Framer Adds Custom Distribution Controls for Staged A/B Testing Rollouts.

Framer has rolled out Custom Distribution for A/B tests, giving site owners the ability to test changes on a limited audience before committing to a wider rollout. The feature addresses a common pain point in experimentation: launching a test at full scale before understanding how it will perform.

With Custom Distribution, users can set the initial percentage of traffic that sees each variant. A landing page redesign, for example, might start at 10% of visitors to the new version and 90% to the control. If early data shows the new design converts better, the distribution can be adjusted to send more traffic to the winning variant without ending the test prematurely.

The update builds on Framer's existing A/B testing toolkit, which already supports variant creation, goal tracking, and statistical significance calculations. Custom Distribution adds another layer of control, particularly useful for high-traffic sites where a poorly performing variant can quickly impact conversion rates or revenue.

When setting up an A/B test in Framer, users now see a distribution slider for each variant. The default split remains 50/50, but this can be adjusted to any ratio. Changes to distribution take effect immediately, allowing for real-time optimization as data comes in.

Framer's approach differs from traditional staged rollouts in that the test continues running even after distribution changes. Metrics accumulate across the entire test duration, not just within each stage. This means users can shift traffic toward a winning variant while still collecting data to validate the decision.

The feature is available to all Framer users with access to A/B testing, which is included in paid plans. No additional configuration or code is required beyond setting the distribution percentages in the test settings.

Starting small reduces risk. A homepage test that tanks conversion rates will do less damage if it only reaches 10% of visitors initially. Once the data validates the hypothesis, increasing distribution becomes a calculated decision rather than a leap of faith.

This approach also helps teams move faster. Rather than waiting for unanimous approval to launch a test at full scale, designers and marketers can propose a limited rollout, gather proof, and expand based on performance. It shifts the conversation from "should we test this?" to "how much traffic should we allocate?"

Custom Distribution also supports gradual rollouts for non-test scenarios. A site-wide design update can go live to 5% of visitors first, then 20%, then 50%, with monitoring at each stage. If issues arise, distribution can be dialed back without fully reverting the change.

Following their recent CMS 3.0 overhaul, Framer continues to add workflow improvements for teams managing high-traffic sites. Custom Distribution fits into a broader trend of giving users finer control over how and when changes go live, similar to feature flags in software development but tailored for site builders.

The feature is live now for all Framer accounts with A/B testing access. More details are available on Framer's site.