Video: Spotify explains how it uses machine learning and humans for curation
Even with the backing of major labels it enjoys, Spotify[1] is no stranger to skepticism.
When the service launched, many thought that it would struggle to attract an audience in the wake of widespread piracy, Apple's dominance of song sales, and a poor track record of streaming music services including Rhapsody and Zune Pass.
But at the New York City launch[2] of its revamped app and enhanced free service last week[3], Spotify spoke about how it had defied critics as a backdrop to an expansion of its free service. Still comfortably ahead of Apple, which has been pushing Apple Music hard as a competitor, Spotify[4] continued with enhancements to its free service, which once wasn't available on mobile devices at all and had been limited to shuffled playlists. Now, though, the company is making available up to 40 hours of on-demand music from 15 playlists per week.
The company noted that the key to making a good free service is not to punish users who don't pony up for the paid service even if there runs an increasing risk of cannibalization. (It's hard to argue that it was not punishing free users when it didn't allow mobile access, but perhaps its thinking has evolved.)
(Image: Spotify)
Its pitch to artists and labels is that music fans matter even if they can't afford a subscription, at least when they are in the younger age ranges that over-index on the company's free service. Spotify can monetize these users now through advertising -- advertising that can be made more precise, as these users use Spotify more. Free users will use