Spotify’s New Privacy Policy: Music’s Not the Product, You Are


In June 2013, Flavorwire ran a piece about Spotify, written in the wake of Thom Yorke’s condemnation of the service for not paying enough money to up-and-coming artists. At the time, we noted that the service was losing money, and that its business model appeared unsustainable, and wondered how it was planning to survive long-term. Two years later, Spotify is still losing money, and its business model appears to remain unsustainable, but perhaps the answer to the question of ensuring its long-term profitability is becoming clear — or, at least, the company’s strategy for doing so is becoming clear. Whether said strategy will work seems questionable, at best.

Last week, we pondered on who Amazon is actually for. The same question could be asked about Spotify: who is it for? Not musicians, obviously, unless getting paid a fraction of a cent every time someone plays your song seems like your idea of a good time. Not Spotify’s private owners, presumably, since the company continues to lose money. So, who?

The obvious answer in this case is: consumers! The idea of paying $10 a month for a magical version of iTunes (Wired‘s words, not mine) seemed almost too good to be true when it came along — no more buying albums! No more paying $1.69 a song from iTunes! Ten dollars a month, and all the music in the world is yours! As with many things that are too good to be true, though, Spotify is proving… well, too good to be true. For a start, it proved that the magic jukebox didn’t have all the songs — it had a lot, sure, but from the start it was missing certain prominent baby boomer artists, and as time has passed, more and more musicians have decided that they’re better off trying to sell albums the old-fashioned way than essentially giving them away on Spotify. This means that the service’s main selling point is undermined — as we observed two years ago, “every [artist] that pulls [their] work from Spotify undermines the service’s biggest asset: its catalog.”

At the time, we wondered exactly how Spotify was going to survive if it couldn’t “make things work paying artists these pissant royalties — and the consensus seems to be that it can’t.” So it has proven. If you’ll allow me to toot my own horn for a minute, I also suggested that the company needed to explore alternative revenue streams. Spotify has clearly been doing so, although sadly they have failed to take on board our frankly genius Netflix-but-for-music idea.

No, Spotify’s idea is this: monetize its users. Not by increasing the subscription fee — the $10 monthly fee is too entrenched now, and to up the fee as the service’s catalog shrinks wouldn’t exactly be great PR. Instead, Spotify is taking a leaf out of Facebook’s, ahem, book, and banking on the fact that its users’ data is its greatest asset. As Forbes’ Thomas Fox-Brewster reported last week, an update to Spotify’s privacy policy suggested that the company was now within its rights — with the user’s permission — to collect photos, contacts, GPS information, voice commands, and more.

The Internet was understandably less than impressed with this news, and once Forbes’ report started to gain momentum as it was aggregated around the world, Spotify retreated from its aggressive data collection policy. Founder and CEO Daniel Ek published a blog post entitled “SORRY.” wherein he set out to explain why Spotify wanted all this data: for “specific purposes that will allow you to customize your Spotify experience.”

This seems… well, at best it seems like a convenient explanation. It’s a tech cliché these days to observe that if you’re not paying for a product, you are the product — but the same aphorism is apparently increasingly true in cases where you are paying for the product. Perhaps this was the plan all along. Paying out negligible royalties to artists is one thing, but when you combine it with huge licensing fees to record labels, it’s always going to be extremely difficult to turn a profit, especially if the labels you’re dealing with would rather that the ground just swallowed you up and you were never heard from again.

The key part of the privacy update, which Ek didn’t address in his blog, is this:

5.2.1 Marketing and advertising We may share information with advertising partners in order to send you promotional communications about Spotify or to show you more tailored content, including relevant advertising for products and services that may be of interest to you, and to understand how users interact with advertisements. The information we share is in a de-identified format (for example, through the use of hashing) that does not personally identify you.

At first glance, the last sentence seems to put the casual user’s mind to rest: there’s nothing here saying that Spotify can notice that you’re playing a million Nickelback songs and bombard you with ads for their new album, or perhaps for tires and college football and whatever else some advertising algorithm says that Nickelback fans also like. But there are a couple of things to note. First, as Fox-Brewster observed, “The vague policy also doesn’t make it clear what kinds of information would or wouldn’t be collected and shared,” which isn’t exactly reassuring.

More notably, though, this is a clear signpost as to what Spotify might do next. It’s been apparent for some time that simply charging a subscription fee doesn’t generate enough revenue to cover the cost of offering a catalog compelling enough to convince users to pay that fee. Tech businesses are notorious for running at a loss for long periods of time, but there’s only so much VC capital you can hoover up before you need to start making money. And when that moment comes, Spotify is going to sell something else apart from music. It’s going to sell: you.