Spotify Now Hosts 70 Million Songs. But It Can’t Keep That Up Forever
Music streaming services have a new favorite number. In recent job ads, Apple Music is telling would-be candidates: “We started with 1,000 songs in your pocket. Now it’s over 70 million.” Last week, YouTube’s Global Head of Music Lyor Cohen announced within a slew of new stats about music consumption: “YouTube Music has more than 70 million official tracks, more than any other music service.”
And in early November, Spotify CFO Paul Vogel told his firm’s investors that the streaming service now has “65 million to 70 million music tracks” in its library — up by 15 to 20 million on the “50 million” estimate that Spotify reported at the close of 2019. This eruption in track volume on the likes of Spotify and its rivals can largely be credited to one segment of the music industry: the DIY, a.k.a independent, artist market.
This comes at a particularly interesting juncture for Daniel Ek and co.
On November 2nd, Spotify announced “Discovery Mode,” a tool that enables labels and artists to gain influence over algorithmic recommendations in exchange for agreeing to a lower royalty rate. The news kicked up fierce debate over whether “Discovery Mode” actually constitutes a form of streaming payola. But “Discovery Mode” also denotes an important new era for the loss-making Spotify, suggesting that the company is under pressure — and willing — to deploy swift margin-improving tricks to convince Wall Street that it’s serious about finding a path profitability. The consequence of failing to convince Wall Street of this seriousness goes beyond mere share price: History tells us Daniel Ek’s CEO standing could become a target should his company stumble in its quest.
Through this lens, the explosion in DIY music is becoming a potential headache for Spotify — because there’s an exponential cost associated with housing 10-plus million more tracks every year. Spotify’s financials don’t make finding this specific expense easy, but in the first nine months of 2020, the firm admits it saw “an increase in information technology costs of €16 million due to an increase in our usage of cloud computing services and additional software license fees.” Spotify uses Google Cloud for its data hosting purposes.
Hosting this ocean of DIY artist tracks isn’t only an expense for Spotify: It’s also an unexplored revenue opportunity. It can’t be long before Wall Street analysts twig that Spotify is allowing the vast majority of revenue associated with the DIY music sector to live entirely outside of its walls. How lucrative is that DIY sector? One only need look to Believe, the parent of indie distribution giant TuneCore, which is reportedly plotting a $2 billion-plus IPO next year.
What, then, should Spotify do in the face of the cost/opportunity conundrum presented by DIY music? To my mind, it has three options.
1. Launch a DIY upload service (again) — but this time, charge for it
In September 2018, Spotify announced it would allow independent artists to upload tracks to its service for free; the following month, it went one better, acquiring a minority stake in TuneCore rival Distrokid and announcing that it would soon let artists upload their tracks not only to Spotify, but rival platforms too. And then, just nine months later, Spotify abruptly closed its upload-and-distribution experiment, going out with a whimper. “The most impactful way we can improve the experience of delivering music to Spotify for as many artists and labels as possible is to lean into the great work our distribution partners are already doing to serve the artist community”, it said.
Profit-wise, Spotify’s big mistake here was offering free uploads in the first place. What would have happened had the platform charged indie artists for distribution, but justified this cost by making additional paid services available on top?
In doing so, it would have emulated its rival, SoundCloud, which last summer — just as Spotify was closing down its DIY distribution operation — acquired distribution and services company Repost Network for $10 million in cash. Today, the basic package of Repost on SoundCloud charges indie artists $30/year, offering digital distribution plus services like playlist pitching and promotional tools. Considering that SoundCloud currently hosts 250 million (!) tracks, the platform’s business model can only make sense if its most prolific uploading artists are paying their way.
And SoundCloud’s business model is starting to make sense: having generated $54 million from its “Creator Business” segment (including Repost) last year, SoundCloud just told Music Business Worldwide that it saw its first-ever EBITDA profitable quarter in Q3 2020.
2. Pull up the quality drawbridge
Back when Pandora was a digital radio network, rather than a Spotify-like interactive platform, it didn’t just accept any old artist uploading via a DIY distributor: Musicians had to be approved by Pandora’s own quality control regimen.
This was the bane of many indie artist’s lives. But perhaps Pandora was ahead of its time. Looking at SoundCloud’s ridiculously large track database (again, 250 million!) suggests that, at some point, streaming services may have to consider deploying a filter system to sift the sublime songs from the shlock. This — the power of scarcity, and the embrace of the role of gatekeeper — could become a winning marketing angle for Spotify in a sea of music streaming sameness.
That said, it might also be a major risk.
Over in China, a market so often ahead of the curve on streaming music trends (tipping, online karaoke etc.), independent artists have never been treated with more respect than they are today — and for good reason. Tencent Music is giving the best DIY acts “financial incentives” to put their music exclusively on its platform. Clearly, Tencent sees this as a potential route to market dominance that doesn’t upset anti-competitive watchdogs. Spotify may not want to pass up a similar opportunity.
3. Roll out fees for hosting the music itself
One route to Spotify gaining a financial slice of the multi-billion-dollar DIY artist market — without re-visiting the idea of becoming a distributor itself — may lie in its relationship with Google. As mentioned, Spotify has to pay Google Cloud significant hosting fees each year for holding its ever-expanding music library. Late last year, Spotify ran the largest Google Dataflow job ever for its Wrapped 2019 campaign.
Presumably, Spotify is paying Google millions more dollars every year because its own music library keeps expanding – and keeps expanding, in the main, because of independent artists. So why doesn’t Spotify pass on the cost of hosting this data to independent acts and their distributors? Could it make song-hosting a service (just like “Discovery Mode”) for which artists are expected to pay?
The answer to that question, naturally, will be rooted in customer satisfaction. But while Apple, Google-owned YouTube, and Amazon are profitable enough in other arenas that they probably won’t ever have to charge for song-hosting, Spotify — with its lack of other businesses to bolster — might see it as the only option left on the table.
Tim Ingham is the founder and publisher of Music Business Worldwide, which has serviced the global industry with news, analysis, and jobs since 2015. He writes a weekly column for Rolling Stone.