removing the one that would have caught something. The result is an estate where real detections are buried under content that has produced nothing for years, which is a worse risk. Make retirement routine: a rule that has fired zero times in 12 months and whose technique is covered elsewhere should be retired by default, with the decision recorded and reversible through version control.
Treat the whole lifecycle as change management. Peer review before promotion, a documented rollback, and a record of who approved what. For organisations assessed under the NCSC Cyber Assessment Framework, this record is directly useful evidence for principle C1, which expects monitoring content to be reviewed and kept relevant rather than deployed once.
At Privia Security, we hand over detection content with its lifecycle metadata attached as part of our managed security services, so a rule inherited from us can be audited by whoever owns it next.
Two weeks covers a full business cycle including a weekend and a month boundary for most organisations, and that is usually enough to see the genuine false positive profile. Extend it for rules targeting activity that is seasonal or tied to a monthly process. Review every result produced during the period rather than only the volume, because a low count can still hide an unworkable alert.
When the technology it targets is gone, when the behaviour has become normal in your environment, when another rule covers the technique more reliably, or when it has produced no true positives over a long period while generating noise. Record the reason and update the coverage map so the resulting gap is explicit. Version control makes the decision reversible if circumstances change.
A named individual, not a team mailbox. The owner is responsible for the rule’s precision, its validation date, and the decision to tune or retire it. When someone leaves, their rules need explicit reassignment, and an ownerless rule should be treated as deprecated until someone takes it. This single piece of metadata prevents most of the drift that affects large rule estates.
Monitor the last-fired timestamp and the firing volume for every production rule. Alert when a rule that normally produces results goes quiet for longer than its usual gap, and when volume shifts by an order of magnitude. Pair that with scheduled validation by executing the technique, since a rule that never fires normally will not be caught by volume monitoring alone.
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.