Customer Experience University - Winning Loyalty & Engagement One Customer at a Time
Aug 7, 2026 · 9 min · 8 segments
A discount that only appears when you threaten to cancel isn’t loyalty — it’s a confession. Joseph Michelli unpacks the “smoke detector test” for predictive churn AI: the same model can serve…
Joseph MichelliHost
Let me start with two devices that do almost exactly the same thing and could not be more different.

It sits up there on your ceiling, watching the air all day, every day for the very first hint of trouble.

Now, a spy also watches closely, constantly, but for its own purposes, not yours.

The moment you discover it's been watching, every ounce of trust you had just evaporates all at once.

The predictive AI that's now scanning your customer base for the earliest signs that someone's about to leave, that AI can be either one of two things.

Does your AI exist to serve the person or to surveil them? And if you get that wrong, the very system you built to reduce churn becomes the reason people leave.

Now, before you think this is soft, let me give you the economics, because retention might be the most leveraged number in your entire business.

Let me start with two devices that do almost exactly the same thing and could not be more different.

It sits up there on your ceiling, watching the air all day, every day for the very first hint of trouble.

Now, a spy also watches closely, constantly, but for its own purposes, not yours.

The moment you discover it's been watching, every ounce of trust you had just evaporates all at once.

The predictive AI that's now scanning your customer base for the earliest signs that someone's about to leave, that AI can be either one of two things.

Does your AI exist to serve the person or to surveil them? And if you get that wrong, the very system you built to reduce churn becomes the reason people leave.

Now, before you think this is soft, let me give you the economics, because retention might be the most leveraged number in your entire business.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.