Build Audience Rules

Audience rules define who qualifies for an audience. Build rules from the business question first, then translate each requirement into a group, condition, operator, value, and time window.

Visual Builder and JSON view

Use Visual Builder for normal editing. It shows conditions as editable controls and is the safest way to build or review logic.

Use JSON when you need to inspect the structured audience definition or copy it for technical review. Do not edit JSON unless your team has confirmed the expected payload shape.

Root match mode

The root match mode controls how the top-level rules work together.

Mode	Meaning	Use it when
ALL (AND)	Users must match every top-level condition or group	The audience must satisfy multiple requirements
ANY (OR)	Users can match at least one top-level condition or group	Several different behaviors should qualify users

For high-risk campaign audiences, start with ALL (AND) unless the business question explicitly calls for alternatives.

Groups

Groups keep related logic together. Use Add group when rules need their own match behavior.

Group logic can include:

Group mode	Meaning	Example
Match all	Every condition inside the group must be true	User is in Delhi and lifecycle stage is lead
Match any	At least one condition inside the group must be true	User viewed pricing or added to cart
Exclude matches	Matching users are removed from the audience	User purchased in the last 30 days

Use exclusion groups for buyers, unsubscribed users, suppressed users, internal test users, or users who are not eligible for the intended channel.

Conditions

Use Add condition to create one rule. A condition contains these controls:

Control	What to select	Example
Source	The entity or source where the field belongs	Events
Attribute	The specific field to evaluate	Event Type
Operator	The comparison to apply	equals
Value	The value to match	add_to_cart
Time window	The period where the condition applies	in last 30 days

Observed source choices include Customer, Orders, Predictions, Purchase, Events, and Campaign.

Operators

Choose an operator that matches the field type and intent.

Operator	Meaning	Example use
equals	Field must match one value	Event Type equals purchase
not equals	Field must not match one value	Lifecycle Stage not equals customer
in	Field must match one of several values	Country in India, UAE, Singapore
not in	Field must not match any listed value	Campaign not in internal test campaigns
is set	Field must have a value	Email is set
is not set	Field must be empty or unavailable	Mobile is not set

Time windows

Time windows make behavior rules specific. The observed builder supports window modes such as in last, in next, and between, with units such as days, weeks, and months.

Use short windows for recent intent and longer windows for lifecycle or retention analysis.

Goal	Suggested window
Cart recovery	Last 1 to 7 days
Paid retargeting	Last 7 to 30 days
Repeat purchase	Last 30 to 90 days
Churn risk review	Last 30 to 180 days

Duplicate and delete conditions

Use Duplicate condition when a new condition is similar to an existing one. Change the field, operator, value, or time window after duplicating.

Use Delete condition when a rule no longer belongs in the audience. After deleting, validate and estimate again.

Clear all

Use Clear all only when the current rule set should be removed. Copy the JSON or duplicate the audience first if another team may need the current definition.

Example: cart abandoners

Use this rule pattern to find users who added to cart but did not purchase.

1. Set root match mode to ALL (AND).
2. Add an include condition: Events > Event Type > equals > add_to_cart > in last 30 days.
3. Add an exclusion group with Exclude matches.
4. Add an exclusion condition: Events > Event Type > equals > purchase > in last 30 days.
5. Validate and estimate.
6. Review preview users for recent buyers. If recent buyers appear, check the exclusion rule and time window.

QA checklist

• Root match mode matches the audience logic.
• Include rules describe positive qualification criteria.
• Exclusion rules remove users who should not receive the campaign or analysis.
• Operators match the selected field type.
• Values are spelled exactly as the source data stores them.
• Time windows match the business question.
• Duplicate rules were changed after copying.
• Empty or outdated rules were deleted.

Common issues

Issue	Cause	Fix
Audience is too small	Rules are too strict or exclusions remove too many users	Switch a group from Match all to Match any or broaden the time window
Audience is too broad	Root mode or group mode allows too many matches	Use ALL (AND), add exclusions, or narrow values
Users who purchased still qualify	Purchase exclusion is missing or uses the wrong window	Add an Exclude matches group for purchase behavior
Validation flags a rule	Required field, value, or time window is missing	Focus the flagged condition and complete the missing control
JSON does not match the visual logic	Visual rules changed after JSON was copied	Copy JSON again after the final validation

Related pages

• Validate, Estimate, and Preview
• Save, Export, and AI Generation