spotify220

In a move to make podcast discovery easier – and encourage more listening to a content type it is spending hundreds of millions of dollars on building – Spotify has launched a new personalized podcast playlist feature. The feature will display a list of recommended shows based on what a user has listened to during the prior three months.

In an announcement Tuesday, Spotify said Your Daily Podcasts will be created using algorithms that analyze a user’s podcast behavior. That includes not only the specific shows they’ve consumed but also the genre of content that they’ve listened to. The algorithm will also look for content from a host the user has listened to and shows of a similar length and format. The new playlist will appear in the “Your Top Podcasts” section of the app’s home screen or in the “Made For You” hub on browsers. It will only be offered to users who have played at least four podcasts during the prior 90 days. The company didn’t say whether shows that are exclusives or produced by Spotify or its wholly-owned Gimlet Media or Parcast would get preferred placement among the recommendations.

Your Daily Podcasts will initially roll out in nine countries, including the U.S., U.K., Germany, Sweden, Mexico, Brazil, Canada, Australia and New Zealand.

In a conference call with investors last month, CEO Daniel Ek said the company continues to see podcast listeners spending more time with the platform, including listening to more music on Spotify than non-podcast listeners. They’re also more likely to become subscribers. “We’re not yet focusing on monetization of podcasts,” said Ek. “The primary thing for us has been to drive up engagement.”

Using artificial intelligence to make podcast recommendations isn’t a new idea, and many apps have already been harnessing user behavior to suggest podcasts. In February iHeartMedia announced it will use thousands of data points to power a new podcast recommendation feature on its iHeartRadio app. And Pandora has created what it has billed as the Podcast Genome Project to help listeners dig through that catalog. It also relies on more than 1,500 attributes such as production style, content type, host profile, timely and evergreen topics, and listener signals including thumbs, skips and replays, to deliver personalized podcast recommendations.

Meanwhile, there’s a new competitor emerging for Spotify and other streaming music services. Amazon Music has expanded to now include an ad-supported option, allowing customers in the U.S., U.K., and Germany to access music without a subscription or Amazon Prime membership. The service is available on Amazon Music’s mobile apps, Fire TV and via its website. The company says it has playlists and thousands of stations, but so far no podcasts.

Amazon still aims to lure users to its subscription service which it says has a catalog of 50 million songs and is more directly comparable to what Spotify offers. Nevertheless, the threat of a well-funded competitor like Amazon expanding its reach into the music market led to a selloff of Spotify’s stock on Monday after the announcement, and sank further during early trading on Tuesday.