Testing AI-Informed Audience Segmentation at Boston Ballet

The Challenge

Boston Ballet’s FY25 single-ticket email program was already delivering strong results through targeted promotions, seasonal campaigns, and audience segmentation.

The team wanted to evaluate whether AI could generate practical audience segmentation ideas, validate existing strategies, and uncover new opportunities worth testing in future campaigns.

Smaller, behavior-driven audience segments often outperformed broader campaign sends, raising an important question: could AI help surface audience opportunities hidden within existing data and provide new ideas for future audience engagement and revenue growth?

The Approach

Boston Ballet combined internal campaign expertise with AI-powered pattern detection to analyze:

• FY25 single-ticket sales performance
• Audience buying behavior
• Purchase timing
• Discount usage
• Email engagement trends

The goal was to evaluate whether AI could help generate actionable audience segmentation ideas that could be tested within Boston Ballet’s existing marketing program.

AI analysis surfaced 55 potential audience segments and behavioral triggers based on:

• urgency patterns
• purchase timing
• pricing sensitivity
• seat preferences
• repeat attendance behavior
• engagement habits

Working together, Boston Ballet and Capacity evaluated each opportunity based on audience size, technical feasibility, expected impact, and implementation effort. The team ultimately narrowed the recommendations to three practical pilot campaigns that could be tested within existing systems and workflows.

What the Data Revealed

One insight quickly stood out: relevance consistently outperformed volume.

Boston Ballet’s strongest FY25 campaigns combined:

• clear urgency
• behavioral targeting
• timely messaging
• audience-specific offers

Meanwhile, broader awareness emails and generic no-offer campaigns generated far lower engagement and revenue.

The analysis also revealed that Boston Ballet’s audience behaved less like one customer base and more like several distinct buyer types:

• early planners
• last-minute buyers
• full-price loyalists
• discount-sensitive patrons
• habitual seat-location buyers
• urgency-driven shoppers

The opportunity wasn’t simply sending more emails. It was matching messaging, timing, and offers to actual buying behavior.

The First Live Test: Early-Access Audience

The first pilot campaign focused on one of the highest-potential AI recommendations: the “Flash-Action Cohort.”

Boston Ballet identified patrons who had previously engaged with urgency-driven campaigns like:

• Cyber Sale
• Halloween Sale
• other limited-time promotions

Those audiences received:

• a dedicated early-access email
• 24-hour access before the public Valentine’s Day sale
• a unique tracking code for attribution

The strategy tested whether behavior-based segmentation could drive earlier conversions and incremental revenue without increasing overall campaign volume.

The Results

The Valentine’s Day Early-Access Test provided an encouraging first look at how AI-informed audience segmentation could be applied within Boston Ballet’s email strategy.

Early-Access Segment Performance

• 6.2K eligible contacts
• 58% open rate
• 3.5% click rate
• 15 purchasing households
• 28 tickets sold
• $2.4K  in attributed revenue

The test demonstrated how AI-informed audience segmentation could be translated into a practical, low-lift campaign test within Boston Ballet’s existing marketing program.

Additional Tests and Learnings

Following the Early-Access Test, Boston Ballet launched two additional AI-informed audience experiments:

• Zone Loyalists targeted patrons with a history of purchasing orchestra seats and emphasized preferred seating availability.
• Full-Price Champions focused on patrons who consistently purchased full-price tickets and highlighted early access to seat selection without discounts.

The tests provided valuable insight into patron preferences and helped inform future segmentation strategies.