Why niche beats reach: the math of high-intent audiences

Reach is easy to buy and easy to waste. A banner ad in front of a million people who scroll past it converts worse than a single host, mid-episode, telling three thousand people why they actually use something.

The difference is trust transfer. A listener who has spent forty episodes with a host has effectively vetted that person the way they'd vet a friend. When that host recommends something, the audience isn't evaluating an ad. They're taking a recommendation.

That trust doesn't scale the way media buys scale. It's built one episode at a time, and it can't be faked with a bigger number. This is why a 4,000-listener true crime show with a specific, obsessive audience can out-convert a 400,000-listener general interest show for the right sponsor: the smaller show's listeners are there for a reason, and the brand fits that reason.

The practical implication for a partnerships team: stop screening by download count first. Screen by fit first, then use size to decide budget allocation across a portfolio of fits. A brand that runs five tightly-matched $3,000 partnerships will typically outperform one $15,000 spray against a broad, loosely-related audience.

This is the entire premise behind how Casey builds a shortlist. Audience size is one input among several, weighted, not the first filter.