Is Podcasting Broken?

Over-saturation, discoverability, and star names

This blog is a response to Lucas Shaw’s piece for Bloomberg, ‘Podcasting Hasn’t Produced A New Hit in Years’. I’ll synthesise his thesis but I do recommend giving his piece a read first or as well.

It’s pretty clear when a piece of industrial analysis strikes a nerve, and this week podcasting experienced that. Lucas Shaw, a Bloomberg reporter specialising on Hollywood, wrote a newsletter with the snappy ‘Podcasting Hasn’t Produced A New Hit in Years’ and basically everyone I follow in our industry had to weigh in with their take on it. Here’s mine.

Shaw basically notes concerns within the industry, particularly from Spotify, about the performance of new shows, originals and exclusives. The problem can basically be boiled down to the following paragraph:

The number of new podcasts has grown more quickly than the podcast audience, and so the number of listeners per show is going down. The list of shows competing to be that program you try on your weekend walk is longer than the backlog of TV shows you want to watch.

Regular readers of this blog will appreciate that I broadly agree with Shaw’s analysis. Podcasting does have a discoverability issue, it does have an over-saturation issue. And the result is that audiences are homogenising around specific brand properties, and not finding new content. But I’d supplement that idea by noting that it has been true since very, very early on in podcasting. And the fact that podcast production growth is outstripping audience growth is also, I suspect, not a new trend. Podcast audiences have risen consistently in recent years, but the number of podcasts produced has risen exponentially. What I mean to say is: this is not a new problem in podcasting.

Interpreting the data

One problem I do have with Shaw’s analysis is his assertion that:

None of the 10 most popular podcasts in the U.S. last year debuted in the last couple years, according to Edison Research. They are an average of more than 7 years old, and three of the top five are more than a decade old.

The problem I have with this is that I don’t think an average is a useful way of looking at this…