//

Today at the Gramercy Theater in NYC, Spotify’s Chief R&D Officer Gustav Söderström announced a brand new free version of the Spotify mobile app.

By leveraging their investment in machine learning, Spotify’s new free tier recommends music to users on the fly. That said, the free tier has always limited users to shuffle. With the new version, users can listen on-demand to whatever song they want, as many times as they want, as long as those songs appear on one of the 15 personalized discovery playlists.

In total, that’s around 750 tracks (>40 hours of music) that Spotify is serving up to users for on-demand listening.

Spotify will also make recommendations in the free mobile version based on existing user-made playlists, from the songs on those playlists to the name of the playlist itself.

Finally, Spotify has built in a low-data mode that cuts data consumption by up to 75 percent. In the past, Spotify didn’t allow offline listening for free, meaning that users were somewhat tethered to wifi if they needed to conserve data.

With the new data consumption system (it’s unclear how this will work), users can actually listen to much more music with wireless data.

In 2014, Spotify introduced its free tier to mobile, letting users listen to their playlists on shuffle with ads. It was a huge part of Spotify’s free-tier growth. In fact, today Spotify has 90 million users on the free tier. And many of those convert to paid users — the company now has 70 million paid subscribers.

The company has focused heavily on mobile since 2014, especially where it concerns the premium mobile player.

Spotify is built upon three tiers: ubiquity, personalization and freemium. Söderström explained that Spotify thinks of itself as the broadcast radio of the 90s, where discovery of great music was supported by ads and drove people to the record stores.

Spotify’s free tier represents broadcast radio for Spotify, and is a critical piece of Spotify’s overall strategy as paid services like Apple Music continue to grow.

from Startups – TechCrunch https://ift.tt/2Hrhq3g