There has long been a clear separation between the two kinds of playlists on Spotify: the human-curated ones like RapCaviar and New Music Friday on one side, and the algo-personalised ones like Discover Weekly and Release Radar on the other. Humans and algorithms working separately, with the former producing a large, multi-tiered network of playlists covering all manner of genres, moods and contexts, and the latter trained on a smaller selection: those above plus the Fresh Finds and Daily Mix lists, and one-off seasonal playlists like Summer Rewind.
Here’s some interesting news, then: Spotify may be testing letting its recommendation algorithms loose within its human-curated playlists too. Well, some of them. Hypebot reported that “Spotify is testing the addition of individually personalised tracks to existing editor curated playlists… In one example, tracks which Spotify’s believes would fit the taste of a particular user were added to an official Spotify hand curated workout playlist”. Billboard named four of those included in the test: Beast Mode, Chill Hits, Dance Party and Metal Ballads which have 4.3 million, 3.5 million, 2.3 million and 121k followers on Spotify respectively.
The emphasis of this test – obligatory “We are always testing new ways to create better listening experiences for more users, while also finding ways for users to discover more music” spokesperson quote included – is thus on mid-tier playlists rather than the real big-hitters of Spotify’s playlists ecosystem. It does make sense around mood and activity-focused playlists: rather than make a Release Radar-sized splash with a weekly, branded, personalised workout playlist, for example, why not just apply the personalisation algorithms to a range of them?
Some labels may worry at this news, especially those that have put more energy in to pitching tracks to these mid-tier playlists – a strategy that can pay off, since some of them are played regularly by followers, with songs sticking around for much longer than on the bigger flagship playlists. Then again, algo-personalising the mid-tier playlists could open them up to a wider range of tracks by more artists who might not have found their way to the curator otherwise, so there may be an overall benefit. In any case, this is just a test for now: the response (in terms of listening) will dictate whether it expands.
Tools: platforms to help you reach new audiences
Tools: Kaiber
In the year or so since its launch, AI startup Kaiber has been making waves,…