For most of its history, podcasting has optimized for one number: downloads.
Downloads drive the charts (and god, everyone loves a good charting podcast). Downloads drive ad pricing. Downloads drive the way the industry talks about success. If a show hits a million downloads, that’s the headline (and the accompanying press release). If it hits ten million, it’s a big hit.
But the dirty secret is that downloads are an imperfect metric. They tell you how many people passed through. They don’t tell you how deeply anyone stayed, how much they cared, whether they engaged.
And that’s where the whole thing starts to get a little strange.
Because if you actually look at listener behavior, podcasting might be one of the most engaged media formats on the internet. Roughly seventy percent of listeners finish most or all of an episode. Nearly half start new episodes within twenty-four hours of release. People routinely spend hours a week with the same hosts, often while doing something else but still listening closely enough to follow long conversations.
That level of habitual consumption is rare in digital media. And yet the industry has largely built itself around a metric that barely captures it.
The Engagement Paradox
Podcasting has clearly achieved scale. Today about seventy-three percent of Americans have consumed a podcast, and more than half listen monthly. Total listening time has exploded over the past decade, growing more than three hundred percent since 2015. By any reach-based measure, the medium has already won (go us!).
Part of the reason is structural. Podcasting grew up in a decentralized RSS ecosystem where downloads were the most visible signal anyone could measure. But reach doesn’t tell you much about the underlying relationship between a creator and an audience. Engagement does. And that difference matters a lot.
And that’s where the broader media landscape starts to look very different.
Look at the platforms that now dominate creator media. YouTube. TikTok. Instagram. None of them primarily optimize for reach. They optimize for engagement behavior: watch time, completion rate, return frequency, session length, shares.
Engagement powers the algorithm. The algorithm produces reach. It’s a knock-on effect (an important knock-on effect but a knock-on effect regardless).
The Internet Runs on Engagement
Podcasting evolved in the opposite direction. The infrastructure of the industry grew up around downloads, charts, and audience size. Those numbers are easy to measure and easy to sell.
God knows how many emails (or LinkedIn posts) I sent while at Paramount touting a #1 podcast. Those weren’t wrong. Internally those numbers mattered, externally they mattered even more. But they’re also a pretty blunt instrument for understanding audience value.
Because I’d argue that right now, engagement is what actually unlocks the full value of the business. Deep engagement creates community. It creates live audiences. It creates merchandise sales. It creates subscription businesses. It creates intellectual property that can travel into other formats. When someone spends five or ten hours a week with a creator, the relationship starts to look less like media consumption and more like affiliation.
That depth is where most creator businesses are actually built.
The Distribution Model Is Changing
And this is where the underlying dynamics may be starting to change. One of the biggest shifts in podcasting right now is where listening is happening. YouTube is now the most used platform among weekly podcast listeners, surpassing both Spotify and Apple. That matters less because of video and more because of how YouTube distributes content.
YouTube doesn’t care how many subscribers you have. It cares how long people watch. If people stay, the system pushes the show further. If they leave, the reach collapses. In other words, the platform distributes shows based on engagement first.
The algorithm is not a human editorial voice. It doesn’t care. It’s pure data, and creators increasingly have to approach it that way. And when engagement works, reach tends to follow.
The Next Podcast Metric
Podcasting may not have been wrong to focus on downloads. They were the most visible signal available in a decentralized RSS ecosystem. But as podcasts migrate into algorithmic environments, the underlying metric stack is changing.
It’s the same engagement behavior we mentioned earlier: Shares. Watch time. Retention. Completion rate. Return behavior. Metrics that capture depth instead of just scale. The strange possibility is that podcasting may have spent twenty years optimizing for the metric that was easiest to count rather than the one that actually drives media businesses.
If engagement becomes the primary signal, the next era of podcasting may look very different from the first.