Apple Was a $5 Billion Company in Debt When Tony Fadell Built the iPod

Tony Fadell tells Lex Fridman the affair started in fifth grade, in a summer class in Grosse Pointe Farms, Michigan. It was BASIC, but not the kind with a blinking cursor: number-two pencils, a stack of punch cards, a reader that sent the program to an IBM machine in the back, and a Texas Instruments paper terminal that printed READY. One card out of order and the whole thing collapsed. The magic, he says, was that you had a language and could create a world — even if Oregon Trail was only text, and you had to imagine the rest.

The hook went deeper on an Apple II. Electronic Arts shipped Pinball Construction Set and Music Construction Set, tools that let a kid make games and scores instead of only playing them. Fadell read the sleeve notes the way people read liner notes on a record: one programmer, often writing in assembly, treated like an artist. The author of Music Construction Set was sixteen. Fadell was twelve or thirteen. If that kid could publish, what could he do?

The Apple II itself was an object of lust. It cost about $2,500 in 1981. His grandfather, who had already taught him tools, offered to match whatever he earned. Fadell caddied all summer. Once the machine was home, nobody could tear him away from it. Early units shipped with schematics. You pulled chips to add a joystick, plugged in a Mockingboard to make synth tones, poked registers, and lost a day's work if the power blinked and you had not saved. He later calls it geeks building for geeks — closer to a Raspberry Pi than to a sealed appliance.

In high school a friend founded Quality Computers, a mail-order house for Apple II software and cards; Fadell was employee two. Turnaround from coupon to box was eight to twelve weeks. Memory cards and accelerators needed patches before AppleWorks would see the extra RAM, so they wrote installers that gave customers a superpower and, not coincidentally, sold more hardware. Between twelfth grade and freshman year he wrote a hard-drive optimizer, some inner loops in assembly, the rest in an early C. He packed boxes, manned support, and learned out-of-box experience in a basement.

Music was the other lifelong obsession. Second-grade albums — Led Zeppelin, the Stones, Hendrix, Aerosmith — got played loud enough to damage his right ear, a loss he says he does not regret. He hacked a headphone jack into a clock radio so he could listen after lights-out. In college he DJed; in Silicon Valley he lugged about a thousand CDs to gigs. At Philips he worked on the Nino and Velo, Windows CE devices. The Nino, around 1996 or 1997, was the first gadget to put Audible books on a mobile device. Memory was tiny, mostly DRAM. If books fit, he thought, why not music?

MP3 arrived around 1997–98 and crushed a lossless track into something like four to eight megabytes. Flash or DRAM could hold a handful of songs; a hard drive could hold thousands. Fadell had already been trying to build that device at Fuse. Apple needed a better Walkman for iTunes. The stars aligned, and he was hired as a contractor to prove it could be made.

He laid the parts out as Lego. He bought calculators and digital cameras and pulled the LCDs so he could compare screen sizes in a row. Batteries went the same way: AA, AAA, lithium-ion, nickel-metal hydride, each weighed and timed. Storage, processors, FireWire, a better headphone jack — block diagrams and a bill of materials, not yet schematics, because it was only him. Mac drawing tools were not good enough, so he used a PC running Visio and 3D software. Then came foam models stuffed with his grandfather's fishing weights, so the thing in your hand had the right mass.

Density, he argues, is a design choice. Open a car door and you want a thunk, not a tinny ding; a pocket player cannot feel hollow. He had to convince himself before he convinced Steve Jobs. Roughly four people at Apple knew the program. As a contractor under NDA he could not recruit inside or talk outside. He also bought every MP3 player on the market and tore them apart. Six weeks, almost no sleep. Advice he was given: present three options, put the two you dislike first, and save the 1.8-inch hard drive with a small screen for last, because Jobs would shoot the others down.

The product was only half the doubt. The other half was whether Apple still had the nerve. In 2001 it was not the company that would later launch the iPhone. Fadell says it was barely break-even, a four- or five-billion-dollar business whose Mac held about 1 percent U.S. share. The company was in debt. Bill Gates had given it a loan. Michael Dell was saying, in public, that Apple should shut down and give the money back to shareholders. There was only the Mac, and the Mac was not doing well.

Jobs flipped through the slides, tossed them aside, and talked. Sony owned every audio category that mattered, and Fadell had already lived through a decade of failure; he was sure Sony would ship an MP3 player and the game would be over. Building the hardware did not scare him. Selling it did. Philips had not known how to market the Nino. Apple's marketing dollars were all on the Mac, which was the company's entire revenue. Fadell waited four weeks after the green light before he signed on.

Jobs's answer was a bet. Build this with the team, he said, and Apple would put at least two quarters of all marketing dollars on this product and nothing else. The iPod would never have happened without iTunes; Fadell is careful to say so. Jeff Robbin's small MP3 app for the Mac had been acquired and become iTunes. The team had been trying to hook it to third-party players — Korean gadgets, the MPMan — and the experiences were horrible. People should not have to burn CDs to take music with them. That is when Apple decided to build its own player, in an Apple fashion, and Fadell got the call.

How do you know an idea is worth years of your life? Fadell's book *Build* has a chapter called "Great Ideas Chase You." For the iPod the test was personal. The pain was hauling thousands of CDs. The painkiller was putting them in a pocket. The superpower was being able to hear any song you loved, anytime, without picking the stack in advance. He draws a hard line between a painkiller and a vitamin: one you need, the other is a marketing story that maybe works. Start with pain, kill it, and if you do it right you leave the customer with an emotional superpower.

Most pains get habituated away. You assume driving is just what people do until you live in a city where you walk. Fadell calls the marketing move the virus of doubt: use the right message to drag someone back to the first time the pain was visceral — remember when you did X and it felt like this — then offer the painkiller and the joy. Nest would later need the same trick for thermostats nobody thought about.

Engineers, he says, are trained to assemble cool whats. They develop, develop, develop, then call marketing to invent a story. Marketing asks why anyone needs it, writes fiction, and the product lands flat because it was never born from a why. A movie starts with a treatment: audience, characters, plot. A product should start the same way. Fadell writes the press release on day one — who it is for, which pains, which features, the price, the virus of doubt — and uses that draft as the bar. Feature creep and late cuts both get checked against it. When you are done, you know you are done.

He splits "marketing" into disciplines as distinct as mechanical and firmware. PR, social, creative, and activation sit on one side. Product marketing, or product management, is the voice of the customer in every meeting, holding the press release so the story stays non-fiction. Nobody reports to that person. They have to earn trust with engineering, design, sales, and support, and they are the hardest hire in the building. On communications, Fadell agrees with Fridman's worry that caution can sand off the magic. Conservatism, he says, starts with the leader. A timid storyteller produces an even more timid PR shop, because those people need to keep their jobs.

Great design, Fadell says, is not a mountain from which tablets descend. If the team has the why, the work becomes a mission, and people bring their own pixels, audio clicks, and mechanical curves. Love goes in; the customer feels love on the other side. He does not treat design and engineering as separate priesthoods. There are electrical designers, AI designers, data-scientist designers. The engineers he wants are artists — often people who came from music — not the rigid ones who say this is the only way it is done.

On something revolutionary, none of you are experts, because the thing has never existed. An expert mindset is exactly what he does not want. Early iPods had ugly risks: spinning media in a pocket that could drop and slap the heads dead. Another was the IDE bus. The device had to be a FireWire hard drive sometimes and an MP3 player sometimes, hot-switching so the drive thought it was still talking to the same host. Fadell prototyped it and let it run for days. Then he invited Apple's mass-storage specialist, drew the command sequence on a whiteboard, and asked for holes. The expert said it would never work, stormed out, and never looked at the prototype.

Fridman laughs because he has lived the same scene. An acoustician walked into his studio, recited reflections and curtains, and priced a hundred-thousand-dollar rebuild. Fridman wanted to know if there were major holes and cheap fixes, not whether the room failed mastering-studio spec. Both stories, Fadell says, are experts who are excellent at *what* — specs, feeds, numbers — and useless at *why*. You can bring in why-experts. Too often teams hire what-experts, who then close the door the V1 most needs open.

Working with Steve Jobs, Fadell says, meant being pushed on details for the right reasons. It was not bullying. "He would critique the work, not judge the person," at least not in front of the group. On a revolutionary V1, almost every call is an opinion held by one, two, or three people. There is no data to fall back on. You still have to tell the team the why, the trade-offs, and why this answer won, or they cannot feel part of it. Most people want a data-driven decision. Consumer V1s do not offer one.

Companies that cannot ship new things, in his experience at Philips and elsewhere, try to turn opinion decisions into data decisions so nobody loses a job. Management consulting, as he describes it, is a machine for that: hand the opinion to someone else, get a slide deck back, and blame the consultants if it fails. Version one of the iPod was not perfect, nor was version one of the iPhone. They got opinions wrong, then used V2 data to modulate — while still adding new opinions, because those are the differentiators.

Two iPhone opinions still sit side by side in his memory. The first was the keyboard. BlackBerry was the productivity device of its time, nicknamed CrackBerry because people could not put it down; the data said hardware keys. Jobs's view was that this device had to show full-screen video, a full-screen browser, and Apple's own apps. A permanent keyboard would steal half the face. There was no 2007 dataset for that. It was a different kind of product, born from entertainment as much as messaging, and the keyboard had to get out of the way.

The second opinion flipped. Jobs wanted no SIM slot — a pure object — and Jony Ive agreed. The team said that would not work. Jobs's counter was the Jobs move: then why do some Verizon phones have no slot? Product marketing, research, and engineering came back days later with how many networks still required SIMs and which way the trend was going. Data killed the purity. The slot shipped. Fadell notes the irony: Apple had created Wi-Fi in 2001 when nobody had it, and eSIM would later make the physical slot legacy. Timing is the whole game. You cannot know, in the room, whether the opinion that lost would have changed the industry.

Fadell's chapter on jerks has one split: ego-driven or mission-driven. The first kind climbs, shoves, and makes it about themselves, aggressively or by cold shoulder. The second kind is unrelenting on the details that touch the customer, not on every detail. Being pushed by that person feels like an attack because they are shoving you past a boundary you were protecting. Jobs, Fadell says, often did not know all the details. He could see on the team's faces that they thought, we could probably do it, but the risk is ugly. Then he would say, list the risks. Then: I am willing to take those risks.

The plastic-to-glass fight is the cleanest example. The team had already argued glass versus plastic and landed on plastic: phones drop, glass breaks, a daily driver should not be fragile. The 2007 Macworld "Jesus phone" under glass was plastic; reviewers who knew better asked, through the museum case, whether that front was plastic. Jobs changed the frame. If keys and coins hairline-scratch a plastic face in a pocket, that is a design failure — the company's fault. If the user drops it and it cracks, that is the customer's doing, and they complain less. The room moved. Corning and the mechanical teams had until June. He was unrelenting, Fadell says, but he was never unchallengeable. You needed a team and data, not a lone speech.

Heat is allowed. Labels are not. Fadell will raise his voice. He will not say "you are the dumbest human on earth," and he will not let anyone else say it, even as a joke between people who can take it. Judging the person adds nothing to the work, and people without thick shields copy whatever the leader lets stand. If you blow it, you apologize in front of the team. That humility is culture too. The difference he felt from Jobs as a leader was technical depth: Fadell came from hardware and systems and grew into product marketing; Jobs came from storytelling and the customer's voice and never went deep on the bits. After Jobs demanded a "beautiful" Macintosh PCB that did not work, he learned not to micromanage traces.

The habit Fadell reversed on purpose was credit. Bring Jobs a good idea and you often got a hmm, or silence. Twenty-four hours later he would return with a slight tweak and announce a genius idea of his own. The table learned to roll its eyes. It trains people to stop bringing things. Fadell now calls out the person in the room: that is a great idea, let's build on it, or it is great but not for now. He still says when an idea is bad. The pendulum has to swing both ways. If it only swings to pain, the team goes quiet.

By the mid-2000s the iPod was about 60 percent of Apple's revenue and held something like 85 percent of its category. Jobs stood in front of employees and said, with pride, that the company was out of debt. Then the feature-phone industry started adding cameras, color screens, and storage. Hundreds of millions of phones a year could swallow the iPod's job. Apple's first answer was partnership: music from the iTunes Store on someone else's hardware. That became the Motorola ROKR, a feature phone that held about ninety-nine songs. Two cultures, weak engineering on the other side, a contractual obligation. Fadell watched it land on Jeff Robbin's face. On stage Jobs spent a minute or two and, Fadell says, threw the phone aside as fast as he could.

If you cannot partner, you have to become one of them to save the thing that pulled Apple out of fifteen years of malaise. Fadell's group built an iPod-plus-phone, click wheel and all, audio through a wired headset. A second project was a full-screen iPod with a virtual click wheel, so iTunes could sell music videos and movies. A third, in another team, was multi-touch meant for a Mac tablet — except there was no tablet software ecosystem, only blown-up Mac apps. Each project had a fatal hole: the wheel could not do data entry, the full-screen iPod had no radio, the tablet had no apps. Welded together, they covered each other.

The OS was, in Fadell's word, Frankenstein — meant as praise. Low-level drivers, manufacturing, and communications came from the iPod. Mac OS was hacked down, new code shoved in; it was not "Mac OS changed a little." Other smartphones of the era obsessed over running Word and Excel. This one was born from entertainment and did not care about that compatibility. It was a new kind of computer.

The keynote everyone remembers — an iPod, a phone, an internet communicator, then the reveal that they are one device — is the form every marketing textbook calls lazy. Jobs had been pitching that story internally from day one, watching faces, trying it on a few trusted outsiders, modulating. By the time he walked on stage it was a play rehearsed for years. He was showing the pain of carrying a laptop, an iPod, and a phone, then offering one object. Showmanship sat on top of a non-fiction story that had been worked until it was true.

Fadell and Jony Ive fought in the distant past; looking back he talks about what he learned by watching Ive's studio. That team lived in materials, form, color, curves, and shadows — not only how an object felt in the hand but how it would look in a marketing photograph. The process was a funnel: open wide, refine, mix opinion with data about what could actually ship in volume. At Nest, with far less money, he still made the mechanical team change curves on the Protect smoke and CO alarm after it was tooled, and he put a custom screwdriver in the box, because those "extraneous" objects are part of the experience. Skip the pain and the customer feels the missing love.

The hard part of Nest was not the thermostat. It was making anyone care. Nobody chose the thing on the wall, nobody installed it, most people could not program it, and they hid it in a hallway and paid the bill. The virus of doubt had to wake that habituated shrug. The worse problem was the channel. In the U.S. the category belonged to the installer: no price transparency, a man in the house naming a $350 unit for a rich-looking home and a $100 unit for the next, kickbacks and Hawaii trips locking the trade, contracts that kept other brands out of Home Depot and Lowe's.

So Nest had to build a disruptive go-to-market as big as the product: sell direct, make self-install something early adopters would actually finish — otherwise the installer would talk them out of it — and fight into retail so later buyers could touch one. Any of those failing, the device never reached a home. It became a hit. Google bought the company for $3.2 billion. Fadell says they considered saying no the whole way. They spent four months with Google on brand, team, IP, and budgets *before* talking price; Google was already an investor, so the range was known. He wanted a home platform that would take three or four years and hundreds of millions. Investors would not fund that. He had watched startups announce a platform, then Apple, Google, Amazon, Microsoft, Samsung each say they were building one too. Developers flee. He wanted to change the landscape before the landscape changed on him.

Can a startup still take on the giants? Absolutely, he says — not on their current turf, on the turf they will want next. Spotify began as an app and became a platform. WeChat began as chat and became wallet, delivery, travel. Google was search; Apple was personal computing, then a phone, then a platform once the App Store existed. "There's no such thing as building a platform company. You build a great app first, and then you can expand it and have the right to become a platform."

Career advice in *Build* collapses to two questions: what do you want to learn, and who do you want to learn it from? People pick a graduate program for a lab; they pick a job for a logo. Fadell's first interview question is never about expertise. It is "What do you want to learn?" Anyone who arrives as the expert in a field that does not exist yet is the wrong hire, because the whole team is not expert either. He talks about versions of himself the way he talks about products: V1 opinions, data from doing, V2 opinions. He took the lowest job at General Magic to be near the right people, barely above the poverty line, eighty-hour weeks. You find what you want by finding what you do not want, and the only way is to do a lot of stuff.

Founding a company, he says, starts with belief in yourself — checked by mentors so you are not a lone mad person. He used to dive into every idea. He now runs away. A great idea stalks you. You set it aside because of some obstacle; a few days later you see a fix, or a conversation circles back. Eventually it becomes a black hole. You cannot think about anything else. That is how you know. Mentors are not a directory you search. They grow out of a network of non-transactional relationships, and they are not all older. His best ones knew nothing about technology as he knew it; they knew human nature, and they pulled him toward empathy.

Money is toxic when the relationship is only a transaction: a VC watching the limited partners, ready to walk if the exit is not two years away, offering no mentorship or network. Smart money is a partner who is on the mission. Do not pick a firm for the brand; pick the partner you will sit with. Lawyers, similarly, have a why. Even mobsters had them. The useful ones work in the gray of interpretation, change a word in an ad, and open a path. The ones who only say no, by the hour, are the PR problem in another costume. Fadell's rule of thumb: if you do not want a no, do not ask — you will still be billed.

Fridman quotes Bukowski — find what you love and let it kill you — and asks about balance. When the work is a mission that changes other lives, Fadell says, it is not a job. It still needs a boundary. At General Magic he was physically, mentally, and socially unhealthy; he put every waking minute into a company that was building the iPhone fifteen years too early. When it collapsed he had nothing left. Balance, for him, is not a three-day week. It is making sure the thinking happens at work, not at 3 a.m. Write the open loops into a page or a tool so the brain can empty. Most exhaustion is rumination — having to remember everything instead of doing the work.

The "Russian question," as Fridman calls it: darkest moments. First, twelve schools in fifteen years. Always the new kid, always the geek with the computer in the 1980s, punished for asking the question that made the class do more work. Latchkey, parents at jobs, a skateboarder brother outside, Fadell alone with the machine. Phone calls were two dollars a minute; there was no internet. He used bulletin boards and hacked MCI and Sprint codes to reach other geeks long distance. Second darkness: the end of General Magic. He had torn up his social ties for the high. He had to get clean.

He is not afraid of death. He is an atheist who still believes in a soul, not as something floating in space but as what you instill in people you worked with. It lives in the stories they tell. The week of the recording, Apple killed the iPod after twenty-one years. Products are vessels, he says, and vessels go away; the love people felt through them does not. Bill Campbell is gone. Jobs is gone. They made him better, and that is the soul he believes in. Why are we here? Because we are naturally curious, and we are always trying to ascribe meaning. That is it. Evolution baked it in. Along the way we make some cool things.