Part IThe Story
Sometime in 1962, a twelve-year-old boy from Seattle walked into a bank in New York City carrying a check for roughly $20 million. The money belonged to his father. It was what J. Elroy McCaw had netted from selling WINS, the New York radio station that had become one of the first in the country to play rock and roll, the station where Alan Freed made himself famous talking over records nobody's parents wanted to hear. The errand was a joke. Elroy was, by every account that survives, an inveterate practical joker, and the image has a joker's logic: the clerk at the window, the small hands on the counter, the gap between the courier and what he was carrying. The record does not say which bank it was. It does not say what the boy wore, or whether he understood the number on the paper, or whether anyone at the teller's cage believed him. It says only that his father sent him in.
Strictly speaking, what the boy was carrying was air. A radio station is a building, a tower, some turntables and a staff. What a buyer pays for is the license to push a signal through a slice of the electromagnetic spectrum over a city of eight million people. The check was the conversion of something invisible into something you could deposit. Craig McCaw would spend the rest of his working life repeating that conversion at larger and larger scale, with more borrowed money and less sleep, until people began to describe him with the same phrase his father's career had suggested: money from thin air.
Seven years after the bank, in August 1969, on a cool, cloudy day, with his mother traveling abroad, the same boy, now nineteen, walked into his father's bedroom in the old Boeing mansion in The Highlands north of Seattle and found Elroy McCaw dead of a massive stroke. He was fifty-seven. He left several radio and television stations, investments in 54 companies, money in 25 bank accounts, a Lear jet and a yacht. Within months he also left creditors. Claims reached $12 million, the bank handling the estate withdrew and pronounced it insolvent, and the mansion (Spanish-style, five wooded acres, eight fireplaces, a ballroom where the boys had once worn coats and ties to a formal dance) was sold along with the yacht and nearly everything else. One small thing was left, partly because it had been placed in trust and partly because almost no one knew it existed: a cable television system in Centralia, Washington, the town where Craig had been born.
The next half century followed from those two rooms, the bank lobby in Manhattan and the bedroom in The Highlands. One taught the boy that air could be sold. The other taught him what borrowed money does when the man who borrowed it is gone.
By the Numbers
The McCaw Ledger
$4.50Early price per 'pop' for cellular licenses McCaw began buying in the early 1980s
900,000AT&T's forecast of total U.S. cellular subscribers by 2000, the document that started it
$54MMcCaw's 1990 pay, making him America's highest-paid chief executive
2 millionCellular One subscribers when AT&T bought a third of the company in 1992
$11.5BAT&T's price for McCaw Cellular, then the second-largest merger in U.S. history
840Satellites in Teledesic's original 1994 plan for an 'Internet in the sky'
$35MPrice McCaw paid in 2012 for a 1962 Ferrari 250 GTO, then a world record
A Wire Through the Sewer
To understand the son you have to start with the father's quarrel, which began in Aberdeen, Washington, on Grays Harbor, in the late 1920s, when Elroy McCaw was a student at Weatherwax High School and decided the telephone company charged too much.
Local oral histories, gathered decades later for a family documentary, tell it this way. Elroy and his friends strung their own wire from the McCaw house to the high school and from there into downtown Aberdeen, down alleys, along the sides of buildings, and through the sewer. The sewer was the clever part. They tied string to ping-pong balls and flushed them down toilets. When a ball floated past a manhole, someone grabbed the string and used it to pull wire through the pipe. The result was a private telephone network, a small kingdom of copper. Elroy began piping music to downtown businesses, then made a deal with Fred Goddard, who managed KXRO, Aberdeen's only radio station, to carry remote broadcasts over his lines for much less than the phone company asked. The phone company got annoyed and had the whole thing shut down. The people who told the story said that was the start of a lifelong grudge Elroy McCaw held against the telephone monopoly.
The rest of Elroy's career reads like that grudge carried forward. He bought a radio station in Centralia in 1937, added KYA in San Francisco, and took over the New York station he converted to rock and roll. He bought and sold stations the way other men traded horses. He went into television, and then into cable in the 1960s, early enough to be involved in an effort to bring cable to New York City. He put his four sons (Bruce, Craig, John, Keith) to work as linemen and door-to-door subscription salesmen. He also, as one later synopsis of his son's life put it, ran a multimillion-dollar business "out of his hat."
That phrase matters. Elroy was a visionary in the way frontier men often are: early, restless, contemptuous of incumbents, and organizationally solitary. The empire lived in his head and in the leverage he kept rolling over. When the head stopped, the leverage came due. Marion McCaw spent roughly the next eight years paying back what her husband had borrowed. By the time she finished, a family that had employed its own cook and a monthly household budget of $7,000 (tuition, servants, club memberships, trips to Europe, flying lessons for the boys) had kept one asset that mattered. There was also a $2 million life insurance policy, which the creditors could not reach.
Almost a quarter century later, standing in New York beside the chairman of AT&T on the day he agreed to sell his company to the most famous telephone monopoly in the world, Craig McCaw summarized his father in one sentence. It is the most revealing thing he ever said about himself, and it is not about himself.
"My father was a visionary who did not hire great people," he said. "The company was too dependent on one person."
Don't Sell
Twin City Cablevision, in Centralia, about ninety miles south of Seattle in a town of roughly 12,000, was not much of an inheritance. In the version O. Casey Corr of The Seattle Times pieced together for a five-part 1993 series, and later for his biography Money from Thin Air, it had 10 employees, served 4,000 homes, and grossed about $8,000 a month. Other accounts put the subscriber count at 2,000. Some say Elroy sold the system to his sons in 1966, when Craig was sixteen. Corr says it was a gift Marion didn't know about. The small company's origins are as murky as the father who arranged them, which seems fitting.
In 1970 The Seattle Times Company offered to buy it. The four brothers met at the mansion, which had not yet been sold, to talk it over. Each stood to clear about $180,000. Bruce, the eldest, was out of college running the family's aircraft maintenance shop, Jet Air, at Paine Field for $400 a month. John and Keith were at Lakeside, the private school their brother had also attended. Craig was at Stanford, majoring in history, and he had spent summers in Centralia doing odd jobs.
"I said don't sell," he told Corr two decades later. "I had worked there . . . that was the only thing I'd ever done in my life, my only job."
There is no romance in that sentence, and also no calculation. It doesn't mention cash flow, the future of cable, or rebuilding the family name, though later accounts would supply all three. It is a twenty-year-old refusing to let go of the one place where he had been useful. His brothers agreed. Craig wanted to leave Stanford and run the system himself, and Marion talked him into finishing his degree and managing it from a distance through a hired manager. In 1971 she married John L. Scott, a Scottish-born real estate broker about eighteen years older than she was, and moved into his house in Bellevue. She bought a smaller ranch house with a pool in The Highlands for the two youngest boys, who were judged too independent to live under Scott's roof. They held parties there that became family legend.
Craig graduated in 1973 and took over. By 1974 he was president and chief executive of McCaw Communications Companies, and the company was already in radio common carrier paging, a business of beepers and dispatch that turned out to be a rehearsal. In June 1975 a senior partner at a Seattle law firm stopped by the office of a young associate named Wayne Perry, looking somewhat apologetic.
"There's this crazy kid down in the conference room whose got this nutty idea," the partner said. "Go down and spend some time with him, but don't take too much time. You're too busy. But don't piss him off because his family's been a client of the firm for years."
Perry was about the same age as the kid. The nutty idea was to use one cable system as leverage to finance the purchase of the next, then use that one to buy the one after, a structure built on tax write-offs and years of planned losses that would look reckless to anyone not running the long arithmetic. Perry thought it was brilliant. He went to work full time for McCaw and later became vice chairman of the company. He became the first of many professionals who walked into a room expecting to humor a rich family's son and walked out as part of his plan.
Perry also left the most precise description of how McCaw's judgment worked and why it frightened people. Inside the company, he said, they talked about needing to be "90 percent" certain before committing to an investment or strategy.
— Wayne Perry, former vice chairman, McCaw Cellular, to The New York Times (2000)But Craig's feeling of 90 percent was often Craig alone. It might be 40 percent to someone else.
By the early 1980s the crazy kid had about 30,000 cable subscribers in Washington and Alaska, places that wanted television and that bigger operators found too remote to bother with. To keep buying he sold a $12 million stake in the company to Affiliated Publications, owner of The Boston Globe, and kept picking up small systems at low prices. McCaw Cablevision eventually became the twentieth-largest cable carrier in the country. It was a good business. He seems to have concluded fairly early that it would never be a great one.
Four Dollars and Fifty Cents a Person
In 1981 McCaw came across a document that AT&T had filed with the Federal Communications Commission. It was a forecast. By the turn of the century, the telephone company's analysts projected, there would be 900,000 cellular telephone subscribers in the United States.
AT&T was dismissing something it had invented. Bell Laboratories engineers had worked out the basic idea of cellular service in 1947 to solve a police radio problem: transmitters that could pass a call from one patrol car's area to the next as the car moved across a city. For thirty-five years the technology sat in the regulatory pipeline while the FCC studied it, the handsets shrank from backpacks to bricks, and a car phone in 1979 cost about $2,500 and was treated as a toy for the rich. Now the monopoly that held the patents was telling regulators, in effect, that the market would be small enough to stay comfortable.
McCaw read the same number and reached the opposite conclusion. He was not betting that AT&T's forecast was right. He was betting that it was wrong in a specific way, that 900,000 was a ceiling drawn by people with no reason to want it higher. Then he looked at what the licenses cost. In the early market a cellular franchise could be had for about $4.50 per "pop," industry shorthand for each person living in the license area. At that price you weren't buying a business, you were buying an option on a future that the dominant player had publicly written down.
"I decided that cellular communications was an extraordinary opportunity," he said years later. "Franchise rights for the entire country were being offered virtually at once. Our strategy was to purchase as much cellular as we could get. People did not realize how valuable it was."
In April 1982, after fourteen years of study, the FCC said it would accept applications over the next ninety days to provide cellular service in the nation's thirty largest cities. Each market would get two licenses. One went to the local wireline telephone company, the "B side." The other was open to independents, the "A side." McCaw was prepared. Of the first six cities where his company applied, Seattle among them, it won all six. As applications multiplied, the FCC switched to lotteries for later markets. Ordinary Americans, dentists and secretaries and investor syndicates, won franchises worth fortunes and often had no idea what to do with them. McCaw entered the lotteries too, but his real skill was in the second step: finding winners and buying their licenses before they understood what they held. Wireless Nation, the industry memoir by James B. Murray, an early broker of cellular licenses, describes this as a frenzy, a land rush staged in hotel conference rooms and on legal pads.
How he financed it set the pattern for the rest of his life. He persuaded bankers that a license bought at $4.50 a pop was worth far more, and he used AT&T's own data to model future cash flows. He then used the licenses as collateral to borrow money and buy more licenses. He borrowed against the cable company and sold shares in it. Through Michael Milken and Drexel Burnham Lambert he did nine bond offerings and stock issuances by 1989. It was his father's leverage with a crucial difference: the borrowing financed a coherent map instead of a scattered portfolio, and a growing roster of people besides Craig McCaw understood where the map was heading. The company had about 750 employees in 1986. It would have nearly 10,000 by 1993.
In 1986 he bought MCI's cellular and paging operations for $122 million and renamed the company McCaw Cellular Communications. In 1987 he sold the cable business to Cooke Cablevision for $755 million (Forbes later said $790 million, and reported that he personally cleared about $250 million at age thirty-seven). The cable systems, built up from the one small company his father had left, went to fund the bet that would replace them.
That same year the company went public. It raised roughly $2.3 billion, according to Corr's reporting, and the family's holdings were suddenly worth more than $1 billion. The offering closed within a week of the eighteenth anniversary of Elroy McCaw's death.
Craig McCaw called it "the worst day in the history of our company."
He meant it literally. For years McCaw Cellular had been buying licenses quietly, ahead of anyone who understood what they were worth. The prospectus put everything on the table: licenses covering 37 million people in 94 markets, twice the nearest competitor, with every gap visible. Anyone who could read now knew what he had, what he wanted, and what it would cost to stop him. He had been made rich and made visible on the same day, and he considered the second worse than the first was good.
Who Are Those Guys?
One of Craig McCaw's favorite movies is Butch Cassidy and the Sundance Kid. In 1992, Corr noted that the mustache McCaw wore then (bushy, drooping, out of place on such a slight man) resembled the ones the outlaws wear in the film. The odd part is whom McCaw identified with. Not Newman, not Redford. He liked the posse, the faceless manhunters who keep coming across the badlands without explanation until the two charming bandits, exhausted and baffled, ask the film's most famous question: Who are those guys?
That was the company he wanted to run, one that made deals so fast and so well, with so little personal fanfare, that competitors could only turn around and ask.
The 1987 prospectus had shown one obvious weakness: McCaw had neither New York nor Los Angeles. Most of both, plus Philadelphia, Dallas and Houston, sat inside LIN Broadcasting, a New York company that owned television stations and controlled valuable cellular licenses, including a 35 percent stake in Los Angeles. LIN was run by Donald Pels, a well-liked executive and an investor in The New Yorker, who had led the company since 1969 and taken its share price from $3.375 in 1981 to more than $100. He was good at his job, proud of it, and liked running the place. McCaw's executives thought LIN's cellular service was poor. McCaw's larger concern was strategic: if LIN's licenses ended up with a Baby Bell, his plan for an independent national network would be finished.
Starting in 1988, helped by $1.5 billion from British Telecom for 22 percent of McCaw, the company began buying pieces of LIN. Craig had lunch with Pels and talked about cooperation. Pels pictured a multimedia combination of LIN, McCaw and Affiliated Publications, with himself in charge. He apparently didn't know that McCaw insisted on being in charge of every alliance he entered.
The task of managing Pels went to John McCaw, younger than Craig and in their mother's estimation the warmest of the four boys. John looked the most like Craig: thin and delicate, with a habit of staring a long time before answering and eyes that darted when he was nervous, a trait the brothers shared. He had started at the family cable system in Centralia in 1975 and become executive vice president for acquisitions, though his title inside the company was "relationship guy." He met Pels ten or fifteen times at Pels's homes in Manhattan and on Long Island, and the two men became friends.
Then Pels publicly floated splitting LIN's broadcast and cellular properties into separate units. That changed everything. Regulations barred Baby Bells from owning broadcast properties, so as long as LIN stayed whole, the Bells couldn't buy it. A spinoff would signal that Bell money was welcome. McCaw had to move, with or without Pels.
John flew to Paris on the Concorde. Pels was staying at the George V. When Pels spotted John in the lobby he smiled and said to his wife, "There's John." Then, in John's telling to Corr, Pels's face went white as he understood why his friend had crossed an ocean.
On June 6, 1989, McCaw Cellular announced a $5.8 billion tender offer for the 91 percent of LIN it didn't already own. BellSouth bid against it, and McCaw won. The final price for control was about $3.5 billion, roughly $350 per pop, nearly eighty times what the first licenses had cost. It was the largest deal in McCaw's history, it nearly broke the company, and it bought the two cities that made a national network possible. Skeptics were stunned by the debt. McCaw and Milken were betting that the network was worth more than the sum of the loans.
During those same years McCaw said something at a company retreat that followed him for the rest of his career. In 1988, on the Snake River, with his top executives assembled, he told them, "I want to be the Wizard of Oz." The line got into print and was read as megalomania: the voice behind the curtain, the man who wants to rule the Emerald City. He spent years correcting it.
— Craig McCaw, to The Seattle Times (1990)I did make that comment, but not in the sense that I seek visibility or the trappings of power. Just the opposite. What I really meant is that I want to make an impact without being particularly visible.
In 1989 the Los Angeles Times noted that recent editions of Who's Who in the West and Who's Who of Emerging Leaders in America listed a son of the late J. Elroy McCaw. It was Bruce, the insurance executive, not Craig, the man who had just upended the Baby Bells. Craig declined to be interviewed or photographed for that story. His picture had never appeared in his company's annual report. He had asked people not to discuss his family with reporters. His 1988 chairman's letter offered one line that serves as either a creed or a disguise: "We approach our work with humility and determination to achieve excellence."
The posse, in other words, wore no badges.
The Six-Foot Cord
Before about 1990 a cell phone was local in a way that's hard to imagine now. Each tower connected to local signaling, local billing, and local land lines. Only towers near a customer's home had the routing information to reach that customer's phone. If you drove from Seattle to Portland, your phone worked, but calls to your number couldn't find you. The network was a collection of separate towns.
This is where the dealmaking turns into something else. The licenses McCaw had collected weren't meant to sit as a portfolio. His holdings in the major markets gave him leverage to trade, swap and buy licenses in smaller markets that others thought unprofitable, usually at a steep discount, so that his territories formed continuous clusters. When the Baby Bells came looking for licenses, they often found McCaw already holding enough of a market to shut them out unless they bought his spare licenses at large markups. The map was the asset, not any one license.
In 1990 McCaw Cellular rolled out SS7 signaling across its network. The technical term, non-facility associated signaling, meant something simple to customers: routing and dialing information could now move across the whole system, so a phone registered in one market could be found in another. The towns became one network. McCaw called it Cellular One, built partly through a partnership with Southwestern Bell and Vanguard Cellular called the Cellular One Group, and introduced roaming charges. It was the first truly national cellular system in the country. That year McCaw was paid $54 million, more than any other chief executive in America.
The money is not the most interesting fact about 1990, though. What matters is what he thought he was undoing. In an interview for Corr's book he described the wired telephone system with a disgust that ran deeper than business.
— Craig McCaw, quoted in O. Casey Corr, Money from Thin Air (2000)With cellular telephony... we saw an enormous gap between what was and what should be. I mean, [the fixed phone system] makes absolutely no sense. It is machines dominating human beings. The idea that people went to a small cubicle, a six-by-ten office, and sat there all day at the end of a six-foot cord, was anathema to me.
Set that next to the boy in Aberdeen flushing ping-pong balls through the sewer to pull a wire the phone company would later tear out. The father fought the monopoly by building his own wire. The son decided wire was the problem. In the 1993 announcement of the AT&T sale he put the same idea in grander terms, saying the dream had been about "uplifting the human being by putting in their hands a more powerful communications tool." It sounds like a press release, and it partly is, but the six-foot cord quote doesn't. It sounds like someone who once sat at a desk and couldn't stand it.
He extended the same impatience to the customer's wallet. In 1987 McCaw Cellular came up with a prepaid product it called GoPhone, for people who didn't want or couldn't get a contract. The name outlived the company that created it, outlived AT&T Wireless, outlived Cingular, and stayed on AT&T's prepaid service until 2017. In 1991 2.5 million cellular phones went into service in the United States, compared with 1.9 million new residential lines. The average cellular bill was $80 to $90 a month, three or four times standard home service, and people paid it anyway. By the spring of 1993 more than 11 million Americans had cell phones and 9,500 new customers were signing up every day. McCaw had about 2 million of them, and his company was the only one with anything close to national reach.
AT&T had forecast 900,000 subscribers for the whole country by 2000. One of its future subsidiaries passed that number by itself years early.
The Man Behind the Curtain
In person he didn't look like what he was. In 1990 a Seattle Times reporter found him in pastels, an Izod-style shirt and loose pants, slightly built, with a soft voice that suggested fragility. Only his eyes, "sharp and confrontational," hinted at the company's appetite. In 1992 Corr described a lean man kayaking alone on Lake Washington on misty mornings. A decade later Forbes still called him "reed-thin." He was dyslexic, and he talked about it more openly than about almost anything else, explaining it in the language of strategy.
"A dyslexic tends to be more conceptual and do things which other people wouldn't see as obvious," he said, in remarks later collected by Tren Griffin, who worked with him for years. "So maybe it's a strategic asset . . . I can't go to a piece of paper and organize things as most people would in a way that they could understand and come up with a plan. I have to explain conceptually what we want to accomplish, and then somebody else has to translate that into a concise organized plan." He also said: "I don't think like other people, so I don't fit very well in a clique. As a result of that I have trouble quantifying people as directly as others. I look at their ideas, rather than at them so much as individuals."
That explains the company he built. McCaw Cellular didn't run on one man's orders, it ran on one man's concepts turned into plans by people chosen because they could do what he couldn't. Unlike his father, he hired strong executives and gave them real authority. "If you pass autonomy as far down in any grouping of people as you can," he said, "you will get extraordinary results if you ask for a lot. The greatest burden you can put on someone is trust."
Those people deserve the same compressed treatment he gave his own father. Tom Alberg spent twenty years in securities law after Columbia Law School before joining McCaw as an executive vice president in 1990, a careful lawyer who joined a company piling up what one account called suicidal levels of debt. John Stanton was McCaw's vice chairman, the executive who said Craig had seen cellular's potential by the late 1970s, and who later built General Cellular. Steve Hooper joined in 1982, just as the license buying started, and worked his way through finance and regional operations to headquarters in Kirkland. James Barksdale came in 1992 after nine years as chief operating officer of Federal Express and imposed a rule McCaw could never have imposed on himself: "If you can't put it on one piece of paper, you can't execute." Each was brought in to complete a man who described himself as incomplete.
The brothers were there too, always deferring. Bruce, the eldest, was the only one who built businesses on his own: Westar Insurance, which became Forbes Westar, and a hand in founding Horizon Airlines. He was a pilot, a big-game hunter, a car collector who drove an old 1986 Mercedes 300E in the One Lap of America, and the owner of a riverfront cabin in Idaho's Frank Church Wilderness, once owned by the supermarket founder Joe Albertson, with no running water and reachable only by water or air. "He's never run the business as if he's the only guy in the saddle," Bruce said of Craig, which is generous and also an admission of who held the reins. Keith, the youngest, never served on the board and later invested in ski resorts. He died in 2002. John left full-time work in 1991 "to get away for a while and get a perspective on things." Craig's comment on his brother's departure is cool: "we simply agreed that if he wasn't going to put his heart and soul in" — and Corr's excerpt breaks off there, so we don't get the rest.
The family complained about the press. "One of the reasons why all of us spend less time in the city," Bruce said, "is because we got run out."
The people around Craig McCaw kept trying, mostly without success, to find the person under the persona. Gardiner Davis had taught him at Lakeside and later worked for the company as a writing teacher. He liked and admired his former student. He also said Craig put on a mask when meeting people. "I'd sort of like to meet the real person and find out what he really thinks." The playful side came out sideways. He kept toys in his office. During a rash of thefts in the building, he hung a fake security camera. He enjoyed food fights and didn't mind being hit. Corr reports that someone named Allen once threw a dinner roll at him; the context suggests it was Robert Allen, chairman of AT&T, though the excerpt doesn't say so explicitly, and it's pleasant to think that a dinner roll thrown across a table might be the most candid surviving record of the biggest negotiation in wireless history.
"The fact is, I'm an introvert," he told Forbes in 2000. A decade earlier he had used a different image: "If I had my way I'd be like the little scientist in the lab with his beakers, happily pouring one into another."
The image is gentle, and it is also accurate in a way he may not have intended. That is what his deals were: transfers of value from one container to the next, cable to cellular, cellular to AT&T stock, AT&T stock to the sky.
Selling the Child
By 1992 McCaw Cellular was generating $1.75 billion in annual revenue and had more than $5 billion in debt. The LIN acquisition had used up its financial flexibility, drained its cash and cut its net income. Craig McCaw controlled more than 60 percent of the voting stock, which meant he had the final say on everything. In every previous deal, with every lender and every partner, there had been one non-negotiable condition: Craig McCaw stays in charge. His family's trust held most of the shares and had always deferred to him.
Robert Allen of AT&T, a quiet man whose temperament matched McCaw's, approached him that year. AT&T sold cellular phones but offered no cellular service. McCaw offered service and sold no phones. AT&T had invented the technology in 1947, forecast it into irrelevance in the early 1980s, and was now going to have to pay to get back in.
On November 4, 1992, the two companies announced that AT&T would pay $3.8 billion for a third of McCaw Cellular, $45 a share against a recent price of $26.75. The next morning the stock jumped $6.50 to $33.25 at the open, and a million shares traded in the first ten minutes. Hidden in the terms was something new. The McCaws sold AT&T an option to buy control of the company within seven years. The option cost $100 million. Exercising it would cost another $600 million plus interest.
McCaw called it a hard personal decision, reached after soul-searching. He said he had put the company's interests above his own preferences. If McCaw Cellular was run well, he argued, AT&T "would not want to fool with a good thing." He framed it as motivation: "It's an additional carrot for us to be as good as we can be." He gathered a few hundred employees to calm fears of being absorbed into a company with 345,000 workers.
"You all know, this deal includes the potential that they can fire me," he told them. "They can't fire you, but they can fire me." They needed to work hard, he said, and Corr noted the smile that came over his face as he finished, "because I never want to be fired."
Tom Alberg, who negotiated much of the deal, was less sentimental. "To be realistic, you have to assume they will exercise the option." Maybe not, he added, given the debt AT&T would have to take on. "You can argue either way."
For nine months the deal went nowhere. Teams of lawyers couldn't work out how to divide revenue, how to handle overlapping markets, whether McCaw could use AT&T long-distance discounts to undercut AT&T's own sales force, or how to structure a 99-year lease of the AT&T name. Barksdale said it couldn't work as written; he'd need a lawyer in every office to explain what was allowed. His instruction to everyone meeting with AT&T: find something he could understand.
In August 1993, on a Friday, McCaw got on a plane to New York unsure whether the biggest deal of his life was about to fall apart. His communications chief, Bob Ratliffe, had already gone ahead with two press plans: one praising a revived AT&T partnership and one celebrating independence. Allen flew in the same day from a week in China, where he had seen how cheaply cellular networks could leapfrog old wired systems in a country with few phones, and he came back convinced he needed the largest cellular company in America. On Saturday morning the two quiet men met, mostly alone, for four hours. The McCaw board (16 people, including John and Bruce McCaw, former U.S. Senator Dan Evans, QFC chief executive Stuart Sloan, John Stanton, and John Giuggio from the Globe's ownership) flew in from Seattle, Boston and elsewhere. By 7 p.m. McCaw and Allen had shaken hands in a hallway of the Waldorf Astoria.
"We have reached agreement," McCaw told his board.
At AT&T's suggestion, the partial deal had become a complete sale: all of McCaw Cellular for AT&T stock valued at about $12.6 billion when announced, structured for tax reasons as a "pooling of interests" merger but functionally a buyout. When it closed in 1994 the price usually cited was $11.5 billion, or about $16.5 billion including the debt AT&T assumed, the second-largest merger in American history at the time, after the RJR Nabisco buyout. Forbes later estimated that McCaw had spent roughly $8.5 billion building the company. The McCaw brothers together took about $2.2 billion. Craig's own share, by one Forbes estimate, was about $800 million.
"At a very personal level, this was a tough decision for Craig," Allen told reporters. McCaw said it was "very difficult" for both companies, "probably more for us than AT&T."
Forbes later called it a distress sale. Alberg later called it surrender. In 2000 Forbes wrote that McCaw had "hit a wall," trapped by $5 billion in debt, and that "against his own instincts, he bailed out." In July 1994, two months before the sale closed, McCaw Cellular sent its employees a letter announcing its first accounting profit in six years as a public company. The business had been built on the long-term value of customers rather than on reported earnings. Money spent acquiring customers was expensed immediately, while the value those customers generated accumulated quietly. The profit showed up just as the owner handed over the keys.
He was supposed to take a seat on AT&T's board and an executive role. He declined the board seat, reportedly because he couldn't stand long meetings. He handed day-to-day control to Barksdale, who soon left to run a startup called Netscape. Hooper took over and by the end of 1997 had made AT&T Wireless the largest cellular provider in the country. Then Hooper left, along with most of the remaining McCaw-era executives, and within a year and a half all six McCaw regional presidents were gone.
What followed reads like a slow version of the posse movie with the posse on the losing side. In 1998 AT&T Wireless introduced Digital One Rate, a flat national price with no roaming or long-distance charges, which in effect turned McCaw's roaming map into a single price. Demand was huge: subscribers grew more than 50 percent while the company added only 350 cell towers, about 6 percent more infrastructure. The network clogged. Churn reached 35 percent in 2000. In April 2000 AT&T sold a tracking stock in the wireless business for $10.6 billion, the largest IPO in U.S. history, implying a value of about $73 billion for the whole unit, six times the price in seven years. McCaw was asked about AT&T chairman Michael Armstrong. "I hope he puts enough lift in the airplane," he said, "before it hits the ground."
It did hit. When the FCC required number portability in 2003, customers tired of years of degrading service left in large numbers, and a new system for porting numbers failed. In early 2004 AT&T Wireless put itself up for sale. Cingular won a bidding war with Vodafone and paid more than $41 billion. On April 26, 2005, the AT&T Wireless brand was retired.
At a trade show in San Francisco in 2004 someone asked McCaw about the company he had sold. He answered with the self-deprecating humor he used when he didn't want to show how he felt.
"When we sold it to AT&T, I said the company was like my child and it was growing up to go to reform school," he said. The audience laughed. Eventually, he added, "you have to step out of the way."
An Internet in the Sky
He didn't step out of the way for long. Even before the AT&T sale closed he had started the most ambitious project of his life, and in its scale and its failure it is the one most like his father.
Teledesic began in California in 1990, based on an idea from an engineer named Ed Tuck that grew out of a Pentagon missile-defense concept called "Brilliant Pebbles." It was led at first by Dave Patterson, its first employee. In 1994 McCaw and Bill Gates each invested $5 million and the company became famous overnight. The Seattle Times later said the pairing was like "Elvis and the pope on tour together." The early team (Russell Daggatt, Elaine Ferguson, Tren Griffin as the fourth employee) worked out of a couple of furniture cubicles in McCaw Cellular's offices. Griffin had turned down a Microsoft job offer the day he joined, and told Gates he needed to do one ambitious startup in his life. Gates said he understood, Griffin recalled, and "probably thought I was nuts."
The plan was 840 satellites in low Earth orbit, more than all the satellites in space at the time, providing high-speed data to any point on the planet, with satellites passing traffic among themselves by laser. The estimated cost before any service was $9 billion. McCaw called it "an Internet in the sky." Field hospitals would send images to doctors in cities. Villages without basic phone service would get broadband from orbit. In 1997, when most internet users were on slow dial-up modems and traffic was said to double every hundred days, this looked to serious people like the next version of what McCaw had already done on the ground: the six-foot cord replaced by a beam from space.
Teledesic did one thing extremely well, and it was the same thing McCaw had always done well. It obtained spectrum. Its executives traveled the world persuading governments to reserve frequencies in the high Ka band, and they got first the United Nations and then the United States to agree. The money followed: Saudi Prince Alwaleed bin Talal put in $200 million, Boeing pledged $100 million, Motorola $750 million after folding its own Celestri project into Teledesic. The company raised roughly $900 million in cash. In 1997 the plan was cut to 288 satellites. Daggatt, as president, dismissed Motorola's competing Iridium system, whose first five satellites had launched that May, as "a technology demonstration for us." Teal Group's Marco Caceres predicted Teledesic would cost "closer to $15 billion, and I'm being conservative." A cheaper competitor, Alcatel's SkyBridge, planned simpler satellites that relayed signals to ground stations rather than routing them in orbit. The Times noted that the Ka band was easily blocked by trees.
Gates was never more than a passive investor. The Seattle Times later calculated that his $5 million was about 0.05 percent of his net worth, roughly what someone worth $100,000 would spend by putting in $50. The famous names gave the project credibility. The satellites still had to be built.
Then Iridium, the "technology demonstration," went bankrupt in August 1999 after costs rose from a reported $1.2 billion to $5 billion and it defaulted on $1.5 billion in loans. In early 2000 McCaw's Eagle River offered about $75 million in temporary financing to keep it running, cut that to $5 million two weeks later, and on March 3 withdrew entirely, explaining that Iridium's voice-only satellites didn't suit a data network. Iridium's 66 satellites and 50,000 customers were left facing shutdown within days. One analyst called the system "likely to be some of the most expensive space debris ever." McCaw instead joined backers in putting up $1.2 billion for control of another bankrupt constellation, ICO Global Communications, which he described as a stepping stone to Teledesic. Others saw it as an exit ramp.
He spoke about it in religious language. "Iridium didn't respect the difference between what they knew and what they had to do," he told Forbes that spring. "With ICO Global, we'll be more conservative. Teledesic is the cross I bear. I'll put my money in. We'll do it, but we won't jump off a cliff."
He didn't jump. The market jumped for him. On the ground he had built the other parts of a planned global system: Nextel, the struggling dispatch carrier whose Motorola radios had first been sold to truckers and cab companies, in which he had taken effective control by April 1995 with a family pledge of up to $1.1 billion. The industry's trade association sent letters to every FCC commissioner calling the announcement "another vote of confidence" in wireless competition. Within four years Nextel had 3.6 million customers. There was also Nextlink, a fiber network for businesses that by 2000 owned enough capacity, McCaw said, to carry all of AT&T's traffic, and that merged with Concentric Network that year to become XO Communications. "We're on a collision course with the old phone business," he said. "You can't defend the past." In the same conversation he said: "When greed becomes this prevalent, something bad always happens. Greed is going to drive people to some very horrible conclusions — they'll bet everything and lose."
He was describing the market, and it described him back. Forbes put his fortune at $7.7 billion in 2000 and $2.7 billion a year later. His Nextel stock fell 82 percent. XO, where his stake had been worth $3 billion to $4 billion at its peak, went bankrupt in 2002, three months after he left its board. He was reported to be selling two homes, a couple of planes, a yacht, a 5,000-acre undeveloped ranch in Carmel, California, and a 780-acre island off Vancouver Island. In 2002 his youngest brother, Keith, died. In the first days of October 2002, Teledesic laid off all but ten employees and suspended its satellite contract. A spokesman blamed the collapse of telecom. In 2003 it sold its spectrum licenses. It had never launched a working constellation.
"This is how a plan to take over the world ended," The Seattle Times wrote, "with a whimper."
There was one twist in the wreckage. In 1999 McCaw had set up Nextel Partners to build out Nextel's network in smaller markets. In 2005 Nextel merged with Sprint in a $36 billion deal, and in 2006 Sprint Nextel bought Nextel Partners for $6.5 billion. The ground networks paid off. The sky didn't. "Pioneers," he told Griffin more than once, "often get arrows in the back."
The Basement in Maryland
The comeback started somewhere modest. In 2003 McCaw's representatives went to a basement office in Maryland to meet Rudy Geist, a lawyer with one client. The client was a Spanish-language broadcaster that happened to be the largest licensee in the country of spectrum in the 2.5 gigahertz band. Since the 1960s the government had given that spectrum away to schools and nonprofits for educational television. In practice most of it sat unused, a public commons nobody was grazing.
McCaw signed a master lease that gave a new Kirkland company, founded in October 2003 and called Clearwire, a foothold in about twenty markets. It went on to sign more than a thousand similar leases, committing to about $5 billion in payments over three decades. In March 2004 he bought Clearwire Holdings, a small Texas company already offering wireless broadband in Jacksonville, Florida. In April he bought NextNet Wireless, a Minneapolis equipment maker. On May 3 he put $36 million into Microcell, a Canadian carrier. He didn't explain any of it, and the industry started guessing again. In 2005 the FCC loosened the rules on the 2.5 GHz band to encourage wireless broadband, and the value of what he had leased jumped. Once again he had bought air before anyone priced it.
"Filling a need that others aren't addressing," he said in those years, "has always been a focus of the companies that I have been involved with." Clearwire first sold its equipment in places like Ivory Coast and Bangladesh, which he pointedly noted were "not the wealthy places," and in sparse communities in northern Canada. The Seattle Times summed up his career in one line: everything he built "required lots of money." He raised it again. In March 2007 Clearwire sold 24 million shares at $25 for $600 million. The stock fell 38 cents on its first day, critics said the underwriters had been too aggressive, and McCaw kept about 49 percent of the votes and 34 percent of the equity. The company lost $284.2 million in 2006 and $727 million in 2007, nearly five times its revenue.
In May 2008 he made one of the most complicated deals of his career. Sprint agreed to contribute all its 2.5 GHz spectrum and its fledgling WiMAX business, called Xohm, to a new company keeping the Clearwire name, and five investors (Comcast, Intel, Time Warner Cable, Google and Bright House Networks) committed $3.2 billion in cash, Intel alone $1 billion. A TechCrunch columnist was harsh enough to put it in a headline: McCaw had "snookered" Google's Eric Schmidt into investing in a technology the writer thought was really fixed wireless pretending to be mobile. McCaw's own statement when the deal closed in November 2008 used his customary large language. "The power of the mobile Internet, which offers speed and mobility, home and away, on any device or screen, will fundamentally transform the communications landscape in our country," he said, and the new company would give the country "the opportunity to return the U.S. to a leadership position in the global wireless industry."
In January 2009 Clearwire turned on its first mobile WiMAX network in Portland, Oregon, the second city after Baltimore. By the end of 2010 it covered 71 markets and more than 110 million people. On December 31, 2010, McCaw stepped down as chairman. The world adopted a competing standard called LTE. Clearwire's spectrum was eventually absorbed by Sprint, and through Sprint by T-Mobile, the way his cellular licenses had once been absorbed by AT&T. His creations kept growing up and leaving to be raised by others.
The Room Facing West
There is a Ferrari to account for. In 2012 McCaw paid $35 million for a 1962 Ferrari 250 GTO, then the highest price ever paid for a car. He and Bruce had once collected about 400 cars between them. This one had been built for Stirling Moss, who never raced it. Others drove it at Le Mans that year: Innes Ireland and Masten Gregory took it through the 24 hours while the man it was built for watched from somewhere else. You can make too much of that, so this account will make nothing of it beyond the fact.
There is also a whale. Starting in 1993, McCaw was the founding donor and underwriter of the Free Willy Foundation, set up with Warner Bros. to return Keiko, the orca from the movie, to the ocean. The Craig and Wendy McCaw Foundation contributed about $3.1 million, and his operating company supplied four active board members. Asked whether that was too much money for one animal, he said it would be "inappropriate, no matter how popular Keiko was and no matter how big children's dreaming was, to spend so much money on one whale if more was not going to come of it." The point was research: finding Keiko's family, which ranged a thousand miles or more from Iceland into Scandinavia, and learning how a family like that communicated and held together. His first marriage, to Wendy Petrak, ended in 1997. He married Susan Rasinski, a Stanford graduate and investment banker who later served as ambassador to Austria, in 1998, and by 2000 he was a father for the first time, with a newborn son and an eighteen-month-old daughter. In 1999 he bought a house in Hunts Point from the saxophonist Kenny G. He joined the Hoover Institution's Board of Overseers, served on the boards of The Nature Conservancy and the Friends of Nelson Mandela Foundation, and gave to the Grameen Technology Center. He gave large amounts to Republican candidates. He told friends, Forbes reported, that "things won't make you happy — but once in a while, you have to test that idea."
In 2016 he dropped off the Forbes 400. In 2025 Forbes listed him among Washington's billionaires at about $2 billion, roughly where he'd been after the crash, though nowhere near where he'd been before it. The money came and went, much as the airwaves did.
The McCaw Cellular alumni, whom Forbes called the "McCaw Mafia," spread through Seattle. Alberg put $50,000 into an untested bookseller named Jeff Bezos in 1995 and went on to co-found Madrona. Rufus Lumry started Acorn Ventures. Hooper co-founded Ignition. Perry, the young lawyer told in 1975 not to waste too much time on the crazy kid, ran Edge Wireless, reselling AT&T service in rural markets. "This clique would rather collude than compete," Forbes wrote. "We got close to the edge a couple of times," Alberg said of the McCaw years. "McCaw was viewed as a great risk taker, but we were also very much infused with a sense that we have to be careful, too." The Cellular One brand, which McCaw built into the first national network, passed through Western Wireless, Alltel, Dobson and AT&T, was sold to Trilogy Partners in 2008, and still operated as recently as the mid-2020s in northeastern Arizona and the Four Corners, a long way from Lake Washington.
The office stayed the same. In 1997 a Washington Post writer found him jet-lagged after a yacht trip along the Turkish coast, in a marble-floored office with living quarters above the marina at Carillon Point in Kirkland, nearly bare except for a huge floor globe and a telescope aimed across the lake. Four years earlier, in 1993, after AT&T bought its first third and before it bought the rest, Corr visited what reads like the same space at the end of a day: a spartan Kirkland office facing west, which McCaw described as a symbol of his focus on the future. He sipped Evian. Cheap electric baseboard heaters ran around the room. He held back almost every personal detail and gave a few only when pressed. Corr wrote that it felt like a chapel, or a library study room meant for whispers.
Late in the day the sun went down behind the Olympics. The glass walls darkened and the marble went cold. He switched on a few lamps, and the room stayed dim.
Part IIThe Playbook
Craig McCaw's career is two stories that use the same tools: a cellular empire built on license arbitrage and leverage, and a decade of post-AT&T bets in which the same methods produced Nextel's success and Teledesic's failure. The principles below come from both. They are meant to be operational: things a founder, operator or investor can test against a current decision, with the evidence from Part I and, where it exists, the evidence against.
The principles
- 1.Read the incumbent's forecast as a price signal.
- 2.Buy the option before the asset has a price.
- 3.Borrow against a map, not a hunch.
- 4.Fix your father's mistake: build what doesn't depend on you.
- 5.Know your shape, and hire the complement.
- 6.Treat visibility as a cost.
- 7.Use the dense core to buy the periphery cheap.
- 8.Stitch the map: the network beats the node.
- 9.Value the customer, not the quarter.
- 10.Hold control until holding it destroys the thing you control.
- 11.Pay for flexibility, and respect timing.
- 12.Find the neglected commons.
Principle 1
Read the incumbent's forecast as a price signal
In 1981 the document that started the McCaw fortune wasn't a market study. It was AT&T's projection to the FCC of 900,000 U.S. cellular subscribers by 2000. Most readers took the figure as information about demand. McCaw took it as information about price: if the company that invented the technology believed the market was small, every license priced on that belief was cheap. By 1992 his company alone had more than twice that number of subscribers.
The general point is that an incumbent's public estimate of an adjacent market is often a statement of what the incumbent wants rather than a neutral forecast. A monopoly that profits from the wired phone has every reason to underestimate the wireless one. Its forecast becomes the market's anchor, and the anchor sets the price of the scarce inputs: licenses, spectrum, talent, distribution.
The risk is obvious. Sometimes the incumbent is right. McCaw's edge wasn't contrarianism for its own sake. It was the combination of a contrarian forecast with an asset whose downside was small at $4.50 a pop.
Tactic: Find the dominant player's published forecast for the market next to its core, ask what it gains by keeping the number low, and price the scarce inputs as if the number were wrong by an order of magnitude.
Principle 2
Buy the option before the asset has a price
McCaw rarely bought businesses at the start. He bought rights: franchises, licenses, leases, frequencies. In 1982 he won all six cellular applications he filed. He bought lottery winners' licenses before they knew what they had. In 2003 he leased 2.5 GHz spectrum from a broadcaster represented by a one-client lawyer in a basement, two years before the FCC relaxed the rules that made it valuable. Even Teledesic's one clear success was regulatory: getting the UN and the United States to reserve Ka-band spectrum.
Rights are options. They cost little to hold compared with what they can be worth once someone builds on them, and their value depends on future events (a rule change, a technology maturing, a competitor's need) that the current holder often can't see. "People did not realize how valuable it was," McCaw said of cellular. That isn't luck. It's the deliberate act of buying the input everyone else values only after the output exists.
Tactic: List the permits, licenses, rights of way or exclusive leases your industry will need in five years, and work out what it would cost to hold some of them now, before anyone has priced them.
Principle 3
Borrow against a map, not a hunch
Both McCaws borrowed heavily. The father's debt financed a scattered collection: 54 companies, 25 bank accounts, stations bought and sold without an overall plan. When he died, creditors found nothing coherent to value, and the estate was declared insolvent at $12 million in claims. The son's debt, including nine Drexel offerings and $3.5 billion for LIN, financed one picture: a national cellular network that bankers could model with AT&T's own data. The licenses served as collateral because they added up to something.
That distinction holds through the failures. Nextel's ground network, financed against an integrated plan, eventually returned billions through the Sprint merger and the $6.5 billion Nextel Partners sale. Teledesic's satellites were financed against a vision with no ground beneath it, and they were never built.
M
Two Generations of Leverage
Father and son used the same tool for different ends.
| Dimension | J. Elroy McCaw | Craig McCaw |
|---|---|---|
| Core asset | Radio and TV station licenses | Cellular licenses, later spectrum leases |
| What debt financed | A portfolio: 54 companies, 25 bank accounts | A map: clustered markets toward one national network |
| Management | "Out of his hat"; dependent on one person | Strong executives with real authority |
| Relationship to Ma Bell | Shut down by the phone company as a teenager | Locked out the Baby Bells, then sold to AT&T |
| Outcome of leverage | Insolvent estate, 1969 | $11.5B sale, 1994; $7.7B fortune cut to $2.7B, 2000–01 |
Tactic: Before taking on debt, draw the single picture all of it serves; if a lender can't value the collateral as one coherent asset, you're borrowing against a hunch.
Principle 4
Fix your father's mistake: build what doesn't depend on you
"My father was a visionary who did not hire great people," McCaw said. "The company was too dependent on one person." He said it at the moment of his own biggest exit, and the timing is the lesson. The empire he sold could survive without him because he had made sure it would: Perry on deals, Stanton on operations, Alberg on law and negotiation, Hooper on finance and the regions, Barksdale on execution. When McCaw left for Teledesic, the company kept going, and under Hooper it became the largest carrier in the country by 1997.
Every founder has a version of the father's mistake: a part of the business that runs only through their own head. McCaw's answer was structural, not inspirational. "If you trust someone, you should burden them with responsibility," he said. "The greatest burden you can put on someone is trust." The McCaw Mafia (Madrona, Ignition, Acorn, Edge Wireless, Cedar Grove) is evidence that the burden made people.
The counter-evidence is just as useful. Once the McCaw-era people left AT&T Wireless, the culture they carried left too, and the company declined. Distributed capability survives a founder's departure. It doesn't always survive a new owner's.
Tactic: Name the three decisions in your company that only you can make right now, and assign each to someone else this quarter, with authority, not just input.
Principle 5
Know your shape, and hire the complement
McCaw was dyslexic and said plainly what that meant: he couldn't take a sheet of paper and organize a plan others could follow. "I have to explain conceptually what we want to accomplish, and then somebody else has to translate that into a concise organized plan." He then hired the exact opposite of himself. Barksdale's rule, "If you can't put it on one piece of paper, you can't execute," is the most McCaw-complementing sentence in the record, and it came from someone else.
This goes beyond "hire people smarter than you." It's precise self-description followed by precise hiring. McCaw knew he was conceptual, introverted, bad at reading people as individuals, and good at seeing patterns in their ideas. He staffed against each weakness: John as the "relationship guy" for LIN, Perry and Alberg as negotiators, operators for execution. Perry's warning about the 90-percent threshold shows the necessary friction. A team that complements a founder also has to be able to tell him when his 90 percent is their 40.
Tactic: Write a one-paragraph description of how your mind actually works, including the failure modes, and check whether your top three hires were chosen to cover them or to resemble you.
Principle 6
Treat visibility as a cost
McCaw called his IPO, which made his family billionaires, "the worst day in the history of our company," because the prospectus told every competitor what he owned and what he still needed. He didn't put his photo in his annual reports, turned down interviews, and wanted to be the Wizard of Oz in the sense of having "an impact without being particularly visible." He identified with the anonymous posse in Butch Cassidy, not the stars.
This isn't modesty. When the strategy depends on buying mispriced assets, every public statement can move the price. A clear, visible plan is something counterparties can bid against, which the Pels episode showed. Secrecy has costs (teachers wondering about "the real person," employees calling him inaccessible, a biographer who got two interviews before McCaw lost interest), and those costs are real. They're usually smaller than the cost of announcing your shopping list.
Tactic: Before any public disclosure (a launch, a raise, a press interview), ask what it tells the people you still need to buy from, and say only what is required.
Principle 7
Use the dense core to buy the periphery cheap
McCaw's major-market licenses did two things. They were valuable businesses, and they were leverage. The Baby Bells often needed his spare licenses to enter markets he dominated and paid large premiums for them. He used the same position to buy, sell and swap licenses in smaller markets others thought unprofitable, usually at a steep discount, building clusters no one else could put together.
The pattern applies widely: a position that is essential to a few powerful buyers lets you acquire what everyone else is ignoring. The core finances the periphery, and once the periphery is assembled it becomes a new core. The LIN deal is the boldest example: $350 a pop, against $4.50 at the start, to buy the two cities that turned a regional cluster into a national network.
Tactic: Identify which of your assets your largest potential counterparties can't do without, and use those negotiations to pick up the lower-value pieces that complete your map.
Principle 8
Stitch the map: the network beats the node
Before 1990 a cell phone was a local device that stopped receiving calls at the edge of its home market. McCaw's SS7 rollout and the Cellular One brand turned isolated systems into one network, and with it created roaming. The licenses were worth more connected than separate, and the customer cared about connection, not licenses. That is what McCaw meant by uplifting the human being with "a more powerful communications tool," and by his contempt for "the end of a six-foot cord."
The later history shows the principle's flip side. Digital One Rate, AT&T Wireless's 1998 flat national plan, extended the stitched map to pricing and attracted enormous demand, then collapsed because the network underneath had grown only 6 percent. A single network is a promise. It holds only if capacity keeps pace with the demand the promise creates.
Tactic: Find the boundary where your product stops working for the customer (a region, an integration, a handoff) and prioritize removing it over improving any single node, provided you can fund the capacity that follows.
Principle 9
Value the customer, not the quarter
McCaw Cellular reported its first accounting profit in six years as a public company in July 1994, two months before AT&T paid $11.5 billion for it. That wasn't an inconsistency. Equipment purchases were depreciated over years, but customer acquisition costs were expensed immediately, so a company growing quickly through customers looks unprofitable precisely because it is creating value. McCaw, like John Malone in cable, managed for the discounted value of long-term cash flows and found lenders and investors who would underwrite that view.
Griffin's warning belongs with this principle: customer lifetime value models can "discover value or hide problems," and their worth depends entirely on their assumptions. Clearwire's 2007 losses, $727 million on revenue about a fifth of that, show what it looks like when the per-customer economics haven't been proven yet.
Tactic: Build a bottom-up per-customer valuation of your business with explicit, debatable assumptions, and add a margin of safety before you let it justify losses.
Principle 10
Hold control until holding it destroys the thing you control
For two decades every McCaw deal had one condition: Craig McCaw stays in charge. He gave that up only once, in 1992, by selling AT&T a $100 million option on control, and then in 1993 by selling everything. He did it with more than $5 billion of debt and a network that needed capital he couldn't raise alone. Hooper later said plainly that AT&T's balance sheet bought licenses "we couldn't have done … as a stand-alone business."
The lesson isn't "never give up control" or "always sell." Control is a tool. McCaw kept it while it protected speed and judgment and gave it up when keeping it would have starved the company. He didn't like it. Forbes wrote that he "bailed out" against his own instincts, and he compared the company to a child headed for reform school. But the alternative was a second insolvency, which he had already seen once.
⏱
The Control Decision
From total control to sale in roughly twenty months.
1987
IPO; McCaw retains majority voting control and calls it "the worst day in the history of our company."
1989
LIN Broadcasting won over BellSouth at about $3.5B; debt passes $5B.
1992
AT&T buys one-third for $3.8B plus a $100M option on control.
1993
Four-hour meeting with Robert Allen; handshake at the Waldorf Astoria.
1994
Sale closes at $11.5B (about $16.5B with debt); McCaw declines the board seat.
2004
AT&T Wireless sold to Cingular for $41B after years of decline.
Tactic: Write down now the specific balance-sheet or competitive condition under which you would give up control, so the decision is made in advance and not under duress.
Principle 11
Pay for flexibility, and respect timing
McCaw's maxims emphasize optionality: "Flexibility is heaven." "We have always tried to have the discipline of being over financed so if you had a problem over a period of time you could survive it." "Pioneers often get arrows in the back." He walked away from Iridium after committing $75 million, cut the commitment to $5 million, and left, accepting criticism rather than putting more money into the wrong satellites. "We won't jump off a cliff," he said about Teledesic.
The record also shows how hard this is. From 2000 to 2001 his fortune fell from $7.7 billion to $2.7 billion, XO went bankrupt, and Teledesic ended. Being overfinanced didn't help when the whole sector collapsed at once. The honest lesson is that flexibility guards against your own mistakes better than against systemic timing, and the best defense against timing is the "via negativa" Griffin attributed to him: learning from those who failed before you.
◎
The Post-AT&T Portfolio
Same playbook, different outcomes.
| Venture | Bet | Result | Verdict |
|---|---|---|---|
| Nextel | Up to $1.1B family pledge (1995) | 3.6M customers in four years; $36B Sprint merger (2005) | Won |
| Nextel Partners | Founded 1999 for secondary markets | Acquired by Sprint Nextel for $6.5B (2006) | Won |
| Nextlink / XO | National fiber for businesses | Bankruptcy 2002; stake once worth $3–4B lost | Collapse |
| Teledesic | 840, then 288, satellites; ~$900M raised | Work halted 2002; spectrum sold 2003 | Collapse |
| ICO Global | $1.2B with backers for bankrupt constellation (2000) | Planned constellation never completed | Collapse |
| Clearwire | 2.5 GHz leases; $3.2B Sprint-led recapitalization (2008) | 110M people covered by 2010; WiMAX lost the standards war | Mixed |
Tactic: For every large bet, write down the cash cushion that lets you survive two bad years and the specific evidence that would make you walk away, then honor both when it costs you.
Principle 12
Find the neglected commons
The 2.5 GHz band had been given away for educational television since the 1960s and had mostly gone unused. McCaw's Clearwire signed more than a thousand leases on it before the FCC's 2005 rule change made it valuable, which positioned him to compete with Sprint for a national portfolio and eventually to absorb Sprint's spectrum. It was the 1982 license strategy again, in a different form: find a resource governed by an old purpose, held by people who don't use it, and imagine its next use.
"Filling a need that others aren't addressing has always been a focus of the companies that I have been involved with," he said. The same instinct sent Clearwire to Ivory Coast and Bangladesh, and the McCaw cable company to remote towns in Washington and Alaska. Neglected resources and neglected customers are often the same opportunity seen from two directions.
Tactic: Look for an asset in your field whose rules were written for a use that no longer exists, and find out who holds it and what they'd accept for a long lease.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Craig McCaw, on his father, at the announcement of the AT&T sale (1993)My father was a visionary who did not hire great people. The company was too dependent on one person.
— Craig McCaw, as recorded by Tren Griffin, 25iqI hate the orchestration of life. Most executives die frustrated (because) they have been orchestrated into boxes.
— Craig McCaw, to Forbes (2000)Iridium didn't respect the difference between what they knew and what they had to do. With ICO Global, we'll be more conservative. Teledesic is the cross I bear. I'll put my money in. We'll do it, but we won't jump off a cliff.
— Craig McCaw, to Forbes (2000)You must have the will to evolve. There isn't any manager who doesn't feel tired, beleaguered. In a changing world, if you keep doing the same thing, you lose.
— Craig McCaw, as recorded by Tren Griffin, 25iqIf you aren't scared you probably aren't doing anything.
Maxims
- Read the forecast, then bet against its author. An incumbent's low estimate of a new market usually reflects its interests, and the price of entry is set by that estimate.
- Buy rights before businesses. Licenses, leases and spectrum are cheap options on futures that most people haven't priced yet.
- Debt needs a map. Leverage behind one coherent picture survives scrutiny, while leverage behind a collection dies with the collector.
- Don't be the company's single point of failure. The founder's job is to make sure the business would continue if he walked into the wrong room on the wrong day.
- Describe your own mind accurately, then hire the opposite. The conceptual founder needs someone who insists on one page.
- Invisibility is a strategy. Every public statement tells a counterparty what you still need to buy.
- Connect the pieces. Customers pay for the network that finds them anywhere, not for any single tower.
- Measure value, not reported earnings. The first accounting profit can arrive the same quarter the company is sold for billions.
- Give up control on your own terms. Decide in advance what conditions would justify selling, before debt makes the decision for you.
- Being first is not always best. Pioneers get arrows in the back, so learn from the failures ahead of you before adding your own.
