Quickstart: Audiences
Build your first audience and activate it across marketing channels. This guide takes you from raw customer data to a targeted segment syncing to your ad platform in under 15 minutes.
Prerequisites
- A connected source with customer data (see Quickstart: Data Pipeline)
- At least one model with customer attributes
- A destination configured for audience activation
Step 1: Define Computed Attributes
Computed attributes are derived metrics about your customers used to build audiences.
- Navigate to Sources > Computed Attributes
- Click Create Computed Attribute
- Choose a computed attribute type:
- SQL Computed Attribute — Write custom SQL to compute any metric
- Aggregation Computed Attribute — Use a visual builder for counts, sums, averages
- Formula Computed Attribute — Combine existing computed attributes with expressions
- Example: Create a “High Value” computed attribute:
SELECT
customer_id,
CASE WHEN lifetime_value > 500 THEN true ELSE false END as is_high_value
FROM customers- Save and run the computed attribute evaluation
Step 2: Create an Audience
Audiences are groups of customers defined by computed attribute conditions and filters.
- Navigate to Activation > Audiences
- Click Create Audience
- Use the visual filter builder:
- Add condition:
is_high_valueequalstrue - Add condition:
last_purchase_dateis within last 90 days
- Add condition:
- Click Estimate Size to see how many customers match
- Click Preview to see sample profiles
- Name your audience (e.g., “High Value Recent Purchasers”)
- Click Save
Step 3: Activate the Audience
- Navigate to Destinations > Syncs
- Click Create Sync
- Select your audience
- Select a destination (e.g., Facebook Ads, Google Ads)
- Choose a sync mode:
- Mirror — Keeps the audience perfectly in sync (adds and removes members)
- Additive — Only adds new members, never removes
- Subtractive — Only removes members who no longer qualify
- Set a schedule
- Click Save & Run
Next Steps
- Explore the filter builder for complex conditions
- Create A/B tests for A/B testing
- Set up priorities to manage overlap
Last updated on