Back in 2003, as the music industry was reeling from widespread, Napster-fueled piracy, Apple CEO Steve Jobs made the record labels an offer they couldn’t resist: Give me a license to sell individual tracks but let me sell them cheap enough to be a viable alternative to free, and I’ll wrap them in DRM for you in a way that consumers will accept, so they can’t be copied.
The labels leapt at the deal and the $1.00 download became the new atomic unit of the business.
Though thrilled at first to have an answer to piracy the record companies eventually came to rue the arrangement once they figured out that Apple was using those inexpensive downloads to supercharge the market for its high-margin iPods and later iPhone hardware, and was reaping far more of the value being created by their music than they were. By then, however, they had become captive to Apple’s ecosystem: Thanks to Apple’s proprietary DRM, the only way to sell music to iPod users — at the time the largest segment of the portable music-player install base — was through iTunes, under terms effectively dictated by Apple.
It took years of tough negotiating, along with the abandonment of DRM, for the labels to wriggle out of Apple’s grip.
By now, of course, streaming has largely supplanted downloads as the main revenue engine of the business, making DRM largely irrelevant, and increasing the labels’ leverage over downstream service providers. But there’s still a danger of becoming captive to an ecosystem they don’t control and losing leverage in the long term.
The labels made two fundamental mistakes in their early dealings with iTunes. First, they let their fixation with fighting piracy blind them to the foreseeable evolution of the market and to make corresponding adjustments to their own business model. Second, they allowed Apple to define and own the dominant DRM, leaving them beholden to Apple to reach its millions of users.
Today, DRM no longer determines how consumers access and listen to music. But other technological factors are beginning to exert a similar influence on consumer behavior. And nowhere is that more apparent than among Spotify users.
With more than 100 million users, including 50 million paying subscribers, Spotify is the largest, if not yet dominant, music streaming service. The $3.9 billion Spotify turned over to the record labels and music publishers in 2016 represented 23 percent of total industry revenue from recorded music.
While Spotify has viable competitors, in Apple, Amazon, Google, and others, it is developing a hold on its users comparable to Apple’s former grip on iTunes users thanks to its increasingly sophisticated recommendation engine, exemplified by its Discover Weekly and Release Radar features, which increasingly determine what Spotify users listen to, and by extension what music they’re exposed to.
In her most recent Internet Trends annual report, Kleiner Perkins analyst Mary Meeker noted that Spotify users listened to an average of 41 different artists per week in January, more than for any other streaming service, which she attributed to its algorithmic recommendation engine.
The popularity of Discover Weekly has surprised even Spotify executives. As senior product manager Matt Ogle told Adweek last year, its success has “completely changed” how the service thinks about the user experience.
“It’s created this Monday morning routine that was certainly never intended to the degree with which it’s taken hold of people—which is something we’re really inspired by and can learn from,” Ogle said.
Since then, Spotify has gone on an acquisition spree aimed mostly at further bolstering its analytics capability. In March it acquired UK startup Songalytic, largely for its audio recognition software, followed quickly by MightyTV to bolster its programmatic advertising capabilities. The following month it acquired metadata service Mediachain, and in May it bought up machine-learning developer Niland.
Last week Spotify hired acquisition specialist Sheila Spence, an indication that its buying spree is far from over, and that it intends to continue investing in its data crunching capabilities.
As noted, Spotify faces stiff competition from Apple, Amazon and Google, none of which are slouches when it comes to data and analytics. All can be expected to follow Spotify’s lead in refining their curation and recommendation systems.
The danger for the labels is that the streaming services will become the gatekeepers that set the terms of success in the marketplace, leaving the tail once again wagging the dog.
In their recent dealings with Spotify, however, the labels seemed once again to be focused on protecting their near-term downside rather than on aligning their long-term interests with Spotify’s. Universal Music’s purported “breakthrough” in its recent licensing deal with Spotify, for instance, was gaining the ability to window certain releases in order to protect initial sales of the album.
In the long run, though, figuring out how to determine what Spotify users will be listening to is likely to prove more valuable than determining when they can listen to it.