Frequently Asked Questions

For A/B testing with randomized number distributions, Statsig's experiments feature is the recommended approach.

To achieve this, you should create multiple test groups within an experiment. Each user who interacts with your application will be deterministically assigned to one of these test groups based on their user ID, which could be either a stableID or userID.

This assignment ensures that each user consistently experiences the same test group during the experiment. Test groups can be configured with associated parameters, allowing for dynamic control over the behavior and experience of each test group.

This setup enables you to conduct A/B/n experiments where users are evenly distributed across the different configurations you wish to test.

For detailed guidance on configuring your groups and parameters, refer to the official Statsig documentation on creating and managing experiments.