Search This Blog

Saturday, May 01, 2010

The Future of Recorded Music

Once upon a time, I wrote, arranged, played the instruments for, mixed, and produced 17 songs on two CDs (here and here - they're free!). I really wrestled with whether or not to try and make money from those efforts for, um, well 15 or 20 seconds - 25 seconds top. Okay, maybe "wrestled" is a little too strong, but hey, I did at least contemplate it.

The reason I decided not to even try to make money is that the success of a band (or artist) and its music is only partly dependent on the music. More important is the marketing: the shows the band does (choreography, pyrotechnics, stunts, and other entertaining moments), the persistence of the band staying in the public's consciousness, the news and buzz about the band (personalities, arrests, drug use, etc.), and so forth. Sure, the music of some popular bands is pretty good, but some popular bands have pretty marginal musicians with pretty marginal songs. In any case, the music of popular bands is no better in any objective sense than the music of many unknown bands.

Relatively recently, I stumbled upon an interesting blog called Cerebellum Blues. Jeff Shattuck, the blogger, suffered a brain injury and woke up with a "rekindled" desire to write songs. He's now recorded his first album and is wrestling with how to make money from his efforts. He's spending a lot longer than 25 seconds thinking about it though. That got me thinking about it again as well.

Unfortunately, I've come to the same conclusion: even though his songs are, in my opinion, quite good, he's going to have a tough, tough time of it. Especially since he can't really do the concert thing due to persistent dizziness due to his brain injury.

The standard business model assumes that the demand side of the equation is the people who want to listen to music and the supply side of the equation is those involved in producing the music to satisfy demand.

However, in the case of music, demand and supply are also inverted in some sense. Humans are driven to create, and creating music is particularly fun and satisfying (at least for people like me). It's even more satisfying when people actually listen to your music. For example, I can't imagine that even a single musician or song writer would say something even vaguely like, "I couldn't care less if people listen to my music as long as I make money at it." Everybody wants to write and play for other people. That's a fundamental need.

In the case of that need, the demand is for people to listen to your music. The supply is then those people who are willing to listen to that music. In other words, regarding satisfying the need of the musician or song-writer to have others listen to his or her music, that musician should be paying the listeners, not the other way around.

However, first let's consider supply and demand from the more traditional perspective. Even from this perspective it's tough and getting tougher to make money solely based on recorded music.


Let's say approx 1 in 10,000 people has wherewithal and motivation to write and record songs with no expectation of generating any income from the recording those songs. Why might these people do that?
1.) Musicians create "demos" to get gigs at various venues (vast majority).
2.) Serious hobbyist (me for example, also silicon-valley entrepreneurs after IPO).
I'm guessing that there are on the order 100,000 such people in the United States alone, and soon there will be 1,000,000 such people worldwide. Let's say each writes 10 songs a year. That's 10,000,000 new songs per year. However, songs are forever (for example, I noticed that my 13 year old daughter has a 40 year old album by the Beatles on her iPod). So there will be around 1 billion new songs created per century.

Cost of Supply

Each song can be replicated for tiny fractions of a penny. With today's equipment, it takes maybe 40 man-hours per song to record and engineer (mix) and the efforts will likely result in a pretty high quality recording. The recording equipment is an iMac (or other computer) with a few hundred dollars of software (or even free if you don't mind some limited functionality and an occasional crash) and a few microphones (for vocals and to mike the instruments, amps and drums). The musicians have instruments which may be quite expensive, but they probably don't need anything special for recording.

So there's very little cost and very little effort involved in recording songs. Other than the musicians' and song writer's effort, the total cost can be almost nothing.


My guess is that an average person is interested in listening to maybe 10,000 different songs in a lifetime. My personal observation is that about half the people can't even really tell if a recording is in tune or not and another 40% can tell but don't really care very much. As a result 90% of people are happy enough with fairly low quality recordings and certainly today's computer based recording equipment is plenty good for that. Therefore, the vast majority of songs being recorded are good enough for 90% of the population.

With 1 billion songs per century and the need for a total of 10,000 songs, that's an oversupply of 100,000 to one! Even dividing by some large number of genres and even if I'm off by orders of magnitude, that's quite an oversupply for 90% of the population.

Connecting Supply to Demand

One argument against free music in the past is that there may be millions of songs, but how does one find songs that they like? Certainly some intermediary like a recording exec needs to do some filtering, right?

Well, no. One problem is that the recording execs can't get through the millions of songs either, so that route guarantees inferior music being distributed.

Technology to the rescue! With the Internet radios (Pandora, Jango,, etc.) people rate songs so you'll be automagically fed songs you like!

More on Supply/Demand Inversion

It's actually already nearly gotten to the point where Supply versus Demand has inverted, where the supply is ears to listen and demand is songwriters who want to have their music heard. Jango has a service call Jango Airplay where song writers can pay to have their music heard. It's $30 for 1,000 plays (i.e. a thousand individuals will be fed your song when listening to Jango). Pretty cool really. And the 90% of the population that's not to picky will have no reason to ever pay for recorded music again. Who knows? Eventually they may even get paid to listen to songs.

Back to Jeff. The best chance he has, in my opinion, is that he has a unique and interesting story (at least to me): "Suffered a serious injury and woke up with a rekindled desire to write songs." Seems like that could get him some human interest stories in major media which might bring people to his site and sell his CDs.

However, from what I can tell, he seems determined to "make it" based on the quality of his songs. Though his songs are good, I wish him luck because I think he'll need it.


Jeff Shattuck said...

Definitely an interesting take. I love the notion of ears being the supply and human desire to create music the demand. I think you're onto something, but I need to mull it further.

As for my chances of making it, I agree 1000%. Sigh...

Hey Skipper said...


Interesting analysis. (Pretty good music, too. Hard to pigeonhole.)

I don’t disagree with it, but I take issue with some of your entering arguments.

Perhaps I’m missing something, but if 1 in 10,000 people in the US has the wherewithal and motivation to write and record songs, and the entire US population is the basis, then the number of producers will be 33,000, not 100,000. Strike everyone under fifteen and over sixty-five (former for lack of wherewithal, latter motivation), then the number will be around 20,000.

Still a long way from your friend Jeff’s ideal number, which would be 1.

Then breakdown that 20,000 into its constituent categories, largely distinct: whatever their relative merits, I doubt there are many opera devotees who would consider hip hop as anything other than Geneva Convention prohibited torture, and vice versa.

Assuming 10 distinct categories of music, that means about 2,000 artists, or 20,000 songs, per year (in the US).

Smaller than what you come up with, but still a long stretch from Jeff’s ideal number, which would be 10.

Here is where the real cutting comes in. Yes, notoriety, flash, and, particularly, luck play a far greater role in musical success than any of us would like. But, that doesn’t leave talent by the wayside; I have no idea of 90% of people couldn’t care for the quality of the recording, but they certainly do know whether they like the song.

Take those 2,000 songwriters in a given category, and play their product for each of the 20,000,000 devotees of that category. I’ll bet that most devotees would share wide agreement about which were the worst, and best, 500. Consequently, the odds of success are far greater for the latter than the former, which makes the consequent supply far less than you suggest.

So, just as before digital music, various filters severely winnow the universe of aspirants to a subset that has that extra amount of “knack” (However you want to define the term, it undoubtedly includes talent.).

The winnowing means may change. AM, then FM used to drive music. Not anymore. RadioParadise more or less hits my constituency, and when I hear something I like, then I go straight to iTunes and get it.

Which means the music industry has disintermediated — recording, reproduction and distribution no longer act as barriers to entry — and returned more to a single release driven model.

A similar pattern affects existing music, except the winnowing agent is time: not very much stands the test of it very well. When is the last time anyone listened to ELP?

Apply the filters of listen-worthy and time, and the supply of music isn’t nearly so large as raw numbers would suggest. The demand has always been artists, and the odds have always been stacked against them.

As for Jeff, of course he will need luck, but a lot less than most. My favorite, Demons & Saints, compares well with James McMurtry.

Bret said...

Hey Skipper wrote: "...but if 1 in 10,000 people in the US has the wherewithal..."

I was a bit (a lot?) sloppy. 1 in 10,000 world wide, the world wide population will peak out around 10 billion, that's 1 million. There's somewhat more than 1 in 10,000 in the U.S. is my guess.

Hey Skipper wrote: "When is the last time anyone listened to ELP?"

I'm sure people listen to ELP everyday.

I don't think time is winnowing quite as much as you think. A lot of new stuff is pretty formulaic and doesn't sound much different than old stuff. In the 1950s - 1980s a lot of new sounds became available because of technology. It seems to me music is evolving a lot more slowly than during that period.

But, maybe you're right. There may be only so many songs that people really want to listen to and they'd be willing to pay for those. However, that still leaves a huge number of artists who might still be willing to pay to have their stuff played.

Jeff Shattuck said...

"Hey Skipper"

Thanks for listening!

By the way, are you guys both stats dudes? Reading your comment and Bret's post reminds of reading a 10K or something!

Cool thinking.


Bret said...


I'm a roboticist and robot software does use a lot of statistics. Before that I was a futures trader (focusing on the trading model end of things) and that requires even more statistics.

Hey Skipper is a pilot, but an incredibly geeky one. :-)

His blog is The Daily Duck and any time there's an airplane crash or malfunction, he usually does an incredibly in depth post about it from a pilot's point of view.