Category Archives: Media Business

Snapchat is normal


Once upon a time, when you talked to your friends or showed them something amusing, you didn’t have to worry that your words and images would be recorded in a permanent, searchable database for your classmates, co-workers, employers, acquaintances, and distant cousins to scrutinize. Surreptitious recordings were possible but rare, and, short of blackmail, embarrassing bits weren’t easily spread.

Then came the Internet. It became more convenient to text, email, IM, Facebook or photo-share than call or meet face-to-face. First employers and schools, then parents and finally kids realized that we are being recorded. Within a few years children were taught not to send messages or pictures that revealed anything they didn’t want their mom or principal to see, and adults learned not to share anything they didn’t want strangers or employers to know.

What a bummer. Who wants to hang loose where every move is being recorded for future examination?  Who wants to grow up knowing that every immature utterance or gawky image from when they were ten or fifteen might come back to haunt them for the rest of their life?

Someone needed to turn communications back from dangerous to safe, to make it personal and transient as it was before Facebook, to make an invasion of our privacy take some effort— at least enough to be a violation of trust.

That’s Snapchat. Snapchat does the best job it can to make our remote communications relatively convenient and safe, even from our friends, family and co-workers, the folks most likely to cause us grief (or vice versa). Snapchat is the anti-Google and anti-Facebook. That’s one reason Snapchat turned down the billions those companies offered. Snapchat, not email, IM, or text may become the new normal. If so, a $3 billion dollar offer might not have been enough, when a less compelling form of Internet communication, Twitter, a simple message rebroadcast service, is worth $22 billion (as of July 2014).

Gollum and the music business


 Part 2: Radio Versus Records

Thanks to our friend Thomas Edison, you can still listen to the first (1878) recording of music here. It took another ten years for Emile Berliner to invent the record, the first mass produced music disk.

Music was the first media to move to the cloud in 1906 with Reginald’s Fessenden’s rendition of “O Holy Night” to ships at sea off the coast of Massachusetts. It took another 16 years for the first radio commercial to air, on WEAF in New York City.

Initially the record industry fought radio, but a symbiotic relationship between radio and recorded music was worked out within a few years. Record labels use airplay to generate sales; radio stations use music to sell advertising.

If music is free on the radio, why do we buy it? Radio only gives us a free taste of a song. If we want to own that song we have to buy it. While music under our complete control costs $.99 a song on iTunes, music under radio station control costs us a mere one cent (of advertising). (see previous post)

Just as Free-to-Play games stimulate demand by offering tantalizing, but limited access to a variety of virtual goods and powers, radio stimulates demand by offering tantalizing, but limited access to a variety of music.  The behavioral economics is the same. Temporarily possessing a song (or magic ring) then losing it triggers loss aversion, a powerful motivator, as Gollum learned to his regret.

The power of radio to stimulate demand for music purchases (through loss aversion) wasn’t obvious to the music industry when radio first blared onto the music scene.  Indeed the bliss of radio’s unlimited free music (and the Great Depression) blasted record sales for fifteen years.

Impact Of Radio On The Record Industry

For 67 years, from 1933 until 2000, radio and recorded music grew up together. By 2000, music’s high water mark, the USA recorded music was a $16B business and radio $20B. Over the next ten years, painless CD ripping and burning, low-cost Internet connectivity,  easy legal and illegal file sharing services, and open MP3 players, cut recorded music revenues in half.


Meanwhile radio held steady (until recently – see next post)


Radio, a primarily free (or more precisely ad-based) medium, is immune to piracy (but not other threats). Recorded music is not. If recorded music sales kept pace with overall growth of the economy (as it had the previous 67 years) then they would have reached $24B in 2013*. However they shrunk to $6B. This is a $18B per year loss to piracy (just in the USA).

* The US economy has doubled since 1997, and recorded music sales in 1997 were $12B (see graph above).

The Volume of Music Acquired Without Payment

Some bloggers defend pirates with surveys showing that pirates buy more music than non-pirates. That defense suffers from the standard correlation versus causation problem. Those who most wish to acquire music do so, legally and illegally. That correlation does not prove that more piracy sells more music. Where radio songs trigger loss aversion, pirated songs do not. In fact pirated songs are essentially indistinguishable from purchased songs, which makes policing (or even remembering they are pirated) difficult.

Perhaps the most elegant defender of piracy is The American Assembly from Columbia University. They rely on absurd data. They claim (on the following graph) that the average American:

  • From age ~11 to ~25 buys ~70 songs/$70 worth of music a year, reaching 1000 songs/$1000 music collections by age ~25.  (That’s triple what our expected purchases are).
  • Then at age 30, the average American reverses his behavior and starts throwing away 39 songs/$39 worth of music a year. [Since we actually buy 20 songs/$20* worth of recorded music a year, the only way to lose 19** songs a year while buying 20 is to throw out 39 songs each year!]


* Americans (300 million music consumers) buy $6.22B worth of songs each year.  That’s $20 of music purchases a year, or 20 songs a year @ $1/song.
** 1044 legal songs at age ~25 down to 376 legal songs by age ~70. That’s a loss of 668 songs in 35 years, or 19 songs thrown out every year for 35 years straight.

This is absurd. Americans buy music at a fairly steady rate throughout their first fifty years, those over forty-five becoming more prominent each year while those under thirty slowly disappear.

Music Sales by Customer Age

Data from 2011 Profile

Music’s sales disaster is no mystery. It’s caused by easy, safe and ad-supported piracy.

Radio drives hunger for music, but it may have doomed the recorded music industry. Why? We expect free music in the air. If you are reading this you probably think … what’s wrong with some more free music? If music is free on my radio, why not on my iPhone?

While the US Government could, as it does with digital audio recorders, simply tax any device capable of playing music a few dollars to compensate for piracy, this is not on the table. Taxes sound scary and we don’t care. What happens to the recorded music industry is someone else’s problem, not ours. [A classic example of Tragedy of the commons.] Our magical thinking about music economics coupled with our endless hunger for free music is pushing the recorded music industry into the crack of doom.

[Next post. Of course music won’t disappear like Gollum or his ring. What will happen?]

Unlike recorded music’s fall in 2000, radio has kept up the economy (until recently…)

Media Share of US Advertising



Does a song cost a penny?

Record (1)

Lincoln Penny

Before the Internet…

Radio and recorded music have been distinct businesses, with ad-based radio driving music disk sales. The power of the Internet is quickly erasing this hundred-year-old division, blending radio and recorded music into an integrated and personalized whole.

Part 1: How much does a song cost?

According to Price Waterhouse Coopers, audio media (radio and recorded music) is a $29B business with three main components: Radio ($19.72B = $16.38B Terrestrial + $3.34B Satellite ), online music services ($3.34B) and recorded music ($6.22B = $3.85B digital and $2.38B physical). Assuming ~300 million Americans listen, that’s $97 per person per year for recorded audio. [I’m not including live concerts, an additional $8.91B business.]

Audio, like video, requires some hardware, $7.76B last year, according to the Consumer Electronics Association, or $26 per person. In addition, audio hardware (575 million devices in the USA) uses 1.8% of residential energy, that’s 190 kWh/year at $.12/kWh or $9 per person.

$97 + $26 + $9 = $124/year = $11/month per listener.

For this we listen more than two hours a day. 

Time Estimates (per day, in hours:minutes)




Radio + Music

US Census


















USA Touchpoints




NAB  report (page 16)









*Averaging 1:59 for 243 radio million listeners, 0 for the 57 million non-listeners.
** Survey of 8-18 year olds

At 67.5 hours per month (2:15 * 30) per person we (300M American listeners) are paying a mere $.15 per hour

$.15/hour seems reasonable compared to $0.50/hour for TV or $2/hour for books. It combines the time and cost of rented music (radio and Internet music services), purchased music (disk and digital) and free (pirated) music.

Audio: Music Service  Versus Music Purchases

Radio, Satellite & Internet music services

Disks & digital music sales

Revenue ($M)



Time per person per day



Revenue per month per listener



Revenue per listener hour



Revenue per 4 minute* song



*Average song length 227 seconds or 3:47

Q: Does a song cost a penny?
A: Yes, one play costs one penny ($.01).

Song length stats

Song length



The Internet of Books. Can ebooks double book profits?

E-readers are the biggest change to the book business in 500 years. Over the next 20 years, as e-readers get better and cheaper ($399 in 2007 to $69 in 2013), they will replace paper books, just as over the past 20 years digital cameras replaced film. PwC projects that the crossover point (in the USA) will come in 2017.


Today, the average ebook costs $6.65, $3.21 cheaper than the average new book ($10.86), so it takes 36 books to offset the average $114 cost of an e-reader.

Lowering the cost and friction of book purchases (from a trip to a store to a click of a button), allowing readers to carry their entire libraries in something the size of a small book, and removing the piles of books that remind us we’ve already got too many books, must increase buying and reading of books. How much? Don’t make the same mistake as McKinsey.

More reading

McKinsey here is statistically illiterate, confusing correlation and causation. They state “Adding a device increases reading” but all they show is that people who have purchased more devices spend more time reading. I bet those owners spent more time reading before they purchased a reading device; in fact, that’s probably why they purchased the device!

At least their friends at Bain have taken statistics. They asked the right question, have ebooks changed your reading habits? 44% said they consume more books (and only 8% consumed less). We know both that e-readers have increased book consumption and that e-reader owners read quite a bit more than average. We don’t know how many more books are consumed after purchasing tablets or e-readers.

Piracy – the downside of ebooks?

Excluding countries without effective copyright enforcement (like Russia, with 92% ebook piracy), books are rarely pirated. While digital books are easy to pirate, the dominance of the walled gardens of Amazon Kindle, the Apple iPad, and the Barnes and Noble Nook, has kept piracy to an acceptable level of around 30%.  Perhaps as important, since you can only read a couple books per month(unlike music where you can listen to thousands of songs) the financial incentive to pirate (or be a pirate) is low. Finally the availability of free public and school library books removes the philosophical justification for book piracy.

All-you-can-eat books?

Since all-the-books-you-can-read by definition provides the maximum possible revenue per customer, why aren’t books offered all-you-can-consume, like Spotify or cable bundles with hundreds of stations of TV? While physical books can’t be offered in unlimited quantities for a reasonable price (except from a library) ebooks don’t have this limitation. Scribd ($9/month), Oyster ($10/month) and Entitle ($15-$28/month) are start-ups offering books (though not from Hachette, Macmillian, Penguin Random House or (except Entitle) Simon & Schuster, 4 of the 5 big book publishers) by subscription. Perhaps more relevant, Audible (bought by Amazon) does offer a good selection and big (though not unlimited) portion of audiobooks for $23/month. Truly all-you-can-eat audiobooks might cost (and produce optimal revenues) at around $49/month.

If subscription is better, why hasn’t it caught on? It takes years to negotiate the contracts required to enable new business models. Copyright holders must take care selling licenses to their property. For instance while Netflix has spent years negotiating, given its $8/month price point, it will never have the breadth of offerings as $60/month cable subscriptions. Any single discount tier provider will be relegated to content that has no better monetization options.

Clearly it’s high time for the book giants to try to take control over their own destiny, like Disney/ABC, Fox and NBC did when they created Hulu in 2007. In fact they are late; books and audiobooks are far closer to being dominated by Internet delivery than TV shows are. Eventually both TV and books will be offered over the Internet at varying subscription tiers, with higher priced tiers offering higher valued content.

The Internet of Books

The highest tier is the most interesting: The ability to search and read (and cross-reference) all books on any connected device, that is “The Internet of Books”. This would be an awesome educational or research tool, worth far more than what the best book customers (book whales) spend on books.

How much would subscription to “The Internet of Books” cost?

$99/person per month would be a safe starting price for a search-and-read-any-book subscription. Testing might yield an ideal price (maximum profit for the book industry) in the $49-$79 per month range. Since ebooks have virtually no inherent cost, any revenue increase should profit boost to the book publisher and author. Three million one-percenters might consider being able to access the Internet of books as important as sending their children to private schools. Businesses might want their top 1% employees to be able to instantly search and read all relevant educational, technical and business books. Graduate schools might consider it a necessity for all their students and faculty. If 1% of Americans bought $1000/year subscriptions to all books, that $3B would more than double the profitability of the entire trade book industry.

Question: Can ebooks double book profits?
Answer: Yes. Access to “The Internet of Books” would be both attractive and valuable, a far better business than selling one book at a time. Other less expensive (and less useful) subscription options would further increase book profits: 

  • $999/year: Search all, read all. “The Internet of Books”
  • $49/month: Search all, read many. ~10 books/month
  • $9/month: Search all, read one. Book-of-the-month club

Netflix and Spotify prove that subscriptions are a better business than one-off purchases. (Too many choices inhibits purchasing.) Since ebooks (and digital video or music) have almost no intrinsic costs, if you sell more, you (author, publisher, distributor) simply make more money. 30% of ebook buyers are ready now:

seven-years-age-of-reason-fig-02-08_full (1)

I want “The Internet of Books”. You do too, big book publishers. Let’s do it!

More info:


Which costs more, a book or a movie ticket?

The typical American read

The average new book costs $10.86, with 2.57B new books sold for $27.9B. The average used book costs about $3, with 1B used books sold for $3B. Combining new and used books we get 3.57B books for $31B, or $8.68 per book (new + used), a bit more than the average $8.38 movie ticket.

On a hourly basis, books are cheaper. With the average book at 64,000 words, and the average person reading at 250 words per minute, that’s $8.68 for about 4.3 hours, or $2.01 per hour of reading, less than half the $4.90 per hour for movies in theatres.

However the $2/hour rate ignores two key items, borrowed books and unread books. Only 48% of people bought their most recent book.

Book source

Adding in 3B (1.5B from public libraries + 1.5B from family and school libraries) borrowed books we get 6.6B books per year or 22.8 per reader. Apparently only half of these 22.8 (11.4) are read.

[A bit below Pew’s estimate of 12.3 books (15 reduced by the 18% of people Pew says read zero books) and a bit above Bain’s 10.3 (see below)], we’re down to 3.3B books read, or $2.17/hour.

In any case, with $31B spent by 280M readers, that’s $110 per year per reader, or $9 per month. for 1 book and ~4 hours of book reading per reader per month.

Question: Which costs more, books or movies?
Answer: It depends:

  • One book ($8.68) costs more than one movie ($8.38)
  • Per hour books are less than half the cost of movies ($2/hour versus almost $5)
  • Overall we spend 3x more on books, $31B for book versus only $10B for movie tickets, or (per person) $9/month for books, $3/month for movies.

More book stats:

Bain estimates 10.3 books (7.3 physical, 3.0 digital) per year


Is the movie business a tenth the size of the TV business?

movies tv














How much do movie tickets cost (on average)?

$8.38 at the theatre.  However we also have to watch ads, which account another for 7% of revenue ($.61), so $8.99 total. At 110 minutes per movie, movies cost  $4.90 per hour.

However unlike TV, only 13% of us watch a movie in a theatre every month. In the USA+Canada 1.36B tickets were sold for $10.8B. With a potential audience of 332M, that’s 4.1 tickets per person per year. The average USA/Canadian spends $3 per month on movie tickets. The average frequent (1 or more movies per month) moviegoer spends ~$12 per month.

Question: Is the movie business one-tenth the size of the TV business?
Answer: No*. In the USA we spend $10B on movies and $140B on TV so it’s only 7%.

*However, if we add in movie rental, sales and subscription service revenue … [future post]

Interesting that movies ($3 or 40 minutes/month) get so much Media attention and TV (~100 hours and $50/month) so little…

Bonus question: I can get a season pass to the ballpark. Why can’t I get a season pass to the movie theatre?

Since all-you-can-eat (e.g. Netflix) is the optimal (maximal revenue) media business model, why don’t theatre owners offer it?

With the typical theatre holding 225 seats, a fraction of the 20,000+ of a sports stadium, theatre owners must ration seats and spread out demand over several weeks. Unlike the steady demand for sports tickets, movies draw the largest audience in the first weekend and smaller audience after that. Theatre owners would like to increase revenue through (all-you-can eat) subscriptions, but keen movie fans don’t want passes that deny them from watching when the movie is most in demand. Nor do wall-sized movie displays offer any advantage to the closest seats, unlike live performances.

Over the past 100+ years, theatre owners have learned to optimize revenue with all-you-can-eat soda and popcorn, but don’t sell the tickets to seats this way, as the number of seats (unlike soda and popcorn) can’t be super-sized at near-zero cost.

Answer: Despite the problems listed above, MoviePass is trying to offer ‘Unlimited Movies” for $29.99 per month, a safe 10x ‘normal’ ARPU (average revenue per user).

What costs more, a cigarette or a TV show?


Part one of a series on what every media reporter and executive should know, but probably doesn’t.

What is more expensive, playing Batman or watching it? What is the real cost of watching TV or surfing the Internet on your smart phone? If you hate video ads or Facebook ads, how much should you have to pay to remove them? How much should it cost to be able to play everything, watching everything, listen to everything?

In the next series of posts I’m going to cover the cost of

  • TV
  • Movies
  • Books
  • Music and Radio
  • Games
  • Magazines & Newspapers
  • Phone service
  • Other services

Unlike many harried journalist I have reliable data sources for my many stats…

First, how much do Americans spend to watch TV?

We pay for TV in 3 ways:

  1. We buy a TV (or other hardware)
  2. We subscribe to a TV service
  3. We watch ads (or buy/rent a DVR)

How much do our TV’s cost?
We buy 37.1 million TVs at an average TV costs of $704. However it’s replaced only about once every 7 years. Per viewer (in the USA there is slightly more than 1 TV per person) this works out to $97/year, plus another $25/year in electricity. That’s a total of $122/year or just about $10/month per viewer.

How much does our TV service cost?
289 million viewers (USA) combine to spend $76B/year, or $22/month per viewer.

And what is the cost of those TV ads?
We watch $64B worth of Ads or $18/month per viewer.

The cost of TV + TV service + TV ads
$10 + $22 + $18 = $50/month.

For this we watch ~3 hours of TV a day or ~$.50/hour. That’s cheap entertainment!

Why ~3 hours a day and not ~5 (what Nielson reports)?


I use Bureau of Labor Statistics (BLS) data and then add 10%. I chose the BLS because Nielsen adds ‘background’ time; when the TV is on but we’re not in the room, when we’re in the room but not watching since we’re sleeping, playing a game, reading, surfing, cleaning, talking on the phone, …


My 10% addition to the BLS estimate represents my estimate that one third of this multitasking time (we spend 30% of our TV hours multitasking) we’re focusing on the TV.

Question: What costs more, a cigarette or a TV show?
Answer: About the same:

  • Since Americans spend $77B for 327B cigarette equivalents that’s $.24 a cigarette. At $.50/hour (above) that’s $.25 for a 30 minute show. So a cigarette ($.24) costs about the same as watching a TV show ($.25). 
  • Since a typical cigarette lasts about 6 minutes, smoking is 4x more expensive per minute than watching TV.  
  • Per month, Americans (on average, smokers and non-smokers combined) spend $34 on Tobacco, a bit more than the cost of the TV, TV service and electricity ($32/month per viewer), but less than the cost of TV service plus TV ad revenue ($18/month, or a total of $50/month). Of course as ‘only’ 44 million Americans smoke, our non-smokers spend $0 and our smokers spend $240 a month to smoke!

Bonus question. Which is worse for you, smoking or watching TV?
Answer: They are equally bad. Either watching a (half hour) show or smoking a (six minute) cigarette will shorten your life by 11 minutes.


  • 317M people in USA
  • 2.6 persons per household

More on TV

Electricity Stats