Amplitude cohort sync
Sync an Amplitude cohort into a Statsig ID List segment so you can target gates, experiments, and dynamic configs at an audience you already defined in Amplitude.
Amplitude cohort sync copies the membership of an Amplitude cohort into a Statsig ID List segment. After the segment exists, you can target feature gates, experiments, and dynamic configs at the same audience you already defined in Amplitude.
The sync carries user identifiers only, not events. Amplitude pushes the membership list on a cadence you choose, and Statsig stores each cohort as one ID List segment.
Cohort sync authenticates with a Statsig Console API key, so it doesn't depend on a link between your Statsig and Amplitude organizations. You configure the sync in Amplitude, not in Statsig.
Prerequisites
- A saved cohort in Amplitude.
- Admin access to the Statsig project, so you can create a Console API key.
Create a Console API key
- In the Statsig console, go to Project Settings > Keys & Environments.
- Click Generate New Key and select Console as the key type.
- Copy the key.
Cohort sync requires a Console API key. A Server Secret Key authenticates the event-streaming connector, not this one.
Configure the sync in Amplitude
Add Statsig as a cohort destination in Amplitude, paste the Console API key, then sync a cohort to it. For the procedure, refer to Statsig (Cohort Sync) in the Amplitude documentation.
Find the synced cohort in Statsig
After the first sync runs, the cohort appears under Segments in the Statsig console as an ID List segment. Later changes to the cohort reach Statsig within about 15 minutes of each sync.
Target a synced cohort the same way you target any other segment. Refer to Using a segment for how to apply a segment in a gate, experiment, or dynamic config.
Size limits
A synced cohort carries the same limits as any other ID List segment:
- A project holds at most 10 million IDs across all ID Lists.
- Lists over 1,000 IDs use a separate synchronization process, so changes take longer to propagate and SDK support varies.
Refer to Adding ID Lists for the detail behind both limits. To target an audience of hundreds of thousands of users or more, pass an attribute on the Statsig user object and target on that attribute instead of syncing a cohort.
Was this helpful?
For AI agents: a documentation index is available at /llms.txt. Append .md to any page URL for markdown, or send Accept: text/markdown.