how much money do artists make from itunes or streaming services

I decided to write this post after stumbling across an article in The Atlantic titled “How Musicians Really Make Money in One Long Graph.” It showed how many monthly sales or streams an artist would have to achieve in order to earn monthly minimum wage. The information, which came from informationisbeautiful.net was unbelievable. But not unbelievable as in really surprising or shocking–literally unbelievable. For example, the article said that an artist would need over 4,000,000 streams per month on Spotify to make monthly minimum wage. After doing more research, I found that this claim is untrue, at least in many situations. This post will focus on how money is made from music downloading and streaming because I doubt physical music has much of a future. Although physical music was more popular than purchasing downloads in 2012, physical sales declined a whopping 12.8%, according to The Huffington Post. This post will use information about Spotify, Pandora, and iTunes because they are the most popular online music services, even though using Pandora when Spotify exists is like using a typewriter when you have a computer. I didn’t actually bother to check the claim that those three are the most popular, but how many people do you know that use Napster or Amazon? (Yes, Napster is still a thing and it does streaming now.)

spotify’s royalties

Spotify is a music streaming service that offers users three options: listening for free on a computer with ads, paying $5 a month to listen without ads, or paying $10 for being able to also listen on mobile devices and download music. The actual formula for calculating royalty payments for Spotify is very complicated, but Rolling Stone gave the royalty arrangement of one band manager. In that situation, if a song is streamed 60 times, the songwriter receives 9 cents and the artist receives 38 cents, which is then split with the label depending on the contract. Using the Rolling Stone rates, I calculated that an artist would make over $25,000, which would be split with a label, if people streamed their songs 4,053,110 times, the amount the Atlantic article says would be needed to make monthly minimum wage. I also calculated the combined amount that an artist and label would make using the Atlantic rates, which was over $7,500, less than a third of the money Rolling Stone reported would be made. Since the actual Spotify royalty formula is extremely complicated and is different for different artists, I checked a couple other sources to get an idea of which figure was generally more accurate. According to NPR, Erin McKeown, who doesn’t have a label, gets $0.004 per play, which means she would make over $16,000 if her music was streamed 4,053,110 times. A second Atlantic article, which came out after the unbelievable one, talks about how cellist Zoë Keating’s songs had been streamed 72,800 times and she only made $281.87, meaning she makes around $0.0039 per play, almost identical to the amount McKeown is payed (note that Keating’s figures are after CD Baby’s 9% commission). In a 2011 blog post titled “Spot the Spotify Payment,” independent artist David Harrell writes,

Without any promotion in the regions where Spotify is currently available, our total number of plays is relatively small, though our Spotify activity seems to be increasing each month. The per-spin payouts we receive via CD Baby are quite variable, ranging from around two hundredths of cent to more than one cent for each stream. (We also had a few spins that rounded out to ‘$0.00000000’ after CD Baby’s commission.) I’m assuming the payout amount depends on free vs. premium listens, as well the subscription prices in each region and currency exchange rates.

Harrell says that he made an average of approximately $0.0029 per play from August 2009 to March 2011 before CD Baby’s 9% commission, meaning he would make about $10,500 if his music was streamed the amount of times the first Atlantic article claimed was required to make monthly minimum wage (Harrell later made an updated post where he said the average payment for stream in June, 2012 was around $0.008). Just to clarify, McKeown, Keating, and Harrell are all independent artists. According to Keating:

Spotify does not pay the same per play to Indie rights holders as it does to Major labels. Majors are shareholders in Spotify and their deals are confidential. That matters to me, but doesn’t seem to matter to others.

Therefore, I can’t disprove the original Atlantic article, but the fact that information from three indie artists and Rolling Stone provide much higher royalty situations at least demonstrates that the Atlantic article greatly oversimplified Spotify’s royalty agreements. Also, it is important to mention that a 2013 New York Times article said the following about Spotify’s rates, further explaining how complicated the company’s payment methods are:

Spotify declined to comment on its rates, but according to a number of music executives who have negotiated with the company, it generally pays 0.5 to 0.7 cent a stream (or $5,000 to $7,000 per million plays) for its paid tier, and as much as 90 percent less for its free tier.

is spotify bad for artists?

Keating says she made over $45,000 dollars on iTunes from October 2011 to March 2012, the same time period when she made less than $300 on Spotify. Does this mean artists shouldn’t use the service? Not necessarily. Keating also explains:

I think Spotify is awesome as a listening platform. In my opinion artists should be view it as a discovery service, rather than a source of income. […] I wish Spotify would do more to facilitate the connection between listeners and artists — i.e show that the artist is playing nearby, or add links to buy music. It’s early days, so maybe this will happen eventually. [sic]

I know Spotify currently provides concert notifications because I can see them on my own account. There are other arguments to why Spotify is good for artists as well. For example, it could decrease the amount of piracy because people can stream as much music as they want on their computer for free, although there are ads. Also, it can lead people to listen to more obscure tracks that they otherwise wouldn’t download. I know this from my personal experience. Before I used Spotify, I would rarely buy music from iTunes, opting instead for using Pandora. When I did buy music from iTunes, I usually only purchased my favorite tracks from my favorite artists. With Spotify, I listen to many of my favorite artists’ less popular songs that I would not have purchased on iTunes and likely wouldn’t have shown up on Pandora. In general, I listen to much more music now that I have Spotify than when I used Pandora and iTunes. People being exposed to a greater amount and variety of music means that they are probably more likely to attend live shows.

iTunes

Rolling Stone explains how iTunes works in this simple image:

The data is based on if an artist receives 16% of sales. Rolling Stone says that artists generally get 12-20% of sales, depending on their popularity. It is also important to note that an artist would get $0.89 if they were not on a label. iTunes does not only use $1.29 downloads, though. The aforementioned New York Times article says that “[o]n a 99-cent download, a typical artist may earn 7 to 10 cents after deductions for the retailer, the record company and the songwriter, music executives say.” However, that download is not specified to be from iTunes.

pandora

Rolling Stone explains how Pandora pays royalties:

The rates go up every year, but the broad formula is that big ‘pure play’ companies, such as Pandora and Slacker, pay either 25 percent of their total revenue per year, or a little more than $.001 per song — whichever is greater. These payments go to a music-business collection agency known as SoundExchange, which then pays 50 percent of it to the copyright owner (usually a record label like Warner or Sony), 45 percent to the artist and 5 percent to non-featured performers. Smaller Internet radio companies pay slightly lower rates.

They also provide a graphic to illustrate this concept.

If you want to learn more about how artists make money in today’s complicated music industry, I recommend checking out David Harrell’s blog and money.futureofmusic.org. The latter site provides very in-depth information on how musicians make money.

Update: Since I wrote this post well over a year ago, there have been many more interesting articles written on this subject. Also, many celebrity musicians have weighed in on whether Spotify and other streaming services are good for the industry. Here are a few of the most notable celebrity opinions about Spotify, Pandora, etc.:

  • Taylor Swift pulled her entire catalog from Spotify. Her label claims that she earned less than $500,000 for US streams from the service in the 12 months prior to her music’s removal, less than she would make by selling 50,000 albums. Spotify says that she was paid $2 million for worldwide streams and would have been able to make $6 million a year if she stayed with the service.
  • Bette Midler made an anti-streaming tweet that got a lot of attention. However, her royalties do not seem typical; if you use the Pandora royalty rates from the Rolling Stone article, Midler should have made around $1,880. A Pandora spokesperson said that the company paid over $6,400 for Midler’s streams (I assume the spokesperson is referring to the total payment before it was divided between Midler and the label). I don’t know why Midler reports that her royalties were so low. Maybe she is in a bad situation with her label.

  • Comedian Joe Mande criticized Spotify’s low royalties on Twitter.

  • Nigel Godrich and Thom Yorke of Atoms for Peace made a series of tweets explaining why they’re pulling music off of Spotify. In an interview, Yorke called Spotify the “last desperate fart of a dying corpse.” Yorke’s other band, Radiohead, self-released their 2007 album online and let fans pay any price for it.
  • Laura Sheeran, Ed Sheeran’s cousin, removed her music from Spotify after she made only £10.84 (about $16) for 40,000 streams.
  • Bono defended Spotify by saying, “I see streaming services as quite exciting ways to get to people. In the end, that’s what we want for U2 songs.”
  • Ed Sheeran said he supports streaming services because they give exposure to his work, which makes it more likely that people will go to his shows.
  • David Lowery of the band Cracker says that he was paid only $16.89 for over a million streams on Pandora of the song “Low,” which he owns 40% of. Using the Rolling Stone rates, he would have been paid a little over $200, so either Rolling Stone‘s or Lowery’s rates are inaccurate.
  • Aloe Blacc, who co-wrote and sang Avicii’s hit song “Wake Me Up,” wrote a piece in Wired calling for streaming services to “pay songwriters fairly.” He says that he was paid less than $4,000 by Pandora for co-writing the song, which was streamed over 168 million times in America. However, Pandora says it paid $250,000 to rights holders for “Wake Me Up,” so Blacc’s low royalties must be due to the way the money is split between the rights holders.
  • Jay-Z will likely acquire Aspiro, which owns both WiMP, a Scandinavian streaming service, and Tidal, the US and UK version of WiMP. Jay-Z’s company will pay $56 million if the deal goes through.

more interesting articles on streaming services

  • Streaming is credited with dramatically reducing the amount of music piracy in Norway. An International Federation of the Phonographic Industry (IFPI) study showed that in 2009, 80% of Norwegians under 30 downloaded music illegally; in 2014, the rate had decreased to 4%. In the first half of 2013, total music sales in Norway increased by 17%, thanks to streaming.
  • Spotify now has a website called Spotify Artists, which explains the company’s royalty policy in detail. Spotify says that, on average, artists earn around $0.006 to $0.0084 per stream. (These rates mean that Laura Sheeran should have made around $240-$336 for 40,000 streams, instead of merely $16.) However, the company stresses that it does not pay a fixed royalty for every stream a song gets. The image below shows Spotify’s formula for paying out royalties. Since the label or publisher is paid royalties from Spotify and it then distributes them to the artist, the amount of money an artist makes from streaming varies depending on the artist’s contract.
  • Time calculated the Spotify royalties for the most popular songs of 2014.
  • Cuepoint, a music blog, published an article with suggestions on how to improve Spotify’s royalty system. The author suggests that profits from users with paid subscriptions should be divided between only the artists that those users listened to. For example, if a user with a paid subscription listens to an obscure death metal band one time and then never uses the service again, that band should get all of the money from that user, instead of it being divided among all of the artists on Spotify, including ones that the user never played. Another blogger, Brendan Moore, made a post explaining what the new payout to artists would be from his subscription to Rdio, a streaming service similar to Spotify, if the company adopted the proposed royalty model.

If you’re interested in this subject and have a lot of time on your hands, you can read some of The Guardian‘s ten recommended articles about Spotify. If you really have a lot of time on your hands, you can read a 25-page essay by Aram Sinnreich, an assistant professor at Rutgers University, called “Slicing the Pie: The Search for an Equitable Recorded Music Economy.”

If you liked this post, you may want to check out my post on how movie theaters make money and if they should change their pricing system.

Advertisements

movieconomics: should movie theaters change their pricing system?

Have you ever wondered how movie theaters make money? Sure, they make it off ticket sales, but how do they split the profits with the movie studio? And why do all movie tickets cost the same, regardless of the quality of the movie or the time of the year? These questions led me to do some research and make a blog post on movie theater economics. My findings were more complicated than expected, but I have tried to simplify the facts by writing a paragraph or two about many of the possible questions one could have about how movie theaters make money and what they could do to make even more.

how do theaters and studios split profits?

This question is more complicated than it seems, since the method of splitting profits depends on the movie and theater. In At the Movies: The Economics of Exhibition Contracts, Filson, Switzer, and Besocke write (emphasis added):

While some contracts are “aggregate” deals, in which each side’s percentage share remains fixed throughout the run, most are ‘sliding scale’ deals. In a sliding scale deal, each week, the distributor gets the maximum of two possible payments: 1) 90% of the movie’s weekly ticket revenue over the ‘house nut,’ which is a flat payment to the exhibitor; 2) a ‘floor payment,’ some percentage of the weekly ticket revenue that typically declines according to a ‘sliding scale’ as the weeks go by – perhaps 70% in the first week, 60% by the third week, and as low as 30% at the end of the run. If the parties anticipate that revenue might peak in the second or later weeks (which can occur when the movie opens before a holiday weekend, for example) the contract includes a ‘best weeks’ clause that ensures that the high floor payments are associated with the high demand weeks.

The difference in profit splitting between different theaters and movies is demonstrated by a study of a theater chain in St. Louis, Missouri from the same article:

The sharing rules vary substantially by movie and theater. For example, in sliding scale deals the exhibitor’s share ranges from 23-70% in week 1, 30-70% in week 4, and 40-70% in week 8. Run lengths depend on performance and vary from 1 to 28 weeks in our data; the average run length is 5 weeks.

According to Forbes, theaters on average keep 45-55% of ticket sales throughout an entire theatrical run.

are concession profits split?

No, the theater takes all of it. Theaters make lots of money at concession stands; around half of their profits. According to Time, around 85 cents of every dollar spent at concession stands is profit for the theater. People often complain about the ridiculous price of items like soft drinks, but high concession prices keep ticket prices lower. The Stanford Graduate School of Business article “Why Does Movie Popcorn Cost So Much?” (the full article is actually worth reading if you’re interested in learning more on the subject) explains (emphasis added):

By charging high prices on concessions, exhibition houses are able to keep ticket prices lower, which allows more people to enjoy the silver-screen experience. The findings empirically answer the age-old question of whether it’s better to charge more for a primary product (in this case, the movie ticket) or a secondary product (the popcorn). Putting the premium on the ‘frill’ items, it turns out, indeed opens up the possibility for price-sensitive people to see films. That means more customers coming to theaters in general, and a nice profit from those who are willing to fork it over for the Gummy Bears.

why don’t theaters just lower ticket prices so they can sell more concessions?

Since theaters get all the money from concessions, but have to share profits from ticket sales, they have an incentive to lower ticket prices so more people will go to the theater and buy concessions. The previously mentioned article At the Movies: The Economics of Exhibition Contracts, discusses this issue:

The exhibitor gains from adjusting p only if the adjustment occurs after the contract is signed. If the exhibitor adjusts p before the contract is signed then the distributor will simply argue to change the sharing rule to ensure that the exhibitor still gets its reservation utility. Given that exhibitors include their proposed p in their initial bid, a subsequent adjustment would violate an implicit contract. Such behavior could cause the distributor (and perhaps others as well) to punish the exhibitor in the future. This encourages exhibitors to keep p up (15).

The letter p represents the price of the ticket.

why do tickets cost the same for all movies?

There is good reason to believe that theaters are making a mistake by charging equal prices for all movies. There is a detailed 22-page article on this subject called “Uniform Prices for Differentiated Goods: The Case for the Movie-Theater Industry,” which was published in the International Review of Law and Economics. I have provided a brief summary of the responses of the article’s authors, Orbach and Einav, to five main perceived problems with nonuniform ticket pricing.

1. Consumers would perceive theaters as being unfair if prices were not uniform.

This isn’t necessarily true. First, theaters started using uniform pricing in the early 70s, so because people used to deal with nonuniform pricing, it is not unrealistic to think that they would be willing to accept it again. Secondly, not all movies are the same and charging different prices for different products is not really unfair. Orbach and Einav explain (emphasis added):

Despite the difficulties that fairness perceptions may present, they cannot justify uniform admission prices. Uniform prices seem fair because of the system’s regularity, not because of any intrinsic justice. No sophisticated schemes and ploys are needed to change the present reference transaction; in fact, simple marketing mechanisms could do the trick. The starting point is the distinction between the movie and the show-time puzzles. All consumers are familiar with the concept of different prices for different products, although for some consumers charging different prices for products they consume at different points of time may appear unfair at first. Thus, charging premiums or giving discounts for unique categories of movies is unlikely to be perceived as unfair. For example, given the unique characteristics and highly publicized production budgets of event movies, charging premiums for such movies probably would not violate fairness perceptions. Indeed, past experience and international markets provide good reasons to believe that patrons would not perceive such a practice as unfair. It also seems unlikely that discounts for films whose demand is relatively elastic, such as documentaries, would be perceived as unfair (145-46).

2. Consumers would be less likely to go to movies with lower prices because they would believe that they are worse movies.

The Avengers is an example of an event movie.

Orbach and Einav believe that this argument is only valid for movies that are similar. If a theater charged more for an “event” movie, which is a movie whose arrival is a big deal and often has a huge budget or is a sequel to a popular film, consumers would probably not believe this action signifies that the other films are worse. Also, if prices were lowered on different types of movies, such as documentaries, demand may actually increase, and consumers would be able to tell that the lowered price does not mean the movie isn’t good (146).

3. Theaters would not be able to know which movies should be priced higher.

This film was so unpopular that the Lone Ranger probably wasn’t the only lone person at many showings.

Frankly, this just seems like a dumb argument. Sure, sometimes movies perform a lot worse than expected, like the Lone Ranger, but in general I’m pretty sure companies can tell that the sequel to The Avengers is going to make a lot more than, say, Ant-Man, which is an actual superhero film scheduled for a 2015 release. In fact, there is an entire website dedicated to predicting how well films will perform at the box office called the Hollywood Stock Exchange. HSX, which may be the subject of a future blog post, is a virtual stock exchange where players buy and sell fake stocks of movies and actors. Orbach and Einav’s response to the demand uncertainty criticism is as follows (emphasis added):

…empirical evidence shows that demand uncertainty is not as great as popularly argued and that the determinants of success in the industry are not totally random (see Section 2.2.1 and Moul, 2004). Producers may be unable to predict the demand for specific movies, but can identify with some level of confidence certain categories of movies with unique demand characteristics, such as event movies and documentaries. In fact, studios make significant investments in studying consumers’ preferences to improve their ability to forecast box-office revenues (Bakker, 2003). Empirical evidence suggests that production budgets, sequels, participation of stars and top directors, ratings, competition from other movies, and advertising are all significantly related to revenues and thus can be incorporated into pricing decisions  (147).

4. Nonuniform pricing could result in people taking advantage of the system by sneaking into higher priced movies.

The Atlantic article “Why Do All Movie Tickets Cost The Same” explains, “[Nonuniform pricing] would create a fascinating incentive for art-house studios to release smaller, cheaper films the same weekend as blockbusters, knowing that thousands of canny consumers might buy fake tickets to their show to sneak into the more expensive blockbuster.” However, Orbach and Einav essentially debunk the claim that this problem is serious enough to justify uniform pricing (emphasis added):

In practice, exhibitors already employ mechanisms that could mitigate much of the arbitrage opportunities. Today, some screens are sold out while others are not, and, therefore, exhibitors must monitor the patrons entering sold-out movies. Otherwise, patrons who could not purchase a ticket to a sold-out movie could use a ticket to another movie. Similarly, exhibitors must prevent moviegoers who buy matinee tickets from watching evening shows and, at the multiplex, they must prevent moviegoers from watching two movies in a row for the price of one movie. One indication of the possibility to administer variable pricing with reasonable monitoring costs is the prevalence of price discrimination across seats in some countries (Cheung, 1977) and at some theaters in the United States, as well as the prevalence of reserved seating in many international markets. Put simply, the arbitrage opportunities under a variable-pricing regime are similar to those that already exist today and, therefore, cannot explain a firm uniform pricing regime (148).

5. Theaters could take advantage of nonuniform pricing and the studios would make less money.

I’ve already addressed the idea that theaters could just lower prices and make more money because more concessions would be bought. Another concern involving theaters taking advantage of varying prices is the fact that they could report that sales for higher-priced movies were actually for lower-priced movies and take the difference for themselves. As Orbach and Einav point out, though, this complaint makes little sense, since theaters could already report more sales for movies which they get a higher percentage of the price from. In addition, they could just under report how many people bought tickets and keep more money (148). The previously mentioned Atlantic article makes a couple of additional criticisms. The first is that theaters already  favor certain movies by giving them more and larger screens. I honestly don’t understand why that is an argument against  varied prices. Secondly, the article says that another reason to not adopt nonuniform pricing is that if one theater raises prices for certain moves, and another doesn’t, some people would go to the lower-priced theater even if they live closer to the one with higher prices. This concern seems legitimate, but it only applies to people who are willing to travel a potentially much longer distance to save what is probably not much money. Therefore, I doubt it is a sufficient reason to use uniform pricing. Also, if all theaters raised the price of a certain movie by the same amount, there would be no stealing of customers.

why do tickets cost the same throughout the year?

Although matinees cost less, movies are otherwise priced the same year-round. However, if demand is greater on certain days or seasons than on others, it makes sense for theaters to raise ticket prices.

Photo from The Atlantic.

On the subject of varied prices based on the time of year or day of the week, Orbach and Einav write (emphasis added):

Much of the reluctance to adopt variable pricing stems from formidable concerns about the costs of administering and policing price differentiation across movies. Such differentiation may present a challenge to the industry, but it should neither be confused with, nor should it supersede, price differentiation across show times. Practical obstacles to variable pricing across movies, if such exist, cannot explain why movies are priced uniformly on weekdays and weekends and throughout the year. Thus, those who remain skeptic of the viability of price differentiation across movies, should still accept the view regarding the likely profitability and viability of non-uniform prices along the show-time dimension (150).

Another idea for how to vary ticket prices is by reducing the price every week or so. A criticism of this idea raised in the Atlantic is that “[y]ou can’t consistently cut prices after a successful opening weekend. If people knew that ticket prices would fall after a big opening, many more would wait until the second or third weekend to see it, which would, ironically, destroy the meaning of opening weekends.” I think this claim is mostly true, but it may not be the case for event movies. I remember a few years ago my math teacher was talking about if movie prices should be nonuniform and asked the class how many of us would be willing to pay $25 to see the latest Harry Potter film on opening day. Around a quarter of the class raised their hands. Therefore, theaters could potentially raise prices to very high amounts early on for event movies, and later the price could become the same as other movies. This way everyone who doesn’t want to pay extra would pay the same price as they would if all prices were equal, and the theater would make more money.

will we continue to see uniform ticket prices in the future?

We may not. According to ABC News, Steven Spielberg recently said, “You’re going to have to pay $25 to see the next ‘Ironman.’  And you’re probably only going to have to pay $7 to see ‘Lincoln,'” suggesting future high-budget films will cost more to see than lower-budget ones.

have movie tickets become more expensive over the years?

Photo from The New York Times.

Yes and no. As the graph to the side from a New York Times article shows, movie prices have indeed increased in recent years. However, the article also states that “Patrick Corcoran, the director of media and research for the National Association of Theatre Owners, points out that a ticket purchased for the average price of $1.65 in 1971 would cost $9.20 today — higher than the actual industry average, if adjusted according to the general inflation rate.”