Feature Toggle

What it means

Feature toggle is the term popularized by Martin Fowler and the continuous delivery community for what most people now call a feature flag. A toggle wraps code so that its execution depends on a boolean or a targeting decision resolved at runtime. Fowler’s taxonomy splits toggles into categories by intent - release toggles that hide unfinished work, experiment toggles for A/B tests, ops toggles for operational control like circuit breakers, and permission toggles that gate features by plan or role. The categories differ mostly in how long the toggle should live and how dynamic it needs to be.

Why it matters

The distinction between toggle types matters because it tells you how to treat each one. A release toggle is temporary and should be deleted the week the feature ships, while a permission toggle for a premium tier may live for the life of the product. Mixing these up is how codebases accumulate hundreds of stale switches nobody dares remove. Thinking in toggle categories keeps flag debt in check and clarifies which flags need percentage targeting versus a simple on or off.

Why the two names

There is no functional difference between a feature toggle and a feature flag - vendors and engineers use whichever term they learned first. Our comparison of feature flags vs feature toggles walks through the history. Tools such as Unleash and ConfigCat implement the full toggle taxonomy with typed flags, so you can model release, ops, and permission toggles distinctly rather than treating everything as a plain boolean.