Thesis challenges music industry to change the way it does business online

Aaron Zimmerman ’04, an operations research and financial engineering major, asserts that “We now have a whole generation of people who are accustomed to stealing.”

He has a point. He refers to the rampant practice of illegally downloading music from the Internet and lays the lion’s share of the blame upon the music industry’s neglect of the online marketplace.

Aaron suggests that record companies boost their online sales by overhauling their pricing system, and even takes that first step in his thesis project called “Pricing to a Different Tune: A Mathematical Model for the Sale of Music in the Digital Marketplace.” His adviser was Warren Powell ’77.

Stealing is stealing
Aaron said that while he was in high school, if he and his classmates wanted the newest music, they would simply go to the store and buy the CD. It never would have occurred to them to slip the CD into their back pocket and sneak out without paying. Yet now, at college campuses across the country, many young people are in the habit of illegally downloading free, bootlegged copies of individual songs.

“You wouldn’t walk into a bank and start taking money,” Aaron said. “And yet stealing music has almost become acceptable.”

Part of the problem, Aaron said, is the music industry’s too-little-too-late response to technological advances in digital music. The Wall Street Journal reports that the availability of and demand for digital music continues to grow while album sales continue to drop. However, Aaron argues, rather than diving into the new marketplace created by the technology, the music industry has only dipped its toe into digital music’s earning potential.

He said that although several record companies own online music stores where digital songs can be purchased, they don’t run them with the same attention they do other businesses. He said the biggest evidence of this is their flat pricing system.

Value varies
“Some sites charge 99 cents for everything, others charge 88 cents,” he said. “Either way it makes no sense to me. So when I started thinking about my thesis, two things came to my mind. Different songs require different prices. Why should track one, which is a big hit with lots of radio play, sell for as much as track 14, which is a B-side?

“Also, a song’s value changes over time, normally dropping. I wondered how prices should be lowered over time.”

These were the problems Aaron sought to address in his dynamic pricing scheme. The model he developed has four main elements: market size, absorption curve, demand, and price.

Record companies routinely assess demographics to determine the size of the market for various albums. The same logic could be applied to individual songs, and the market size would thus represent the highest possible number of copies sold.

The absorption curve represents the changing value of a song over time. Each point on the curve represents the percentage of the total market size interested in purchasing the song during a particular week.

“When a song first comes out, people do not know it, so its value is low,” Aaron said. “But if it gets some radio play, the value goes up because buyers are willing to spend more to get the song. Then, after a couple months of radio play, people may only be willing to spend a fraction of that.”

Demand is a measurement of what percentage of consumers want to buy a particular song for a particular price.

Lastly, there is the price. Is it 15 cents? Is it $1.25? What dollar amount will maximize profit, or at least minimize losses?

Aaron used empirical music sales data gathered by the University of Pennsylvania to develop his model. After completing the model, he ran and compared three different scenarios: the perfect-information scenario, the flat-pricing scenario, and the dynamic-pricing scenario.

The perfect-information scenario is the ideal situation, where a company can make the maximum amount of revenue possible for one time period. Clearly this is an ideal, but companies hope to get as close to that revenue as possible.

The flat-price scenario is the current scenario of giving all songs the same price at all points in the time period.

The dynamic information scenario falls somewhere in between. At the end of each week, the company analyzes the revenues received and uses the information to predict the best price for the next week. It is this final scenario that Aaron believes should be used by the digital music companies.

“These companies keep very detailed records about the popularity and radio play of a song and the demographic market for an artist,” he said. “They have the historical data and the current data, and I believe that with that they could develop a dynamic pricing scheme that would greatly improve their business. It’s absurd to think that every song has the same value throughout time.”

Aaron is not sure where to take his model now. He has considered sending his model to people in the music industry, but he is not sure that the industry is as yet receptive to change.

“Should people who are in the industry look at it? Definitely!’ Aaron said. “They had control over pricing before, and now they don’t. They can’t charge whatever price they want to, because people will just steal the songs instead. They need to listen to the consumer, stop suing people for downloading free music, and find some type of pricing system that works.”