If Apple were to try to sell music itself, Steve would have to convince the heads of all five major record companies that an independent online store operated by Apple was their best, and perhaps their only, choice, given the sophistication of the digital onslaught they faced. Even then, given their temerity, he’d have to bend over backward to give them a comfortable way to try it out.

Selling music online was a complicated challenge. Apple’s engineers needed to adapt iTunes so the music could be bought and organized easily, so charges could be recorded and billed appropriately, and so purchased tracks were encrypted to prevent buyers from copying and sharing purchased music indiscriminately. This last bit, a measure that would protect the labels from further piracy, was actually the most straightforward. Software companies had been working to address such security problems for more than a decade, and had developed all manner of digital locks and online verification tricks to protect their own software. Depending on what the label heads would eventually decide they wanted, Steve could easily customize the encryption, or watermarking, of MP3 tracks. It was much easier for Apple to tame that technology into a simple, foolproof lock than it was for the labels.

The more significant challenge facing the store’s developers was billing. This seemingly simple problem was profound—existing billing systems might have cost music purveyors more for each transaction than the profit they could earn. This was in large part because of an issue that was becoming as vexing to the industry as piracy—namely, that online buyers were showing a preference for buying individual singles rather than higher-priced albums.

Napster’s own traffic had demonstrated this new consumer behavior. When music fans could download whatever music they wanted, they liked to cherry-pick their favorite tunes, rather than get an entire album. This was a complete reversal of what happened to the music business in the late 1960s and early 1970s, when the recording industry all but did away with the single and focused instead on albums that commanded a much higher unit price. Many artists embraced the change and recorded “concept” albums, such as the Beatles’ Sgt. Pepper’s Lonely Hearts Club Band, The Who’s Tommy, or Pink Floyd’s The Wall. But labels abused the concept and regularly released albums with just one or two strong tracks, knowing that committed buyers would spend $10 to $15 on the whole album just to get those tracks.

Steve knew that there was no turning back from the “Napster effect.” Now that listeners had the option, they would nearly always choose singles over the albums padded with forgettable tracks. Steve thought singles should sell for 99 cents, which more or less represented the imputed value of a track on an album, since the average conventional CD in the 1990s had a dozen or more tracks and sold for about $15. The price also appealed to Steve’s nostalgic streak, since it was the same price that he and others our age had paid for the 45 rpm singles we’d purchased in the 1960s.

There was one problem with Steve’s idea, however. Historically, Visa and MasterCard charged 15 cents, plus around 1.5 percent of the transaction value for a single purchase; while American Express charged 20 cents plus 3.5 percent of the transaction value. That’s not such a big deal when the sale price is in the tens or hundreds of dollars, but when a single song costs just 99 cents, a transaction fee of 17 to 24 cents would be ruinous.

If Apple was going to become a significant music e-tailer, it needed to figure out how to process charges for small purchases without forcing the credit card companies to radically alter their commission structures. (Apple wasn’t the first to face this conundrum of finding an affordable way to process and pay for non-cash “microtransactions” of less than a dollar. It had befuddled just about everyone except the phone companies, who solved it by aggregating their own internal accounting and billing for customers’ individual phone calls once a month.)

Eddy Cue figured out a couple of ways to get around the problem. First, he suggested that the iTunes music store periodically bundle groups of purchases from an individual customer to send to the credit card clearing companies as a single transaction, rather than post them individually. That wouldn’t always be possible, but as the store’s traffic increased, the credit card charges could be consolidated into fewer separate transactions. Also, Cue had the store offer a simple way for parents to set up “music allowances” to prepay for their kids’ purchases, which would provide up-front payments in large enough increments to cover the cost of reconciling transactions as they trickled in later.

These kinds of intricate answers delighted Steve. When Apple took on a major project, he wasn’t just concerned with the design and marketing. He wanted to know everything about the project, and he expected his employees to attack every conceivable problem—from design and engineering to seemingly mundane tasks such as packaging and billing—with creativity. Steve told me he was just as proud of the microtransaction solution as he was of the redesigned iPod models he would introduce in conjunction with the opening of the online store.

Cue’s team made another crucial decision: Apple would build the iTunes digital “storefront” right into the iTunes application, rather than create a public website to serve as its music retail site. If you look for “www.itunes.com” online, you come to an Apple.com marketing page for iTunes, which describes its many wonders but doesn’t allow you to buy music. The only way to get to the store is via the iTunes application, which at that time was available only for Macintosh computers. This appealed to Steve for several reasons. It gave Apple control of all the technology behind the store, and it cemented a direct commercial relationship with customers. The simple transaction of buying a song, and of handing over a credit card number to Apple in order to so, became part of what Steve had begun calling “the Apple experience.” As a great marketer, Steve understood that every interaction a customer had with Apple could increase or decrease his or her respect for the company. As he put it, a corporation “could accumulate or withdraw credits” from its reputation, which is why he worked so hard to ensure that every single interaction a customer might have with Apple—from using a Mac to calling customer support to buying a single from the iTunes store and then getting billed for it—was excellent. Steve had told me back in 1998 that the only reason for companies to exist was to build products; he was now using his company to build more than just products. Apple was now creating a holistic customer experience. Everything the company did, from technology development to the design of its stores, offline and on, was in service of that customer experience. Apple’s broad-based, intense focus on this was far ahead of its time, and would have wide cultural implications. After seeing and experiencing the uniform excellence of Apple’s products and service, customers would increasingly demand the same from other companies. Apple redefined the word “quality” and forced other companies to wrestle with the higher expectations of their customers.

There was another key short-term benefit to building the iTunes store into the iTunes application: the limited reach of the iTunes store would be reassuring to the nervous music industry executives Steve had to woo. Half a million iPods had been sold, enough to create a meaningful niche but not nearly enough to affect the broader economics of the entire music industry. After all, Mac users accounted for a measly 4 percent of all personal computer users. For once, that minuscule market share was a competitive advantage. Since online sales of digital music represented a fearsome change to the label chiefs, Steve went to them with a simple, seemingly safe proposition: Why don’t you experiment with selling music downloads, to gauge demand and learn the customer and marketing dynamics, in my safe and tiny “walled garden”?