Writing

Seven Months Independent: The Audience Is Compounding

Just over seven months ago, The Chain of Thought Podcast became fully independent. Same show, same guests, now running on its own channels and its own audience. That created real challenges - we had to start from scratch on YouTube just as the podcast was catching on via Galileo’s YouTube. But it also created opportunities: in focus, in exploration and in imagination. So what does that look like in our numbers?

The short version: organic demand is compounding, on both YouTube and the podcast feed.

Let’s look at our YouTube first. A podcast feed auto-delivers every episode to every subscriber, so reach is guaranteed. YouTube guarantees you nothing. It reruns its discovery test on every upload and decides, each time, whether to show your video to anyone at all. You largely don’t inherit reach for being subscribed to (unless subscribers turn on notifications); you earn it episode by episode. That makes organic YouTube traction one of the cleanest signals of demand there is, and earning it is most of what the last seven months have been about: titles that win the click from the small slice the algorithm shows you, and a concentrated launch in the first hour to pass its discovery test.

There have been some learning moments, but overall, it’s working!

YouTube Studio analytics for the Jerry Liu episode: 15.2K views, 1.8K watch-time hours, and +470 subscribers in its first week, with 89.5% of views coming from YouTube recommendations.

The Jerry Liu episode’s first 5 days on YouTube.

My newest episode with Jerry Liu, “The AI Framework Era Is Over”, crossed 15,200 views in under a week, 89.5% of it from YouTube’s own recommendation engine via Browse and Suggested. Not ads, not me posting links. That one video added nearly 500 YouTube subscribers in four days, with the click-through rate climbing as it scaled - it’s on track to be our most-watched episode yet. And Jerry isn’t alone. The AMD episode with Anush Elangovan is near 6,000 views, an episode on building an AI coworker with Sterling Chin is past 3,600, both pulled by recommendations. We’ve established a pattern of organic breakouts from the trendline.

Recent guests

The podcast feed tells a quieter version of the same story, but I’m particularly proud of the progress we’ve made there. Because every subscriber auto-downloads, the floor rises on its own as the base grows. Our release-window downloads are up about 15% over the last few months. That engine is healthy and self-reinforcing. It just doesn’t make for as flashy a chart.

A little more on the numbers:

  • 61 published episodes
  • Over 150,000 downloads and views total
  • Lifetime average north of 2,500 an episode that keeps climbing
  • Our new YouTube channel has gone from 0 to 2,800+ subscribers in 5 months
  • Top episodes are now clearing 10,000 downloads/views combined

Now I need to solve the floor. What I want is for every episode to reach a real audience, not just the ones the algorithm happens to love that week. Closing the gap between the breakout and the median is the whole project for the back half of this year.

The audience itself is the part I’d put up against any show in the category. About two-thirds of it is 35 to 54 years old, senior operators and decision-makers. For a B2B AI show, that’s exactly the audience we want to reach.

Chain of Thought audience by age on YouTube over the last 28 days, concentrated in the 25-54 brackets.

Audience by age on YouTube, last 28 days. The 35–54 bracket drives roughly two-thirds of watch time.

So that’s where seven months of Chain of Thought independence leaves me: organic demand compounding on YouTube, a download floor that climbs on its own, repeat breakouts proving the format travels, and a clear target for what comes next.

Come join us! Subscribe here, or reach out to sponsor the show.