Feature Management

What is Gradual Rollout?

Q: What is Gradual Rollout?

A release strategy where a feature is incrementally released to larger percentages of users over time.

A gradual rollout (also called a percentage rollout or progressive delivery) is the practice of releasing a feature to a small percentage of users first, then incrementally increasing that percentage as confidence grows.

This approach limits the blast radius of potential bugs or performance issues. If something goes wrong at 5%, only a small fraction of users are affected, and you can quickly roll back. As metrics look healthy, you increase exposure until reaching 100%.

Gradual rollouts are implemented using feature flags with percentage-based targeting. The flag evaluation is typically deterministic per user (using a hash of the user ID), so each user consistently sees the same experience.

Example

You roll out a new payment flow: Day 1 at 5% (internal testing), Day 3 at 20% (monitor error rates), Day 5 at 50% (check conversion metrics), Day 7 at 100% (full release). At Day 3, you catch a currency formatting bug affecting European users and fix it before wider rollout.

Related Terms

Feature Flag

A toggle that enables or disables a feature in your application without deploying new code.

Canary Release

A deployment strategy that releases changes to a small subset of users or servers before full rollout.

Percentage Rollout

A feature flag targeting strategy that enables a feature for a specific percentage of users.

A/B Testing

An experiment that compares two or more variants of a page or feature to determine which performs better.

Ship features with confidence

Use feature flags to control rollouts, run A/B tests, and instantly kill problematic features.

Explore Feature Flags