User Guides
Features and Rules
How to create and manage experiment features and traffic allocation rules.
Features and rules are the core building blocks of Mojito experiments.
Features
A feature represents an experiment or feature flag. Each feature has:
- A human-readable name
- A unique key used by the SDK to identify the experiment
- One or more variants representing different experiences
- One or more rules that control traffic allocation
Rules
A rule defines how traffic is distributed across variants for a given feature:
- Sample rate — The percentage of eligible users who enter the experiment
- Variant distribution — How included users are split between variants
- Lifecycle state — Rules progress through Draft, Staging, Live, and Stopped states
Rule Lifecycle
| State | Description |
|---|---|
| Draft | Initial state. Configuration can be edited freely. |
| Staging | Published for QA testing. Not yet serving live traffic. |
| Live | Actively serving traffic to users. |
| Stopped | Experiment completed. A winning variant can be selected. |
Editing Live Rules
Live rules are protected to help preserve experiment integrity while they are serving traffic.
- Allowed while live: update traffic allocation, edit rule-level payloads, and edit variant payloads.
- Not allowed while live: rename the rule, change variant names or IDs, or add or remove variants.
To make structural changes to a rule, create a new one, update it in staging, and replace the live rule when it is ready.
Managing Features
From the property dashboard, you can:
- Create new features
- Edit feature names and keys
- Archive features that are no longer needed
- View the history of rule changes on the timeline