Why Spotify can never be profitable: The secret demands of record labels

Imagine a new hot-dog selling venture. Let’s also say there’s only one supplier to purchase hot dogs from. Instead of simply charging a fixed price for hot dogs, that supplier demands the HIGHER of the following: $ 1 per hot dog sold OR $ 2 for every customer served OR 50 percent of all revenues for anything sold in the store.In addition, the supplier requires a two-year minimum order of 300 hot dogs per day, payable all in advance. If fewer hot dogs are sold, there is no refund. If more than 300 hot dogs are sold each day, payments to the supplier are generated by calculating $ 2 per customer or 50 percent of total revenues, so an additional payment is due to the supplier. After the first two years, the supplier can unilaterally adjust any of the pricing terms and the shop can never switch suppliers.

Would this imaginary hot dog establishment be able to generate a profit? Never, because the economics are one-sided. The supplier will always elect the formula that captures the largest amount of money for themselves, completely disregarding the financial viability of the store. If the store miraculously managed to generate a profit, the landlord would simply raise the rates after two years.

Such economic demands may be imaginary for the hot dog business, but they are the stark reality that every digital-music subscription service such as Spotify, Rhapsody, MOG, Rdio, and others must confront. These details aren’t well-known because digital music service deals are always wrapped tightly with strict non-disclosure agreements.

For the first time, people are talking, and these previously secret demands are being made public. The specifics are even more onerous than the hot dog example cited above. Together they doom online audio companies to a life of subjugation to the labels, as you will learn below.

Here are some specific demands that digital music companies are compelled to agree to:

  1. General deal structure: Pay the largest of A) Pro-rata share of minimum of $ X per subscriber, B) Per-play costs at $ Y per play, C) Z percent of total company revenue, regardless of other business areas. As stated previously, this means labels de facto set retail price (they also regularly negotiate floors on price, giving even less wiggle room), which limits the ability of the music service to develop ancillary revenue streams that aren’t siphoned off by the labels.
  2. Labels receive equity stake. Not only do labels get to set the price on the service, they also get partial ownership of the company.
  3. Up front (and/or minimum) payments. Means large amounts of cash are necessary to even get into the game. In my experience, this further stifles innovation in services and business models.
  4. Detailed reporting, including monthly play counts. This seems rational enough — you would assume this information is necessary to pay artists and make other business decisions. The problem is, the labels each make additional demands, including providing additional reports unrelated to payment, including overall market share of sales in various categories. I doubt that, for example, phone manufacturers demand Best Buy provide the percentage of sales of competitors’ phones. The labels effectively offload their business analysis (and the cost of such analysis) onto the music services. I can’t think of another industry where that is standard practice.
  5. Data normalization. Labels all provide their data and files in different formats. That data is constantly changing as labels make available new material and make unavailable old material. This might seem trivial. It’s not. Without standard naming conventions and canonical methods for referencing artist, tracks and albums (ISRC and UPC don’t cut it), the services are left to try and match artist, track, album names provided by one label with those of another. It’s incredibly inefficient, as each service must undergo this process separately (although there are now companies that provide a service for doing this for the retailers).
  6. Publishing deals. Once you’ve signed deals with the labels, you then need to cut deals with the publishers. Determining ownership is a complete nightmare and there are huge holes in the licensable catalog. The data issues here are worse than with the labels. The long and short of it: Although you may have the rights to stream from labels, you sometime can’t get the rights to stream from the publisher, or worse, even find the publisher.
  7. Most favored nation. This is a deal term demanded by every major label that ensures the best terms provided to another label are available to it as well. This greatly constricts the ability to work out unique contractual terms and further limits business models. It is a form of collusion since each label gets the best terms the other label negotiates. It’s also why it’s easy to get one label (typically EMI) because they’ll provide low-cost terms knowing that others will demand higher rates, which EMI will then garner the benefit from.
  8. Non-disclosure. Every contract has strict language prohibiting the digital music company from revealing what they pay to the labels. If they speak publicly about any of the licensing terms, they jeopardize invalidating their license which would torpedo their business. Since labels license on behalf of the artists any payment to the artist comes from the labels not the digital music company. This is the main reason music services, not the labels, have been getting heat from the artist community. Music services can’t defend against accusations about low artist payments because they pay the labels who don’t disclose what they’re paying to the artists.

With most other businesses, if a supplier makes unreasonable demands, a retailer can turn to other providers. Since copyright law gives record labels and publishers a government-granted monopoly, no such option is possible with music. Digital vendors have only two options: Accept the terms or not include those songs in their offering.

The sale of EMI to other music companies means there will shortly be only three major labels. If a music service rejects terms offered by a label, then that service’s offering will have an enormous hole in their catalog of 25 percent or more of popular songs. In the business world, a monopoly leads to lopsided economics, and the subscription digital music business is a poignant illustration of that.

Final note: Online radio services such as Pandora take advantage of a government-supervised license available only to radio broadcasters thus sidestepping dealing with record labels. While the per-song fees are daunting, they bypass virtually all of the terms listed above.

A 15-year veteran of the digital music business, Michael Robertson is the founder and former CEO of MP3.com and is currently CEO of personal cloud music service MP3tunes as well as the radio recording service DAR.fm. He can be reached at michael@michaelrobertson.com. He would like to thank Paul Petrick for his contribution to this piece. 

Image courtesy of Flickr user walknboston

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

  • Carrier IQ and the continued erosion of operator trust
  • Defining work in the digital age: an analysis by GigaOM Pro
  • Forecast: the evolution of the e-book market



GigaOM