Podcast Discovery

Somewhere around five or six years ago Jeremy Cherfas invested a good deal of time to start a podcast called Pick of the Podcasts. This was a show that had sections from as many as four shows featured in order to encourage people to subscribe to those shows. The effort was incredibly time consuming and resulted in the podcast owners not really having a way to measure how many people had listened to their show, as the episode audio was not being hosted from their site but a separate file. Pick of the Podcasts had a very short run but, as a mechanism for podcast discovery, it seemed a rather interesting way to approach the problem of "podcast discovery"; finding shows that we might want to subscribe to.

In the half-decade that has passed since this attempt to help people find new shows, not much has changed to help people find new shows to listen to. The directories that exist all seem to promote the big networks, such as PRX, NPR, BBC, Gimlet, NHK, and the like, while simultaneously burying the independent creator who could very well be making something truly unique that we might enjoy. There should be a way to give everyone an equally weighted score when trying to suggest productions to people.

Oddly enough, while doing the dishes tonight, I started tossing around the idea of creating a system that would track podcasts published in the last week or so, look for present and historical keywords weighted by frequency, then generate a custom RSS feed that I could subscribe to with any podcast player. This feed would contain 4 new episodes per week for shows that might be of interest based on various factors provided by the RSS subscriber.

Would this work, though?

Discovery has long been a problem with creative endeavours as there has been an absolute plethora of content online since the Internet's inception. Jeremy is right when he strongly states that podcast discovery is just as much a problem today as it was in 2015. There must be a solution to the problem, and I wonder if something like this would bring us one step closer to an ideal solution.