All music streaming platforms claim to use both artificial intelligence and manual curation to discover new songs from emerging artists. However, users often have to listen to many songs just to find a track they like. This is because the user has no control over the recommendation algorithm. Romanian developer duo Alex Ruber and Andrei Patru have improved this process and created an app called Smores that makes it easy to add new music to your library.
Smores is a free iOS app that allows users to listen to short clips of songs based on their listening history. Users can skip tracks using vertical feeds like TikTok.
Image credit: S’mores
The app connects to your Spotify account and uses the Spotify API to discover new songs. If you enjoy a clip of the song, tap the like button and it will be added to a playlist called “Smores discovery” in your Spotify account. Alternatively, you can add songs to one or more existing playlists.
The developers told TechCrunch in an email they set themselves up to build a new music discovery app. So last September he released the first version of Smores.
“We love discovering new music, but we got stuck in the recommendation bubble and spent too much time sifting through the sheer volume of new music that was coming out. I had a hunch that just by listening to the “correct” snippet, I could tell if I liked it. Shazam’s popularity shows that this is the case,” they said.
The duo said they wanted more control over their discovery algorithms and built transparency into the app. To that end, Smores includes a number of built-in controls for modifying a user’s recommendation feed. Users can filter out suggestions based on the top 6 microgenres of the month. These change as you listen to more music in the app and like more songs.
Image credit: S’mores
You can define the snippet length (from 5 to 60 seconds) in your app’s advanced settings. Limit discoveries based on the number of followers an artist has on Spotify. Filter songs by BPM (beats per minute), song key, and release date.
Image credit: S’mores
One of the great things about this app is that you never listen to the same song twice. Additionally, the developers say they have tweaked the algorithm to figure out the “best” part of the song to play with the snippet. They say that when the app plays a good preview, many users tend to love the song in as little as five seconds.
User Retention and Future Plans
Music discovery apps are fun to use, but building an audience that uses them on a regular basis is difficult. Despite this challenge, the developer has managed to retain a good number of users (at week 8 he had 7%) and says he hears positive feedback from regular users.
“It’s true that music discovery apps in general have a low cadence. Casual listeners actively discover new music maybe once every three months. They rave about how much they love the ease, speed, convenience and quality of their recommendations,” they said.
The team is currently focused on building features like Smores Radio and integrating Apple Music and other streaming platforms. In the future, we would like to introduce an Android version and possibly a premium tier, but we have not disclosed any paid features yet.
Increase in AI in music
Music fans have often complained about the growing role of AI in music discovery and distribution. Nonetheless, businesses and app developers are relying more and more on AI, using it for more control over their algorithms with buttons and filters.
Bytedance’s music app Resso (currently only available in India, Brazil and Indonesia) leverages vertical feeds and the company’s proven AI capabilities to help casual listeners discover new artists. The Chinese tech giant also aims to launch TikTok Music globally. AI-powered music suggestions could play an important role in this service.
App developers are also bringing features to their music apps with the help of AI. LineupSupply, an app that converts festival posters into playlists, has changed its name to Playlist AI. The app also introduced a new feature that allows you to write prompts such as ‘What dance artists were popular in the 1990s’ to generate playlists.