Peloton’s Big Bet on Personalization with Agentic AI Decisioning

Session Summary

In this session from MoEngage’s Agentic Marketing Tour, Steven Balagula, Director of Global Lifecycle Marketing at Peloton, and Paul Meinshausen, CEO of Aampe (now part of MoEngage), discuss how Peloton is moving beyond static, rules-based lifecycle marketing toward autonomous agentic decisioning. Steven shares how he built executive buy-in by connecting personalization directly to member retention and customer lifetime value, launched a 50/50 pilot on a high-stakes audience, and overhauled Peloton’s creative content pipeline to support autonomous variation across tone, value propositions, and send timing. Together, they outline what it takes to manage AI agents, establish human guardrails, and evolve marketing teams into high-performing agentic organizations.

Top Takeaways

• Tying AI Pilots to Executive Value Metrics: Secure leadership buy-in by connecting personalization directly to retention and customer lifetime value rather than pitching AI as an isolated channel tactic.

• Testing High-Stakes Audiences with Controlled Baselines: Validate agentic decisioning by running disciplined 50/50 split tests against top-performing legacy campaigns to prove incremental lift on core business metrics.

• Overhauling Content Architecture for Machine Learning: Autonomous decisioning requires deep, modular content libraries categorized by tone, value proposition, and offer so algorithms have sufficient variety to test and optimize.

• Evolving Roles to Manage Agentic Systems: Marketing leadership is shifting from manual campaign builds toward defining agent reward metrics, setting brand guardrails, and overseeing multi-channel orchestration.

• Expanding from Single Channels to Omnichannel Experiences: While email often serves as the initial proving ground, the greatest value of agentic systems emerges when extending 1:1 personalization seamlessly across apps, web, and owned platforms.