Business Matters: The Butterfly Effect and Why iTunes Radio Hasn't Increased Track Sales

A 1972 research paper by Edward Lorenz titled "Predictability: Does the Flap of a Butterfly's Wings in Brazil Set Off a Tornado in Texas?" introduced the term "the butterfly effect" into popular lexicon. It's generally understood to mean small events (a butterfly flapping its wings in Brazil) can have larger, unintended consequences (a tornado in Texas). Another way to say it is, "One thing leads to another." 

A new report by Music Forecasting shows why iTunes Radio isn't the boon to track sales people thought it could be. Listeners of Internet radio aren't opposed to buying music. They simply don't want to interrupt the listening experience to hit the buy button. And it could end up having unintended consequences. 

Music Forecasting believes it knows the problem. A new Music Forecasting report claims that iTunes Radio listeners use the service for a "lean-back" listening experience. Like other forms of radio, iTunes Radio provides listeners a variety of music for very little effort. Such an experience is not conducive to building a music collection. 

This finding should deflate the some of the optimism that originally surrounded iTunes Radio. Launched Sept. 18, Apple's answer to Pandora was thought by many in the music business to have potential for incremental track purchases. The service tightly integrates a buy button that allows the listener to buy a track from the iTunes Music Store. It got off to a quick start, attracting 20 million listeners in about a month

But no sales boost ever materialized. As I noted in November, a dip in track sales coincided with the launch of the service. Track sales were down 12.9% in the fourth quarter and fell an equal amount in the first three weeks of 2014. 

Buying music on iTunes Radio clashes with the nature of radio. Users simply don't want to lean forward to buy music when they're enjoying iTunes Radio's lean-back listening experience. The availability of free music doesn't prevent people from buying tracks, explains Sam Milkman, executive VP at Music Forecasting. The interviews revealed that most people still buy music -- both digital and physical -- and want to own a music collection. 

iTunes Radio will result in some level of purchases. The 40 to 50 people interviewed in Philadelphia and Raleigh, North Carolina are "very aware" of the button and think it's a good idea. Other Internet radio services such as Pandora have long had buy buttons that take the listener to either iTunes or Amazon MP3. But Milkman doesn't believe iTunes Radio's buy button will result in a large increase in incremental purchases. Instead, purchases will be casual and sporadic. "It seems more like a trickle than a flood." 

iTunes Radio's inability to sell tracks isn't the main problem. The butterfly that causes a storm could be the difference between actual sales and expected sales. Recall that Apple negotiated directly with record labels for iTunes Radio rather than rely on the compulsory license used by Pandora. Apple's licensing terms works out to roughly 0.13 cents per stream in the first year, according to Billboard's estimates. (Pandora currently pays 0.12 cents for advertising supported stream and 0.22 for streams derived from paying subscribers.) Apple also got concessions that reduce its royalty burden: iTunes Radio can play up to two non-royalty-bearing tracks an hour that are promotional in nature. 

Why didn't labels get more from Apple? One plausible explanation is they believed iTunes Radio's buy button would result in incremental sales. Without an expectation of incremental sales, labels would have demanded higher royalties. Like the butterfly in Brazil, these licensing terms could have unintended consequences. 

In years past, the Copyright Royalty Board would look to royalties paid by on-demand services when setting statutory rates, not exactly an apples-to-apples comparison. This time the CRB will use iTunes Radio licenses -- set using optimistic sales expectations -- as a point of reference when setting rates for 2016 to 2020. Anything that changes future statutory rates will have very real financial consequences. If you assume statutory royalty payouts will rise 20% a year until 2015 and then 10% until 2020, a 0.1-cent increase in the statutory rate would pay an additional $183 million to labels and artists from 2016 to 2020. What started as a flutter ends up a multi-million tornado.