Podcasting has emerged as a major billion-dollar industry, with advertising revenue in the US expected to reach $2 billion this year alone—a figure that will double by 2024. Against that backdrop, major players in the field are beefing up their podcast arsenal. With Spotify, it recently spent about $85 million on two companies focused on podcast measurement and analytics, while Acast recently announced that Podchaser – the “IMDb for podcasts” that provides advertisers with deep data insights – for a $27 million deal.
But as the big platforms lock horns in the podcast treasure hunt, smaller players continue to come into the scene with their own ideas for how to advance the podcast medium for creators and consumers alike.
One of them is Snipd, a Swiss startup building a podcast app that uses AI to transcribe content and note-taking apps; Auto-generate book-style “chapters”; And, starting this week, serve up podcast highlights in a TikTok-style personalized feed.
Search and subscribe
Like other so-called “podcatcher” apps, Snipd works by finding and subscribing to podcasts that users are interested in – this could be anything from true crime to history and sports. But Snipd is striving to be more than just another podcast in terms of how it analyzes the content of episodes to help listeners dig in and get to the heart of the details that matter.
For example, Snipd can create “chapters,” which divide each episode into sub-topics by its own title, as well as generate transcripts of entire episodes.
In addition, users can manually create “snips” while listening to an episode, allowing them to save their favorite moments and add notes to each clip.
With Snipd’s latest launch, available this week on Android and iOS, the company is channeling its inner Tiki Talk by offering a variety of highlights from what it believes are the most memorable moments from its many podcasts. It then assigns an AI-generated headline to each clip, and provides a feed that users can navigate by scrolling up and down.
From there, listeners can save each clip to their library, or — if you like what you hear from a short episode provided by Snipd — jump straight to a full podcast episode.
It’s worth noting that with the app’s latest update, users are now asked to select their favorite topics (for example, “Story” or “Music”), which Snipd uses to create these highlights. This means that the episode feed is not based. Only Because Snipd also pulls content from users’ podcast subscriptions He thinks Among other “signs” they will be interested based on their chosen topics.
Snipped founder Kevin Smith told TechCrunch, “The goal of the algorithm is to provide content that the user is looking for — and we use a variety of signals for that. “Whether a user is subscribed to a particular show or not is a strong signal, which is why a lot of the content seen comes from user subscriptions. But there are many other important indicators such as what the user has listened to, highlighted and saved or is currently being viewed among other users.
While this can be seen as a positive step by those looking for help finding new and useful podcasts, it can be especially annoying to users who want to watch content they’ve subscribed to. But Snipd eventually plans to give listeners more granular control over what appears in their highlights feed, including the ability to filter out unsubscribed clips from podcasts.
It’s worth noting that Snipd’s new feed is specifically focused on podcast episodes released in the past two weeks – in the future, it plans to follow a more similar approach to YouTube, which Snipd thinks is relevant and interesting in suggesting older content.
Away from the new TikTok-inspired highlights feed, Snipd users can get AI-driven highlights for every episode in their main subscription list, regardless of the episode’s current status.
The app automatically generates highlights for popular podcasts using criteria such as how many users have subscribed to a show. And for new or less popular podcasts, users can “ask” Snipd to work its magic, giving them highlights, chapters, transcripts, and all in around 20 minutes.
AI at work
But what exactly does Snipd look for when evaluating what content to present in its “highlights”? How does one know which classes are more qualified than others? According to Smith, it all boils down to how users have historically interacted with the segments — it analyzes which types of content are most interesting and then feeds this data into an AI-training method.
“Our AI learns by analyzing the content of older episodes, and comparing which episodes were most popular with our users and which episodes weren’t,” Smith said. “The most insightful episodes are frequently highlighted by our users, while the less interesting episodes are often skipped and not highlighted. Our AI has learned to use the actual content of the conversation to identify these episodes and can recommend them in new episodes.
Smith added that Snipd creates most of its AI models in-house, and for language models in particular, it starts with large pre-trained models like GPT-3, which can understand a lot about text and language.
“We then adapt these models to our specific use cases,” says Smith. “Other models, we develop completely from scratch.
We use feedback signals from users to improve the models over time.
Smith says that in the company’s early findings, users seem to use highlights to decide which episode they want to listen to — so they’ll flick through different clips until they find something that catches their eye, then skip to an entire episode. The problem, ultimately, is one of overload — just like how Netflix “recommends” new shows based on subscribers’ viewing habits and offers program previews on the main menu screen, Snipd is trying to help. listeners Filter by podcast audio.
“Our users sometimes have more than 100 shows recorded, especially the most data-rich ones, like Lex Friedman Podcast or the The Tim Ferriss Show” said Smith. “These classes are up to five hours long. This makes it very time-consuming for listeners to find the episodes they are most interested in.
Some studies suggest that 74% of listeners use podcasts to “learn something new,” 71% cite “having fun” as their primary purpose, and 51% for entertainment.
And that’s why Snipd’s self-proclaimed 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 of engaging with knowledge in podcasts and try to improve it. From finding the best content, consuming it, saving the knowledge the user wants to remember, to sharing it with friends.
Before the latest app update, Snipd mostly focused on letting users highlight and save specific knowledge they find, so they can revisit it later. As such, the app is compatible with headphones, so joggers (for example) can triple-click a button on their headphones to create and save a clip with an automatically generated title, summary, and transcript. And given how popular podcasts are among drivers, Snipd also recently rolled out support for Apple CarPlay, allowing users to stream podcast highlights while driving.
Snipd supports its “unlocking knowledge” mission in other ways as well. For example, if users want to read episodes or transcripts from their podcasts, they can integrate and sync Notion with Snipd, a reading and note-taking app. On top of that, users can manually export Snipd content to Obsidian, Logseq, Bear, and Markdown.
Show me the money
Based out of Zurich, Snipd will be formed by a five-member team, including five founders and two employees. The first iteration of the app launched last August, and in the intervening months the company has raised $700,000 in pre-seed funding from early-stage Swiss venture capital (VC) firm Wingman Ventures, as well as previously invested in billion-dollar companies like Square, Pinterest and Wish in the US. Acequia Capital, a VC based at Smith said Snipd plans to raise the seed at some point in the “not-too-distant future.”
All of this leads us to a very important question around finances – 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 to the There are other similar podcast apps, so this could mean a basic free version supported by ads or promotional content, with some fun AI-powered smarts pushed behind a paywall.
This 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 cool, but it’s unclear whether Snipd can gather enough of a user base to build a significant business. Additionally, there are similar companies out there, like Moonbeam, a podcast discovery app that uses machine learning and human editing to provide personalized podcast recommendations. There’s also Airr and Fathom.fm, which are similar in helping listeners get more out of their podcasts, either by aiding discovery or by allowing them to pull out the episodes that interest them the most.
In fact, Snipd may be acquiring or hiring in the process. Spotify, for example, already offers recordings for its original podcasts, and it’s not unusual for podcast-focused startups to spend millions of dollars. Amazon recently released podcast recordings.
In a crowded space, it’s clear that the major 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 that.
“We see podcasts as one of the largest knowledge-bases in the world, and so we focus on the knowledge-seeking community,” Smith said. “While our competitors treat podcasts as music they listen to from start to finish, we see them as a series of rich moments of knowledge.”