Growth Marketing Glossary

Feature Flag

fea·ture flagnoun

A switch to turn features on or off without a deploy - the mechanism behind gradual rollouts, instant rollback, and controlled experiments.

featureon/offtoggle withouta deployturn a feature on or off without redeploying
Schematic — Feature Flag
Term
Feature flag (feature toggle)
Is
An on/off switch for a feature
Without
Deploying new code
Enables
Gradual rollout, rollback, A/B tests

Forms & parts of speech

feature flag · noun
A code switch for a feature.
"Behind a feature flag, we rolled the change out to 5% of users, watched the metrics, then ramped it up."

Definition in plain terms

A feature flag, also called a feature toggle, is a mechanism in software that lets a team turn a particular feature on or off without changing or redeploying the code.

The feature is built and shipped but wrapped in a conditional switch; whether it's active - and for whom - is controlled by a configuration that can be changed instantly. This unlocks several powerful capabilities.

Gradual rollout: release a feature to a small percentage of users first, monitor it, then ramp up. Instant rollback: if something goes wrong, flip the flag off immediately rather than scrambling to deploy a fix. Targeting: show a feature only to specific users or segments.

And A/B testing: show different variants to different groups to measure impact. Feature flags decouple deploying code from releasing a feature, which makes releases far safer and more controllable, and they're a core tool of modern experimentation and continuous delivery.

Why it matters to growth leaders

Feature flags are directly relevant to a growth leader because they're a core enabler of experimentation and safe, controlled rollouts - both central to growth.

The ability to release a feature to a small slice of users, measure the effect, and either ramp it up or instantly turn it off is exactly the controlled, low-risk approach growth experimentation requires.

Feature flags are also the mechanism behind much A/B testing in product, letting different users see different variants so the team can measure which performs better.

For a growth leader, this means features and experiments can be rolled out gradually and reversibly, dramatically reducing the risk of any single change and enabling continuous, data-driven iteration.

Understanding feature flags clarifies how a growth and product team can move fast while staying safe - shipping behind flags, testing on subsets, and learning from real usage before committing fully.

They're a foundational tool for combining velocity with control, which is the heart of disciplined growth experimentation.

Worked example. A growth leader wants to roll out a significant new feature but is wary of the risk of releasing it to everyone at once, and feature flags provide exactly the controlled, low-risk path needed.

The team ships the feature wrapped in a feature flag - a switch that controls whether it's active and for whom, independent of deploying the code.

Rather than a risky big-bang launch, the leader releases the feature behind the flag to just 5% of users first, watches the key metrics closely for problems or negative effects, and finds it performing well.

The team then ramps the flag up gradually - to 25%, 50%, 100% - monitoring at each step, with the safety net that if anything goes wrong they can flip the flag off instantly and roll back without a frantic deploy.

The same flag mechanism lets them run it as an A/B test, showing the new feature to one group and the old experience to another to measure the real impact.

The growth leader recognizes feature flags as foundational to combining velocity with control: features and experiments roll out gradually and reversibly, the risk of any single change drops sharply, and the team iterates continuously on real-usage data.

Using feature flags, the leader ships fast while staying safe - the heart of disciplined growth experimentation.
Failure modes to watch. Releasing features to everyone at once instead of rolling out gradually behind a flag; lacking instant rollback when a change causes problems; not using flags to run controlled A/B tests; and conflating deploying code with releasing a feature when flags decouple them.

Synonyms & antonyms

Synonyms

feature flagfeature togglefeature switch

Antonyms

big-bang releaseall-or-nothing launch

Origin & history

A feature flag toggles a feature on or off without deploying code, decoupling deployment from release; it enables gradual rollouts, instant rollback, targeting, and A/B testing - a core tool for safe, experimental, continuous delivery.

Etymology: source.

Usage trends

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Common questions

What is a feature flag?
A switch in software that lets a team turn a feature on or off — for everyone or specific users — without deploying new code, enabling gradual rollouts, instant rollback, A/B tests, and safer releases.
What can you do with feature flags?
Roll a feature out gradually to a percentage of users, instantly turn it off if something breaks, target specific segments, and run A/B tests by showing different variants to different groups.
Why do feature flags matter for growth?
They enable controlled, low-risk, reversible rollouts and much of product A/B testing — letting a growth team ship fast while staying safe and iterating on real-usage data.

Related tools & calculators

Resources & people to follow

Curated, non-competitor resources verified per term.

Related training

Disciplines

Areas of marketing where feature flag is a core concern:

Sources

  1. trendsGoogle Trends — "feature flag"