Targeting Rule

What it means

A targeting rule is the logic a feature flag evaluates to decide what value to serve a given user. Instead of a flag being simply on or off for everyone, targeting rules match against a context of attributes - user ID, email, subscription plan, country, device, or any custom property you pass to the SDK. Rules are evaluated in order: the first one that matches wins, and a default value serves anyone who matches nothing. This is what lets a single flag serve different variations to different segments at the same time.

Why it matters

Targeting is what makes flags powerful beyond a basic switch. Percentage rollouts, beta programs, plan-based entitlements, regional launches, and A/B test bucketing are all just targeting rules. Good targeting is deterministic - the same user always lands in the same bucket via consistent hashing - so a person’s experience does not flicker between page loads. It also keeps sensitive logic like premium gating out of scattered if-statements and in one auditable place, where product managers can adjust who gets what without a code change.

In practice

You might write a rule that serves a new dashboard to all internal emails, then 10 percent of users on the Pro plan, defaulting everyone else to the old view. LaunchDarkly offers a rich rule builder with segments and prerequisites, while GrowthBook ties targeting to its experimentation and analysis layer. Keep rules readable and audited - our feature flags best practices and implementation guide cover attribute design and avoiding rule sprawl.