Discord Experiment Rollouts

What are Discord Experiments / Rollouts?

Discord Experiments are upcoming features or interface changes that Discord is currently testing and developing. These experiments allow Discord to introduce potential new features to a limited number of users or servers to monitor how they perform before deciding whether to launch them to everyone. The technical process often involves A/B testing, where certain users or guilds (Discord servers) are randomly or selectively given access to new features or changes, while others continue to see the regular version. This enables Discord to compare behaviors or gather feedback to improve the platform.

How Discord Experiments and Rollouts Work

• A/B Testing Approach: Experiments are rolled out using an A/B testing methodology, where groups of users (or entire guilds) are assigned to "treatment” (receiving the new feature) or "control” (not receiving it).
• Random or Filtered Assignment: Assignment to experiments can be random or may be filtered by certain criteria, such as location, client version, or account age.
• Rollout Tracking: Each experiment maintains a "rollout status,” showing how much of the user base or how many guilds have the feature enabled. Rollouts can be very limited or gradually expanded to broader populations as testing progresses.
• Technical Details: At a technical level, rollouts are defined in Discord's admin tools, using configuration files that mark which experiments go to which users or guilds (populations).
• Persistence: For users not yet signed in, Discord uses a "fingerprint” (a unique identifier) to track which experiments a user should see. When a user creates an account, this fingerprint can persist the experiment assignments through registration.

Types of Discord Experiments:

• User Experiments: Targeted at individual users; for example, introducing a new settings page or UI element just for them.
• Guild Experiments: Targeted at specific servers; for example, giving a server access to polls or slowmode options.

Examples

Recent and past experiments have included:

• Emojis in Channel Names
• Skill Trees
• Custom Username Fonts
• General Voice Channel Notifications

How Rollouts Are Assigned

• The assignment can utilize a hash based on your server ID or experiment name, producing a deterministic but apparently random distribution across Discord’s user base. For guild features, this means some servers get access while others do not, regardless of activity or Nitro status.
• There is no way for most users to opt-in or guarantee early access to experiments. Sometimes, accessing early features is possible by using Discord's Canary or Beta clients, but this does not always work for all experiments.
• Some technical users may use unofficial tools or modified clients to visually enable experiments, but these do not enable real API interaction and are unsupported.

This tool is not in any way associated or endorsed by Discord.