Using AI-Powered Predictive Audience Plays

LXRInsights' AI-Powered Predictive Audience Plays help marketing teams activate customer intelligence through channel-ready audience recommendations. Each play is built using customer value, purchase behavior, predicted future value, and engagement signals.

The framework begins by segmenting customers into three groups:

High-Value Customers (HVC)

Customers with the highest revenue contribution, strongest lifetime value, and highest purchase potential.

Common use cases include:

• Loyalty initiatives
• Premium product promotion
• Repeat purchase campaigns
• Lookalike audience creation
• High-value customer acquisition

Mid-Value Customers (MVC)

Customers with strong growth potential and opportunities for increased engagement and spend.

Common use cases include:

• Upsell campaigns
• Cross-sell initiatives
• Repeat purchase programs
• Loyalty nurturing
• Future VIP identification

Low-Value Customers (LVC)

Customers with lower engagement levels, limited purchase history, or lower projected future value.

Common use cases include:

• Re-engagement campaigns
• Second-purchase initiatives
• Product recommendation campaigns
• Audience suppression strategies
• Customer nurture programs

Understanding the Customer Score

LXRInsights assigns every customer a proprietary score between 0 and 100 based on a combination of behavioral, transactional, and engagement signals. The score is designed to reflect a customer's overall value and future revenue potential.

Factors considered include:

• Purchase frequency
• Average order value (AOV)
• Lifetime value (LTV)
• Recency of purchase
• Product preferences
• Site engagement
• Predicted future purchase behavior

Customers are then grouped into value tiers:

High-Value Customers (HVC) | Score: 75–100
Top revenue contributors with the strongest lifetime value and highest purchase potential.

Mid-Value Customers (MVC) | Score: 55–75
Customers with strong growth potential who may be ready for upsell, repeat purchase, or progression into higher-value segments.

Low-Value Customers (LVC) | Score: Below 55
Customers with lower engagement, limited purchase history, or lower projected future value.

These scores are refreshed regularly, allowing marketers to track customer movement between segments and activate the most relevant audiences for acquisition, retention, reactivation, and loyalty initiatives.

‍

Audience Plays in Action

LXRInsights generates pre-configured audience plays that can be activated across Google, Meta, Email, SMS, and Loyalty channels.

Examples include:

• Build Google Performance Max seed audiences from High-Value Customers
• Create Meta lookalike audiences from top-performing customers
• Identify customers most likely to purchase again
• Launch churn prevention campaigns
• Retarget one-time buyers
• Develop Mid-Value Customers into future VIP segments

Each audience play includes recommended filters, activation channels, and suggested marketing actions.

‍

Example Audience Plays by Segment

High-Value Customer Plays

Find Customers Most Likely to Purchase Again

Recommended Channels:

• Email
• Meta
• Google
• SMS

Audience Criteria:

• High purchase probability
• Recent purchasers
• Above-average AOV
• Multiple orders

Recommended Actions:

• Replenishment campaigns
• Product recommendations
• Upsell initiatives
• SMS reminders

Why it works:

These customers already demonstrate strong purchase intent and typically deliver the highest return on retention efforts.

Build Google PMax Seed Audiences

Recommended Channels:

• Google Ads
• Performance Max

Audience Criteria:

• HVC score above 80
• High conversion probability
• Strong revenue contribution

Recommended Actions:

• PMax audience seeding
• Smart bidding optimization
• New customer acquisition

Why it works:

Instead of letting Google's algorithm learn from all customers equally, brands can teach it which customers matter most.

Scale Best Customers into New Acquisition

Recommended Channels:

• Meta
• Google

Audience Criteria:

• HVC score above 85
• Strong revenue contribution
• High future value prediction

Recommended Actions:

• Lookalike audiences
• Prospecting campaigns
• Customer expansion initiatives

Why it works:

Your best future customers often resemble your best current customers.

Recover Customers Before They Churn

Recommended Channels:

• Email
• Meta Retargeting

Audience Criteria:

• High churn probability
• Declining engagement
• Previously strong spend behavior

Recommended Actions:

• Win-back campaigns
• Personalized offers
• Dynamic retargeting

Why it works:

Retention is often more profitable than acquisition.

Mid-Value Customer Plays

Identify Future VIP Customers

Recommended Channels:

• CRM
• Email
• SMS

Audience Criteria:

• MVC score between 65–75
• Strong purchase activity
• High future value prediction

Recommended Actions:

• Loyalty programs
• Early VIP segmentation
• Personalized nurturing

Why it works:

These customers frequently become tomorrow's highest-value customers.

Low-Value Customer Plays

Retarget One-Time Buyers

Recommended Channels:

• Email
• Meta

Audience Criteria:

• Single purchase history
• Recent activity
• Low engagement

Recommended Actions:

• Product follow-up campaigns
• Cross-sell offers
• Bundle recommendations

Why it works:

Moving a customer from one purchase to two purchases often has a significant impact on lifetime value.

Suppress Low-Quality Acquisition Traffic

Recommended Channels:

• Google
• Meta

Audience Criteria:

• Low predicted future value
• Low AOV
• One-time buyer behavior

Recommended Actions:

• Reduce prospecting spend
• Exclude from acquisition seeds
• Shift budget toward higher-value audiences

Why it works:

Not all customers are equally valuable. Growth improves when acquisition focuses on quality rather than volume.

‍

Operationalizing Customer Intelligence

Predictive Audience Plays provide a structured way to activate customer intelligence across marketing channels. By organizing audiences based on customer value and intent, teams can align acquisition, retention, reactivation, and loyalty efforts with the customers most relevant to each objective.

The result is a repeatable framework for turning customer insights into actionable marketing programs across the customer lifecycle.

‍