Podcasting has become a major multi-billion dollar industry, with ad revenue in the US alone expected to reach $2 billion this year — a figure that is set to double by 2024. Against this background, major players in the field are strengthening their podcasting arsenal, with Spotify recently gave about $85 million for two companies specializing in the measurement and analysis of podcasts, while Acast recently picked up Podchaser — “IMDb for Podcasts,” which gives advertisers deeper insights — in a $27 million deal.
But as the big platforms go in search of podcasting riches, smaller players continued to arrive at the location with their own ideas on how they can evolve the podcast environment for both creators and consumers.
One of them is Snipd, a Swiss startup creating a podcast app that uses AI to transcribe content and sync with note-taking apps; automatic generation of book-style “chapters”; and starting this week, provide podcast highlights in a custom TikTok-style feed.
Beyond Search and Subscribe
Like other so-called “podcatcher” apps, Snipd works by having users search for and subscribe to podcasts they’re interested in—this could be anything from true crime to history to sports. But Snipd aims to be much more than just another podcatcher, in terms of how it analyzes episode content to help listeners prepare and get to the heart of the important details.
For example, Snipd can create “chapters” that divide each episode into self-titled navigable segments, while also being able to generate transcripts of entire shows.
Additionally, users can manually create “snippets” while listening to an episode, allowing them to save their favorite moments and add notes to each clip.
With the latest launch of Snipd which is available at Android and iOS this week, the company is channeling its inner TikTok by presenting users with a sort of highlights bar, automatically pulling out what it deems the most memorable moments from multiple podcasts. It then assigns an AI-generated title to each clip and presents them in a feed that users can navigate by scrolling up and down.
From there, listeners can save each clip to their library or—if they like what they hear from the short segment presented by Snipd—jump directly to the full podcast episode.
It’s worth noting that with the app’s latest update, users are now asked to select their favorite topics (eg “story” or “music”) that Snipd uses to generate these highlights. This means that the episode feed is not based clean on users’ podcast subscriptions, as it also downloads content that Snipd thinks they will be interested based on their chosen topics, among other “signals”.
“The goal of the algorithm is to present the user with content that they’re interested in – for that we use different signals,” Snipd co-founder Kevin Smith explained to TechCrunch. “Whether a user is subscribed to a particular show is a strong signal, so much of the content shown comes from the user’s subscriptions. But there are many other important signals, such as what a user has listened to, highlighted and saved, or what is currently trending among other users.
While this may be interpreted as a positive move by those looking for help finding new and useful podcasts, it may annoy users who only want to see content they’ve specifically subscribed to. But Snipd eventually plans to give listeners more granular control over the content that appears in their highlights feed, including the ability to filter out clips from podcasts they haven’t specifically subscribed to.
It’s also worth highlighting that Snipd’s new feed focuses on newly released podcast episodes, particularly those released in the previous two weeks – in the future there are plans to take a more YouTube-like approach in terms of offering older content that Snipd believes is relevant and interesting.
Outside of the new TikTok-inspired highlights feed, Snipd users still have access to AI-driven highlights for every episode in their main subscription list, regardless of how recent the episode is.
The app automatically generates highlights for the more popular podcasts using criteria such as how many of its users are subscribed to a show. And for new or less popular podcasts, users can manually “tell” Snipd to do its magic so that it gives them highlights, chapters, transcripts and everything else within about 20 minutes.
AI at work
But what exactly is Snipd looking for when evaluating what content to feature in its “highlights”? How can it know which segments are more worthy than others? According to Smith, it’s all about how users have interacted with episodes historically – he analyzes what type of content generates the most interest and then feeds that data back into his AI training engine.
“Our AI learns by analyzing the content of old episodes and comparing which parts of those episodes are highlighted the most by our users and which parts are not,” Smith said. “The most insightful parts of an episode are often highlighted by our users, while the less interesting parts are often missed and not highlighted. Our AI has learned to use the actual content of the conversation to identify these parts and can recommend them in new episodes.”
Smith added that Snipd builds its AI models mostly in-house, and for language models specifically, it starts with large pre-trained models similar to GPT-3who can already understand a lot about text and language.
“We then refine these models according to our very specific use cases,” noted Smith. “Other models, we train completely from scratch.
We then use user feedback signals to improve the models over time.”
Smith said that in the company’s initial findings, users seem to use highlights to decide which episode they want to listen to — so they’ll go through different clips until they find something that grabs them, and then jump to the full episode. The problem is ultimately choice overload – similar to how Netflix “suggests” new shows to watch based on subscribers’ viewing habits and presents a preview of the show on the main menu screen, Snipd tries to help listeners filter out podcast noise.
“Our users are sometimes subscribed to over 100 shows, especially those that are very information-rich, like Lex Friedman Podcast or The Tim Ferriss Show” Smith said. “These episodes are up to five hours long. It takes a long time for listeners to find the parts that interest them the most.”
Some studies suggest that a full 74% of listeners tune in to podcasts to “learn something new,” compared to 71% who cite “having fun” as their main motive and 51% who cite relaxation.
That’s why Snipd’s mission is to “unlock the knowledge” in podcasts.
“The main problem we’re solving is getting knowledge from podcasts,” Smith explained. “We look at the entire user journey to engage with knowledge in podcasts and try to improve it. From discovering the best content, consuming it, saving the knowledge the user would like to remember, to sharing it with friends.”
Before the app’s latest update, Snipd was mostly focused on letting users highlight and save specific bits of knowledge they come across so they can review it later. As such, the app is compatible with headphones, so joggers (for example) can triple-click the button on their headphones to create and save a clip with an automatically generated title, summary, and transcript. And considering how popular podcasts are with driversSnipd too recently introduced support for Apple’s CarPlay, allowing users to generate podcast highlights while behind the wheel.
Snipd supports its mission of “unlocking knowledge” in other ways as well. For example, users can integrate and synchronize Snipd with the read-it-later service Readwise and a note-taking app concept if they want to read segments or transcripts of their podcasts. Additionally, users can manually export Snipd content to Obsidian, Logseq, Bear, and Markdown.
show me the money
Based in Zurich, Snipd is a team of five, including three co-founders and two employees. The first iteration of the app launched last August, and in the following months the company raised “oversubscribed” $700,000 pre-stage funding round from backers including early-stage Swiss venture capital (VC) firm Wingman Ventures, as well as US-based VC Acequia Capital, which has previously invested in billion-dollar companies such as e.g. Square, Pinterest and Wish. Smith said Snipd plans to raise a seed round at some point “in the not too distant future.”
All of this brings us to a rather important financial question – how exactly does Snipd make money? The short answer is that Snipd doesn’t make money… yet. But in the future, the company plans to adopt a freemium business model à la other similar podcast apps, so this could mean a basic free version supported by ads or promoted content, with some of the funky AI-powered smarts pushed behind a paywall.
It also raises questions about how easy it will be to thrive in a market that includes well-established (and well-funded) incumbents like Apple, Spotify, Acast and Pocket Casts. Snipd’s AI-powered features are certainly nice, but it’s unclear whether Snipd can gather enough of a user base to build a significant business. In addition, there are already comparable companies, such as Moonlightpodcast discovery app which combines machine learning and human processing to deliver personalized podcast recommendations. And there is also Airr and Fathom.fmwhich are similar in that they help listeners get more out of their podcasts, either by aiding discovery or by allowing them to pull out the parts they find most interesting.
In fact, Snipd could be an acquisition or hire acquisition in the making. Spotify e.g. now offers transcripts for his own original podcasts and is no stranger to give away millions of dollars for startups focused on podcasts. Amazon also recently released podcast transcripts.
In a busy space, it’s clear that the big podcast players will continue to look for new ways to add value and differentiate themselves from the competition, and helping their subscribers “unlock knowledge” could be another way to do just that.
“We view podcasting as one of the largest knowledge bases in the world, and therefore we are focused on the community of knowledge seekers,” said Smith. “Whereas our competitors treat podcasts like music you listen to from start to finish, we look at them as a series of knowledge-rich moments.”