The CMO Podcast with Fexingo · 2026-07-08 · 10 min
In this milestone 100th episode, Lucas and Luna dig into a specific marketing analytics tactic: using machine learning models to predict customer churn before the customer ever signals intent to leave. They walk through a real case from a mid-market SaaS company that reduced churn by 22 percent in two quarters by acting on AI-generated risk scores. Lucas explains the difference between a basic logistic regression and a gradient-boosted model in plain language, why most churn models fail because of stale data, and the one metric CMOs should track instead of retention rate. Luna pushes back on the cost of implementation and asks whether smaller teams can realistically build these models in-house. The hosts also touch on the ethics of predictive targeting and what happens when a model flags a customer who wasn't actually at risk. Concrete, actionable, and grounded in a real number.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
AI Meets the Mid-Market: How PE-Backed Companies Are Leapfrogging with AIDisambiguation · on Predictive analytics89 / 100
Mo (Mohammed) Saadat from StratahubEnergytech Startups · on Predictive analytics86 / 100
Privacy, Power, and Prediction: Reclaiming Control of Your Personal Data, with Carissa Véliz | Lawyerist PodcastLegal Talk Network · on Predictive analytics85 / 100
B2B Storytelling: Michael Brenner on Empathy-Driven ContentBeyond B2B Marketing · on customer retention83 / 100
Finding the Students Schools MissTech & Learning Conversations Podcast · on Predictive analytics81 / 100
Why B2B Brands Are Using AI for Customer Health ScoringThe Growth Operator with Fexingo · on Churn prediction77 / 100
How Glossier Built a Community-First Beauty Brand72 / 100How CMOs Are Using Gamified Loyalty Programs75 / 100
How CMOs Are Using AI Agents for Customer Retention85 / 100How CMOs Use AI to Predict Campaign Performance Before Launch82 / 100How CMOs Are Using Branded Newsletters to Drive Revenue73 / 100