In the early 1900s, jazz musicians refused to record phonograph records because they feared rivals would cop their best licks. We can laugh at their shortsightedness, but it's reminiscent of today's music industry, which is so afraid of piracy it still hasn't figured out how to incorporate digital downloads into a sustainable business model. Each year record companies ship about 800 million compact discs —nearly 10 billion songs. That sounds like a lot until you compare it to the 13 billion songs that were available (according to download tracker BigChampagne) for free on peer-to-peer networks in 2004.
The one bright spot for the industry has been Apple's iTunes store, which has sold 600 million songs since 2003, accounting for 80 percent of legal downloads in the United States. Piracy is clearly here to stay, but as iTunes has shown, the record companies' best strategy is to provide an easy-to-use service that offers music downloads at a fair price. But what price is "fair"? Apple says it is 99 cents a song. Of this, Apple gets a sliver—4 cents—while the music publishers snag 8 cents and the record companies pocket most of the rest. Even though record companies earn more per track from downloads than CD sales, industry execs have been pushing for more. One option is a tiered pricing model, with the most popular tunes selling for as much as $3. After all, the music honchos reason, people pay up to $3 for cell-phone ring tones, mere snippets of songs.
Steve Jobs, who has been willing to take a few pennies per download so long as he sells bushels of iPods, calls tiered pricing "greedy." That view is shared by millions of consumers who believe the record companies have been gouging them for years. From the buyer's perspective, however, Apple's 99-cents-for-everything model isn't perfect. Isn't 99 cents too much to pay for music that appeals to just a few people?
What we need is a system that will continue to pack the corporate coffers yet be fair to music lovers. The solution: a real-time commodities market that combines aspects of Apple's iTunes, Nasdaq, the Chicago Mercantile Exchange, Priceline, and eBay.
Here's how it would work: Songs would be priced strictly on demand. The more people who download the latest Eminem single, the higher the price will go. The same is true in reverse—the fewer people who buy a song, the lower the price goes. Music prices would oscillate like stocks on Nasdaq, with the current cost pegged to up-to-the-second changes in the number of downloads. In essence, this is a pure free-market solution—the market alone would determine price.
Since millions of tunes sit on servers waiting to be downloaded, the vast majority of them quite obscure, sellers would benefit because it would create increased demand for music that would otherwise sit unpurchased. If a single climbed to $5, consumers couldn't complain that it costs too much, since they would be the ones driving up the price. And enthusiasts of low-selling genres would rejoice, since songs with limited appeal—John Coltrane Quartet pieces from the early 1960s, for example—would be priced far below 99 cents.
The technology for such a real-time music market already exists. The stock exchanges keep track of hundreds of millions of transactions every day and calculate each stock to the quarter-penny in real-time. Banks are able to do the same with hundreds of millions of ATM withdrawals. A music market would actually be much simpler. When a trader on the Chicago Mercantile Exchange buys soybean futures, he has to take into account weather, crop yields, supplies in other parts of the world, and the overall economy. On the Digital Music Exchange, there is only one input: demand.
The interface could look something like Apple's iTunes, where users search for songs they want. One important addition would be a ticker that calculates the number of times a track has been downloaded. Click on the icon to see how much it costs right now. Click again and you freeze the price—we'll give you something like 90 seconds to make up your mind—and make the purchase. If you buy a track for $1, that doesn't necessarily mean the price goes up for the next person. Just like on the stock market, it might take a lot of transactions to move the market. Another potential feature, stolen from Priceline: If you tell the system how much you're willing to pay for the new 50 Cent single—say, less than 50 cents—it could send you an e-mail alert when the market is willing to meet your price.
This is all really just a corollary to Chris Anderson's Long Tail theory. In the material world, stores sell goods that generate a satisfactory return on the space they eat up. According to Anderson, the editor in chief of Wired, your run-of-the-mill record store has to sell at least two copies of a CD per year to compensate for the half-inch of space it takes up on the shelf. But in the digital realm, there is no shelf space. Infinite amounts of product are available. Instead of a hit-driven culture, we experience what a friend of mine calls "an embarrassment of niches." A record company doesn't have to depend on one album to rack up sales of 5 million. They can make the same money selling 500 copies of 10,000 different titles, or, for that matter, 5 copies of 1 million titles.
Of course, there are modest fixed costs associated with this pricing model: bandwidth, servers, office space, electricity, and the salaries of people who maintain the business. That means there would have to be a price floor, perhaps 25 cents a song. But each obscure indie rock or klezmer song that gets sold for a quarter is almost pure profit, and the bargain-basement price would induce people to download even more tunes.