Audience Segments

The Audience Segments node computes and retrieves audiences — the set of records that currently match a saved segment. A segment is a saved filter rule over a data entity (for example, "customers in the EU who ordered in the last 30 days"), not a static list. Use it to target campaigns, drive personalization, or feed a downstream sync with exactly the people who currently qualify.

Overview

Three operations are available. Understanding which to use — and when — prevents stale or truncated results.

Audience Segments Node configuration

Compute Segment

Runs the segment rule and stores the resulting audience for later retrieval. Returns a count, not the member records. Can run asynchronously: the node kicks off the computation and returns without waiting for it to finish.

Audience Segments Node additional configuration

Output: The count of records that matched the segment rule. The audience is stored and can be retrieved with Get Audience.

Compute and Get

Runs the segment rule and returns the matching members in the same call. Intended for small segments only. Results are effectively truncated beyond approximately 1,000 records — do not use this operation for large audiences.

Output: An array of the matching records, returned inline. Truncated if the audience exceeds the size limit.

Do not use Compute and Get for audiences that may grow beyond ~1,000 records. Use Compute Segment followed by Get Audience (with pagination) for any sizable audience.

Get Audience

Reads back the members of an already-computed audience. Supports pagination via cursor and page size. Returns whatever was last stored by a Compute Segment run — if the segment has never been computed (or was deleted), the result will be empty.

Output: A page of audience member records and a cursor for the next page (empty when the last page is reached).

Delete Audience

Removes a stored audience (the computed result set). The segment definition itself is not affected — only the stored audience data is deleted.

Output: Confirmation that the stored audience was deleted.

Notes

The Audience Segments node works best when you match the operation to your audience size and keep data freshness in mind:

• The reliable pattern for large audiences: run Compute Segment first, then use Get Audience with pagination to page through the members.

• Get Audience returns stale data if recomputed infrequently. It reads whatever was last stored. Always compute before reading if you need current membership.

• Compute and Get is not suitable for large segments. Results are truncated beyond approximately 1,000 records. Use it only for small, bounded audiences.

• Asynchronous computation (on Compute Segment) means the node returns before the audience is fully stored. If the next step immediately calls Get Audience, the audience may still be computing. Add a brief wait or design the flow to compute in an earlier step.

• Deleting a stored audience does not delete the segment definition. The segment can be re-computed at any time.

A well-designed Audience Segments workflow uses Compute Segment for large audiences and always re-computes before reading when current membership is required.