(FORTUNE Magazine) – When Terry Semel showed up at Yahoo's Sunnyvale, Calif., headquarters on May 1, 2001--his first day as CEO--he had barely used a PC or surfed the Net. He read e-mail by having his assistant print out a stack of pages, and responded by hand, scrawling notes in the margins for her to retype. This was the man co-founder Jerry Yang had chosen to rescue the company? To many investors and staff, Yahoo was taking another step toward the abyss rather than away from it.

As a former co-head of Warner Bros., Semel was known in Hollywood as one of the movie industry's most daring, creative, and successful businessmen. He took chances on movies like Batman--and made them pay off big. He backed aging stars who wanted to direct, like Clint Eastwood, and won Oscars. He cut deals that seemed crazy but are now widely imitated. And he was so good at handling the industry's giant egos that many say they actually liked working with him--a rare comment about a studio boss.

But three years ago, as the air was rushing out of the dot-com bubble, it was hard to see how a career hanging out with movie stars would be of much help to Semel in Silicon Valley. If anything, it seemed a handicap. In Hollywood he had worked in a mahogany-paneled office half the size of most homes, with two assistants tending to his every need. At Yahoo everyone, including the CEO, worked in cubicles. Semel believed in spending money to make money, liberally doling out rides on Warner's fleet of private jets and stays at the company's compounds in Acapulco and Aspen to make his stars feel pampered. Yahoo looked as if it needed a cost cutter, someone who could give the business--which was in shambles--a good shake. And unlike other media moguls, Semel was publicity-shy: In front of a big crowd his Brooklyn-accented monotone made most people's eyes glaze. Silicon Valley, home to Oracle's Larry Ellison, Apple's Steve Jobs, and Sun Microsystems' Scott McNealy, likes its CEOs on the flamboyant side.

The culture shock was palpable. The company had been a quintessential high-tech startup. Yang and fellow engineer David Filo founded Yahoo while they were students at Stanford University in 1994. Begun as a mere directory of websites, it became a darling of the high-tech bubble and grew rapidly into a global Internet portal--the online equivalent to a TV network. Yet even after the bubble burst in 2001, it was still a company where geeks ruled.

As many investors now know, Semel has proved most of his doubters wrong. A thorough reorganization, a series of shrewd deals, and a rebound in the online-advertising market have combined to send Yahoo's growth and profits soaring. The company lost nearly $100 million on $717 million in revenues the year Semel arrived; in 2003, 22 years later, it made $239 million on $1.4 billion in revenues. This year it made the FORTUNE 1,000 for the first time, and if it grows 20% a year--the projected growth rate of online advertising and roughly a third its current growth rate--it will be a FORTUNE 500 company within five years. It now has 150,000 advertisers, including 70 of the country's 100 biggest accounts. Registered users--those who use Yahoo at least once a month--have more than doubled since Semel arrived, to 133 million, and the money they spend with Yahoo has almost doubled. Yahoo's stock price, which fell below $10 a share shortly after Semel arrived, has been as high as $50 since then and is more than $40 right now.

But it's not Semel's high-tech acumen that is winning him plaudits. It helps that he now knows how to use a computer, of course, but it's his analog skills, the ones he learned in Hollywood, that are behind it all. How? Cut through the perks and trappings of Semel's Hollywood life and what you get is a born dealmaker, and a tough but hands-off manager that people like working for. Indeed, Semel is quite happy to tell you that his lack of tech savvy is one of his greatest assets. "I see myself as a typical user who wants things to happen with ease and comfort," he says. "When it takes me more than ten minutes to sign up for one of our services, I know something is wrong." In other words, contrary to what everyone first thought, Semel's background turned out to be exactly what Yahoo needed.

The story of how a luddite like Semel ended up running one of Silicon Valley's signature startups is pretty straightforward. Semel had been an informal advisor and friend to Yang since the pair met at Herb Allen's annual Sun Valley, Idaho, media conference in 1997. Yang wanted to pick Semel's brain about the media business. Semel wanted to know more about the Internet. For example, the two talked extensively when AOL and Time Warner merged about what Yahoo should do in response. In April 2001, less than a month after Yahoo decided to replace chairman and CEO Tim Koogle, Yang offered Semel the job over lunch. (Although Yang has no operational role at Yahoo, he is at headquarters all the time and, with 6.5% of the stock, is the most influential nonexecutive board member.) He knew Semel's background would make him a controversial hire but says he believed those fears wouldn't last. "Everyone talks about what he did with movies and entertainment," Yang says, "but what he really did was pioneer how to take a piece of content and get it out there. He has a distribution mentality, which at the end of the day is what Yahoo does on the Internet. And so when we started talking about Yahoo generically as a distribution company, we both just went, 'Gosh, this is going to be really cool.'"

Semel accepted the job almost immediately. For one thing, he could afford to take a risk. When he and co-CEO Bob Daly quit Warner Bros. in 1999, many estimate that they were worth hundreds of millions of dollars. (There is still debate about whether he and Daly left Warner of their own accord--as they say--or were pushed out by then-CEO Gerald Levin.) More important, his lack of skill in using a PC or e-mail notwithstanding, Semel understood that the Internet was growing into a communications and entertainment medium. He had backed the launch of Warner's website in 1996, after all. And he correctly understood the impact of new technology when he supported Warner's push into DVD technology against stiff industry resistance. Indeed, one attraction of the Yahoo job, Semel says, is that in spirit Yahoo reminded him of Warner Bros. in the 1970s: "It had a great brand and great people, but its business was broken." Meanwhile, the way Yahoo made money--through advertising--was as old as the entertainment business. "The idea of being CEO didn't scare me because I felt like I had been there before."

It would be tempting to conclude that Semel first got into entertainment, almost 40 years ago, because he desperately wanted to inject some glamour into an otherwise unglamorous Brooklyn upbringing. In truth, he wasn't thinking about glamour at all when his friend Dan Romanelli told him about the Warner Bros. sales-training program in 1965. An accountant by training, he was deathly bored and saw the Warner job as a "chance to learn about marketing and sales, which I was interested in."

His early movie jobs were arduous. During stints in New York City, Cleveland, and Los Angeles, all Semel did was sell movies to theater chain owners. He'd show up with a list of the movies Warner was going to release over the next few months, and each owner would bid on the movies he wanted. There were few national chains in those days, so Semel had dozens of theater owners to cut deals with, and he traveled constantly. He became such a whiz that by the time he was 32, he had already had jobs running CBS's and then Disney's distribution. Lured back in 1975 to head Warner's distribution, Semel and Daly were managing the entire studio within five years for Warner Communications CEO Steve Ross.

The 20 years that Semel and Daly ran Warner will probably go down as one of the longest and most successful partnerships in Hollywood history. The pair didn't just greenlight movies; they changed the way movies were made and marketed. Semel pushed the idea of renting Warner's excess movie-distribution capacity to other moviemakers as a way of creating a less risky income stream. Almost every studio now has such an arrangement. He and Daly backed releasing and selling movies inexpensively on DVD before anyone else, despite concerns that the move would destroy the videocassette-rental business. Instead the market expanded. They made Batman, an expensive, unconventional adaptation of the comic-book character by an untested director, and turned it into a merchandising bonanza. Now merchandising has become an integral part of the way movies are financed and marketed. During their tenure revenues grew tenfold, to more than $11 billion.

Semel was so skilled at negotiating deals that Hollywood even has a name for it: "getting Semelized." Producer Arnon Milchan, who made roughly 40 movies with Semel, describes it like this: "First he doesn't return your calls for a while, so you get desperate. Then, when he finally agrees to see you, he makes you wait two hours. By the time you are ushered into his office, and he's given you a big hug and asked you about your life, you're so emotionally drained that you've almost forgotten why you came. He, however, has not."

Virtually every studio made a play for Semel after he left Warner. But the Internet bubble had piqued Semel's interest in the new medium. He formed a small investment group called Windsor Media with four younger colleagues from Warner's online division and set out to look for deals. But Jim Bannister, one of the partners, says the discussions always came back to Yahoo. "It was always on the radar screen."

As it turned out, investors and Yahoo's staff were skeptical about Semel for a long time. Before he arrived, there had been speculation that he'd been hired only to sell Yahoo to a studio. The staff was abuzz that he would require everyone to call him "Mr. Semel." His decision to turn a small conference room into an office and not work in a cubicle like former CEO Koogle caused a stir. Many assumed he'd had a tantrum, and Semel's quiet, methodical style only fanned those flames at first. He denied the rumors about "Mr. Semel" and his office space tantrum, but he seemed aloof and isolated--he never roamed the halls to say hello the way Koogle had. And outside of the company's heavily scripted quarterly earnings staff meeting, he rarely appeared in public. During the day he went from closed-door meeting to closed-door meeting, often eating lunch while watching a presentation. At night he'd be driven in his black Range Rover back to his apartment in San Francisco's Four Seasons hotel. And on weekends he'd hop on his private jet to see his family in Bel Air.

Semel, who at 58 was twice as old as the average Yahoo employee, had so much to learn that even some of his senior staff worried he would not get up to speed fast enough to lead effectively. "He'd say some things where you'd just go, 'Oh, my God, he doesn't know,'" says Jeff Mallett, Yahoo's former president and now a co-owner of the San Francisco Giants. "We'd be talking to him about buddy lists and Yahoo's server protocols [the standards the company's thousands of computers used to communicate], and he'd say, 'What's a buddy list? What's a server? What's a protocol?' He didn't know that you could log on to AOL from the Internet instead of using AOL client software." Mallett says Semel called him into his office at 8:30 one night because he couldn't figure out how to log in to Yahoo's Geocities website.

But while his image was taking a beating, privately Semel moved to lay the foundations for a turnaround. Yahoo's growth had been explosive during the bubble. Revenues, largely from Internet advertising, grew from $84 million in 1997 to $1.1 billion in 2000. But when the bubble burst, half of Yahoo's advertising revenue disappeared in the space of a year. Indeed, the company's business fortunes turned so suddenly that when Semel arrived, some investors worried that the company might soon disappear entirely, like so many of its dot-com brethren. Semel's vision for the business wasn't fancy. He believed that Yahoo's goal of building a new way to entertain and inform people was sound, but he thought the company wasn't going at it the right way.

One of his most important moves in those early days involved Mallett, who as president had openly campaigned for the CEO job and who made no secret of his unhappiness at not being chosen. In most situations like that, the loser quickly leaves the company. Mallett stayed an entire year and to this day speaks of Semel with admiration. "I know it's a cliche, but he listened, and that won me over. He has this way of making you feel comfortable, even in tough situations." He says Semel's stamina while learning about Yahoo's business and technology was incredible. One-hour meetings turned into six-hour marathons. "Sometimes we'd just walk him over to the computer and show him what we were talking about. Then he'd say, 'Let me repeat it back.' And we'd go over it again. He'd stay in that conference room for hours until he got it. I think he learned three years of information in six months."

You can tell how much Semel has learned about technology just by listening to him. He can rattle off the names of search companies, like Espotting and FindWhat, that most people have never heard of. And he gets wide-eyed talking about music streaming and downloading. "Yahoo averages 150 million music streams a month. Think about how many music videos you'd see on MTV in a day--150?" he says. He's even begun to send his own e-mail--typically at 11 on Sunday night. As Semel got up to speed on the technology, he also set about reorganizing the way Yahoo did business. He looked for new ways to diversify its revenues, which were completely dependent on one kind of advertising. He created premium services that customers would pay for, and expanded the kinds of advertising that appeared on Yahoo's pages. Today, 17% of Yahoo's revenues are from fees, 36% from display advertising, and 40% from the web equivalent of classified ads, according to Morgan Stanley estimates. And he gutted Yahoo's sales force, bringing in Wenda Harris Millard, who had had a long career in traditional branded advertising, to run it.

Yahoo had grown so fast that it didn't have an ad sales team in any conventional sense. Few of its salespeople had any industry expertise or a Rolodex of relationships. Most just took orders via e-mail. The premise that you might have to make a formal business presentation about an idea or call on an advertiser or agency personally seemed a colossal waste of time. And with so much ad business coming in unsolicited, who had time for lunch when you could be booking business? Of course, when Yahoo's ad business dried up, its salespeople had no relationships to fall back on, nor enough experience to recast their sales pitch. Semel recalls, "I'd say to the head of entertainment advertising, 'Pitch me,' and after 15 minutes I'd have to say, 'I have no idea what you're talking about. Even worse, you have no idea either.' '' Today at Yahoo you're not doing your job if you're not constantly on a plane visiting ad agencies and their clients and racking up big bills entertaining them. If you have a new idea, you'd better know how to write a business plan and pitch it in formal meetings for it to see the light of day.

Semel went to work reorganizing Yahoo's structure too. One of his strengths as a dealmaker is that he knows how to get negotiations focused on two or three points instead of ten. Too many points to negotiate keeps a deal from getting done, he believes. And when he looked at the way Yahoo was structured, it looked like a deal with way too many points. "The company had 44 business units when I got there," he says. "I'm not sure if GE has 44 business units." After methodically meeting with every one of Yahoo's divisions, he slimmed the organization to five, then four, groups--media and entertainment, communications, premium services, and search. He also created what has come to be known as the product council. Before Semel arrived, new products grew out of the brains of engineers. But there was little coordination or planning across divisions. The product council's goal is to ensure that every division head knows about every major product in the pipeline. Yahoo today is very much the New Age media company Semel envisioned. Technology is an important thing but not the only thing. Many of his top executives, like COO Dan Rosensweig and CFO Sue Decker, are not techies but have backgrounds in media or finance.

Ultimately, it is Semel's dealmaking that has had the biggest effect on Yahoo's business and morale. Semel's management style hasn't changed. Many employees have never met him face to face, and his speeches are still stultifyingly boring. But his string of bold and successful partnerships and acquisitions has made him seem more like Gary Cooper than Peter Sellers. In late 2001, Semel launched a hostile takeover of Hotjobs, a move that catapulted Yahoo into the No. 2 spot in the hugely profitable online-personals business, behind Monster.com. At the same time he convinced the toughest, most intransigent CEO in all telecom, Ed Whitacre of SBC Communications, to partner with him in selling co-branded broadband connections to SBC customers. In 2003 he acquired Inktomi and Overture, deals that have enabled Yahoo to challenge Google for dominance in search.

The search deals are classic examples of how Semel thinks, manages, and cuts deals that work. Eighteen months ago Yahoo wasn't even a player in search. It relied on partners: Google for search results, Overture Services for search-related advertising. Now, after buying Overture and Inktomi, Yahoo has its own offering and is Google's biggest competitor. AOL and MSN, the other two Internet portals, would love to be the greatest threat to Google but can't make that claim yet. "I can't tell you how many times in meetings someone said, 'Google is too far ahead. We can't catch them,' and Terry said, 'We have to,'" says a former Yahoo executive. Yet once the deals were done, he refused to rush a product to market, despite intense internal pressure to do so. Instead, he had Yahoo's engineers rebuild its acquired search technology almost from scratch. Now, even though Google still gets all the buzz, Yahoo's search is actually better in features like local search and shopping. "We didn't get into search to do what everyone else is doing. We got into search to change the game," Semel says.

Not everything has gone Semel's way. He's had to close Yahoo's business of selling portal technology to corporations. His big bet, called Platinum--that consumers would pay a premium for entertainment streamed to them at their computers--has flopped as well. And he is under continual pressure from investors to expand Yahoo's relationships with broadband carriers. The deal with SBC was a watershed because it gave Yahoo access to a third of the nation's homes. But that deal is now two years old. Despite smaller deals with British Telecom in Britain and Rogers Communications in Canada, investors are starting to wonder when Semel will do his next big deal. It's a legitimate concern. But as competitors have discovered throughout his career, Semel is an easy man to underestimate and a very tough man to beat.

FEEDBACK fvogelstein@fortunemail.com

FORTUNE.COM SUBSCRIBERS ONLY Read more about Yahoo online, including how it has positioned itself to challenge Google, and the company's complete FORTUNE 1,000 stats.