Create Audiences

Audiences are saved segments of your contacts. Use them to target specific groups when sending emails or starting automations.

Two Ways to Create an Audience

Define the audience

Create with AI
Build by hand

On the Audience page, describe what you want in the Ask Brew to filter for box. Brew generates the filters and a descriptive name automatically.Examples:

Customers who signed up in the last 30 days but haven’t opened any emails yet.

Contacts who clicked our product launch email and visited our pricing page.

Enterprise customers who have been subscribers for more than 6 months.

Review the generated filters and name, and adjust anything before saving.

On the Audience page, click + Filter
Add filters one at a time:
- Pick a property or behavior
- Choose an operator (equals, contains, before, after, greater than, etc.)
- Enter the value (the table and count update as you type)
Combine filters with AND or OR

Available filter dimensions:Contact properties

Email domain
Location
Any custom property you’ve added

Custom properties are what make the richer filters possible. Signup date, plan, feature usage, and last purchase are all things you sync onto the contact first, then filter on. “Signed up in the last 30 days” works once signupDate exists on your contacts.A date property supports before, after, on_date, between, is_empty, and is_not_empty. Filter dates are fixed and never roll forward on their own. For “the last 30 days”, filter after a date and move it forward when you reuse the audience. on_date matches the whole UTC day, while before and after compare against midnight. So a time on that day counts as both on and after it.Email engagement

Opens (did open / didn’t open)
Clicks (did click / didn’t click)
Sent (was sent / wasn’t sent)

Subscription status

Subscribed / unsubscribed

Review who matches

Brew shows the filters that define your audience plus a preview of matching contacts and the total count. The count includes unsubscribed members; a send skips them.

Name and save

Click Create new audience in the bar at the bottom of the page and name it. The audience is now available when sending emails or starting automations.To change a saved audience, open it and edit its filters. The bar then also shows Save, which updates the audience in place.

Create with an Agent or API

Call create_audience over MCP, brew.audiences.create in the TypeScript SDK, brew-cli audiences create --input in the CLI, or POST /v1/audiences with a name and filter definition. These paths save the same audiences as the app. Saved audiences resolve omitted filter types from the brand’s field registry before counting and saving. If tier is registered as a number, this filter excludes tier 10:

{
  "name": "Everyone except tier 10",
  "filters": {
    "logicalOperator": "and",
    "filters": [
      { "field": "tier", "operator": "not_equals", "value": 10, "type": "number" }
    ]
  }
}

Explicit types remain supported. Unknown fields use string comparisons; supply their type when you need a typed comparison. For date ranges, pass type: "date" so request validation can parse the operands.

Size and Edit an Audience from an Agent or Code

Current size. An audience’s stored count is cached. GET /v1/audiences/{audienceId}?include=count (or list_audiences with include: ["count"]) computes it fresh. Either way it counts every member, unsubscribed ones included.
Counts and breakdowns. search_contacts or POST /v1/contacts/search with count: true returns how many contacts match. Add groupBy (up to two fields, or emailDomain) or bucket (day, week, or month) for a breakdown.
Copy. create_audience with sourceAudienceId, or POST /v1/audiences/{audienceId}/duplicate, copies the filters. Contacts aren’t copied; both audiences match the same contacts.
Add or remove specific people. update_audience takes addEmails and removeEmails.

Using Saved Audiences

Saved audiences appear in two places:

The Audience page. Click the chip showing the selected audience to switch between saved ones
The send dialog. When sending an email or starting an automation, pick a saved audience as the recipients

For any selected audience, you can:

Download as CSV. Export the matching contacts
Validate contacts. Run a deliverability check on every contact in the audience (2 credits per address)
Delete audience. Remove the audience definition (your contacts stay; just the saved filter is removed)

Deleting an audience is permanent. The filter and configuration are gone. Your contacts aren’t affected.

Audiences are dynamic. Brew stores the filter criteria, not a frozen list of contacts. As contacts are added, updated, or change engagement, they automatically enter or leave audiences based on the filters.One exception: an audience built from more than 100 email addresses saves a snapshot of the contacts that existed then. A contact added later isn’t backfilled, so build a new audience to include it.

Engagement Snapshots

A saved audience filters on what a contact is: their properties, their status, whether they have ever opened or clicked. It cannot filter on when that happened. Engagement lives in your event history, not on the contact record. For a cohort like “everyone who opened in the last 30 days”, build it from events instead. Over MCP, create_audience_from_events freezes the matching contacts into an audience. From code, POST /v1/audiences/from-events takes the same fields:

Events. One to 10 event types, such as opened or clicked.
Window. A fixed date range of up to 90 days, or the window of one send.
The opposite cohort. cohort.exclude keeps everyone in a base audience who did not match. That’s how you build “no opens in the last 60 days”.
Wait for it. The audience can’t be sent to until its build is ready. Check with list_audiences and include: ["build"].
What it saves. Each build adds a date field that stamps every member with their latest matching event. Don’t delete that field while a build is queued, or the build fails. A failed build needs a new snapshot.
The base is copied. cohort.exclude copies the base audience’s filters when the build starts. Editing or deleting the base later doesn’t change the snapshot. A base whose filters use OR is refused.

The tradeoff is in the name. A snapshot is frozen at the moment you create it, so it doesn’t re-evaluate the way a saved filter does. Rebuild it when you need it fresh.

Common Audience Patterns

Engagement-based

A time window on opens or clicks needs an engagement snapshot, because a saved filter can’t tell when engagement happened. Rebuild the snapshot to refresh it.

Highly engaged. Opened or clicked in the last 30 days
Moderately engaged. Opened or clicked in the last 90 days
At risk. No opens in the last 60 days but previously active
Inactive. No engagement in 90+ days

Behavioral

Power users. Used a specific feature more than X times
Recent buyers. Last purchase date after a fixed date, such as 30 days ago
Cart abandoners. Started checkout but didn’t complete
High-value. Total spend above a threshold

Demographic

By role. Group by job title or department
By industry. Target specific verticals
By company size. Small business vs enterprise
By geography. Region or country

Lifecycle

New subscribers. Joined in the last 7 days
Trial users. Currently in trial
Recent converters. New paying customers
Long-term customers. Subscribed 6+ months

To send one email in a version for each segment, follow Send a Segmented Newsletter.

Best Practices

Start with engagement tiers

Build basic engagement segments (highly engaged, moderately engaged, dormant) before getting fancy.

Use descriptive names

Name audiences so your team understands their purpose. Active Enterprise Trials Q2 beats Audience 7.

Smaller is better

Targeted audiences usually outperform broad sends.

Test before scaling

Validate messaging with a small audience before sending to a large one.

Need Help?

Our team is ready to support you at every step of your journey with Brew. Choose the option that works best for you:

Self-Service Tools
Talk to Our Team

Search Documentation

Type in the “Ask any question” search bar at the top left to instantly find relevant documentation pages.

Ask ChatGPT or Claude

Click here to open the Copy page menu at the top of this page, then send the page to ChatGPT, Claude, or your editor.

Get Started

Emails and Automations

Review and Send

Brand, Assets, and Domains

Audience

Analytics

Settings and Security

Two Ways to Create an Audience

Create with an Agent or API

Size and Edit an Audience from an Agent or Code

Using Saved Audiences

Engagement Snapshots

Common Audience Patterns

Best Practices

Start with engagement tiers

Use descriptive names

Smaller is better

Test before scaling

Need Help?

Search Documentation

Ask ChatGPT or Claude

Schedule a Call

Call Us Directly

Slack Channel

Email Support

​Two Ways to Create an Audience

​Create with an Agent or API

​Size and Edit an Audience from an Agent or Code

​Using Saved Audiences

​Engagement Snapshots

​Common Audience Patterns

​Best Practices

Start with engagement tiers

Use descriptive names

Smaller is better

Test before scaling

​Need Help?

Search Documentation

Ask ChatGPT or Claude

Schedule a Call

Call Us Directly

Slack Channel

Email Support

Two Ways to Create an Audience

Create with an Agent or API

Size and Edit an Audience from an Agent or Code

Using Saved Audiences

Engagement Snapshots

Common Audience Patterns

Best Practices

Need Help?