There is so much gold in podcasts—interviews with primary sources, super-well-researched and vetted topics, and hosts of experts sharing their knowledge.
The problem is, they are very hard to find. You probably don’t know it, because when you search for something, you always get a result. But the truth is, for every podcast episode you find when searching, there are 100 more that don’t show up.
The answer to this problem is simple.
Podcasts, at their most primitive level, are .mp3 files. They are no different than the file format a Beatles tune, radio commercial, or exported voicemail are saved in. And while there are players to play them and voice-recognition programs to transcribe them, there is no search engine that can read them when they are not playing.
When podcasters record shows, they save them as .mp3 files and upload them to their podcast host. Since that file can’t truly be read, the search engines really only have the title, image file, subtitle, description, and tags from which to index the episode. (On occasion, they have the transcription, as well.)
So, if a podcaster interviews Neil DeGrasse Tyson, but decides to name the episode “Talk Shop with the Nation’s Top Space Doc” and makes the podcast description about the discussion itself without mentioning Neil, the chances of it being found on an Apple Podcasts search is very slim.
That “problem” has opened the door to a variety of third-party search engines with better ideas.
Audio Junkie, Listen Notes, Ivy, Mixtape, and OmniSearch.ai are all vying to be your podcast search engine of choice.
Listen Notes is probably the biggest and most well-known podcast search engine, and for good reason. Wenbin Fang, its creator, built something amazing. Launched in 2017, Listen Notes now boasts 1.4 million users each month, and with that number comes some amazing ways to help you find the topic you’re looking for.
Wenbin realized that technology is not the only way to read what a podcast is about—he’s got 1.4 million users to do it for him. Listen Notes gives its users the ability to create playlists for themselves, which sounds pretty normal, right? But from a search standpoint, when a user puts a particular podcast episode into a “Current Events in Afghanistan” playlist, a search engine can then see that intent and index those podcasts as such.
Listen Notes also takes note of what its 1.4 million users are listening to right now, adding relevance to podcasts with really obscure titles and descriptions.
Wenbin also found power in his employees and their recommendations, which has resulted in another source of indexed shows.
All in all, Listen Notes leverages eight or nine different ways to index podcasts for its search functions.
Audio Junkie found themselves with a similar asset—users. Audio Junkie collects user reviews and gives reviewers a chance to tag shows with their most relevant hashtags. That category cloud of options exists on their homepage and is an easy way to search for the topic you’re interested in most. And you don’t have to type into the search bar… clicking around on a list of categories still allows you to search for something good.
Ivy.fm is another unique search engine. Ivy’s algorithm turns every episode into a series of tags (that the podcaster is able to edit). It takes the episode title and description and uses them to pull out searchable tags, and its algorithm adds tags that seem relevant. For instance, if an episode claims to be about the Cuban Missile Crisis, Ivy’s algorithm will add JFK as a tag. That means it is not totally relying on the few words the podcaster writes about the episode.
Mixtape and OmniSearch are the most ambitious of the search engines. With a $125K grant from the European Union, Mixtape set about to create a process that listens to every podcast uploaded, automatically transcribing and indexing the content, and then using that to aid its search feature. OmniSearch thought the technology could be used internally by companies like Dropbox to make every file, regardless of type, searchable.
As goes the world of start-ups and social media companies, Apple and Google are very likely working on implementing this technology or buying one of these companies as we speak. Fortunately, in the meantime, these companies are filling that hole nicely.
October 2021 Issue