Segmentation and Activation

Segmentation groups customers based on traits, behaviors, predictions, or campaign context. Activation sends those audiences to channels where the team can personalize messages, suppress users, or trigger journeys.

Segment building blocks

Segments can use:

• Profile traits: lifecycle stage, location, plan, account type, language.
• Behavioral events: viewed product, started signup, purchased, renewed, churned.
• Campaign context: source, medium, campaign, ad group, creative.
• Predictive signals: conversion likelihood, churn risk, LTV band, engagement trend.
• Eligibility rules: consent, channel permission, suppression, frequency caps.

Common segment examples

Segment	Example criteria	Use case
High intent	Pricing page viewed and no purchase in last 7 days	Retargeting or sales follow-up
Abandoned cart	Add to cart without checkout	Email, SMS, or paid retargeting
Churn risk	Engagement down and subscription renewal approaching	Retention campaign
Loyal customer	Repeat purchase and high engagement	Upsell or advocacy
Fraud review	Suspicious source or event pattern	Exclusion or manual audit

Static vs dynamic segments

Static segments are snapshots. Use them for one-time analysis, audits, or controlled campaign uploads.

Dynamic segments update as customer data changes. Use them for ongoing lifecycle journeys, real-time suppression, and behavior-triggered activation.

Activation controls

Before sending a segment to a channel, confirm:

• The audience has a clear business purpose.
• Consent and channel eligibility rules are applied.
• Suppression lists are included.
• Frequency or recency limits are set when needed.
• The destination channel accepts the selected identifiers.
• A test audience or preview is reviewed.

warning

Do not activate audiences that contain sensitive, restricted, or poorly understood attributes unless your organization has explicitly approved that use.

AI-driven segmentation

AI-assisted segmentation can help teams move from static rules to intent-aware audiences. Examples include:

• Users likely to convert in the next period.
• Customers showing churn risk.
• Users whose engagement is increasing.
• Customers likely to respond to a specific channel.
• Cohorts with high predicted LTV.

Use predictive signals as decision support. Validate performance through reports, holdouts, or experiments before making large budget shifts.

Measurement after activation

After launch, monitor:

• Audience size and match rate.
• Delivery rate by channel.
• Conversion rate.
• Revenue or value generated.
• Incrementality where a holdout exists.
• Unsubscribe, opt-out, or complaint rate.
• Segment fatigue over time.