(FORTUNE Magazine) – It surely wasn't happenstance that Ivan Seidenberg chose Las Vegas as the place to up the ante in the communications war. During the Consumer Electronics Show in January, the chairman and CEO of Verizon mystified attendees when he laid down his latest bet: a two-year, $2 billion spending plan to start digging up and replacing the last leg of the nation's century-old copper-wire telephone network. Sounding like an information-highway-crazed visionary of a decade ago, Seidenberg unveiled a scheme to bring fiber-optic cable all the way to businesses and homes, creating a high-speed infobahn from which commerce, entertainment, education, and health care would thunder into living rooms everywhere. This would be the start of the "all-broadband, all-the-time lifestyle," he declared, adding that the industry was "at the beginning of a communications revolution."

The only thing missing from Seidenberg's pronouncement was a couple of showgirls and some feathers. After the lights came up, people in the industry were scratching their heads. Could this really be the strategy of a $68-billion-a-year phone and cellular giant that has become the company to beat in telecom and yet is under pressure on a half-dozen fronts? Had Seidenberg been asleep during the dot-com era, when companies like Global Crossing and Williams Communications lost billions laying fiber-optic cable that hardly anybody ever used? What about the rising competition confronting Verizon's $22.5-billion-a-year wireless business? And what about the ongoing regulatory threat? Rival telecom companies are still lobbying the government to lower the fees they must pay Verizon and the other Baby Bells to send calls through their local networks--fees that constitute one of Verizon's most lucrative revenue streams.

But if the grand proclamations in Las Vegas raised more questions than they answered, that's Ivan Seidenberg for you. In the past decade the enigmatic CEO has transformed the ugliest of the Baby Bells into a major force in both the sleepy traditional phone market and the red-hot world of wireless--and he hasn't done it by showing his cards. Verizon's revenues are about a third bigger than those of its nearest peer, SBC, and the company has a hefty $22 billion in cash flow. It is not only the largest local phone company (see chart) but also the No. 2 long-distance provider, behind AT&T. It is No. 1 in the fast-growing wireless industry, where its momentum is so great that it helped push archrival Cingular to acquire AT&T Wireless. And in the past six months Verizon's stock has risen 15%, outperforming the shares of its competitors.

Through all that, Seidenberg has emerged as telecom's consummate poker player, one who, despite a penchant for flashy pronouncements, plays close to the vest. He has made huge, daring bets in the past--witness the company's $52 billion merger with GTE in 1998--and he has reason to bet boldly now. Verizon's former monopoly--traditional local voice service, which makes up about 42% of its total revenues (see chart)--is besieged. In the past three years Verizon has lost eight million of its 63 million access lines as customers have switched to resellers like AT&T and MCI or to cable companies offering voice services, or as they have simply dropped their phone lines for wireless.

What Seidenberg did in Las Vegas was send a message to the cable companies trying to steal his customers: Verizon will spend to stay in the game. His broadband plan could take Verizon into a wide range of services that cut into cable industry turf, including HDTV programming, interactive gaming, and perhaps even home monitoring and security. The goal, for now, is to roll out fiber to the curbside of one million homes by the end of the year, at a cost of $1 billion--more than 8% of Verizon's total capital expenditure budget for 2004 --and another two million homes next year. That's a significant investment. But if Seidenberg does wind up broadly rolling out fiber to Verizon's 32 million other customers, the stakes could go much higher--to $40 billion, some analysts calculate. A bet that big would take years to recoup, which is one reason Susan Kalla, a longtime telecom analyst at Friedman Billings Ramsey, thinks it's a bluff. "He's not going to do it," she says simply. "The numbers, they just don't make sense." Yet like other veteran Seidenberg watchers, Kalla knows it's all part of a bigger game.

Sitting at a corner table in the 33rd-floor cafeteria of Verizon's Midtown Manhattan headquarters, Seidenberg smiles thinly and contemplates his lunch (a salad and a banana) as he weighs the reasons so many investors question his broadband plan. "In the early stages of anything new, management shouldn't be particularly bothered by investors who are skeptical about the strategy," he says, shrugging. Those doubts, however, make him a bit defensive. "Most investors only understand that which has already been done. They never really like things that haven't been done before," he says, digging into the salad. "That's why Christopher Columbus had so much trouble getting financing."

Does Seidenberg consider himself a visionary? Ask him, and he'll recoil. He believes telecom is an industry that has had its fill of messiahs whose grandiosity defined and ultimately destroyed the companies they ran. "You want a word for me? Operator. Grinder. Manager. Those are words I'd rather be called," he says. "Plodding. Slow. Those are good words too." But there are plenty of industry insiders who call him prescient. Says management advisor Ram Charan, who has known Seidenberg for 20 years: "He has the mental agility to reframe issues, ideas, and perspectives from multiple angles. That's a rare leadership talent."

Seidenberg, 57, has an unlikely background for a FORTUNE 500 CEO. He was raised in the Bronx, where his dad was an air-conditioning and refrigeration installer. His first job was as a building janitor. Later, in 1966, he started climbing into manholes as a cable splicer with New York Telephone. After Army service in Vietnam, he returned to the phone company while attending classes at night to complete his undergraduate and MBA degrees. Seidenberg joined AT&T in 1974, but after the 1983 breakup of Ma Bell he moved to Nynex. There he began a methodical climb up the ladder until he was named chairman and CEO in 1994.

Not that Seidenberg will regale you with tales of his past. He is extremely guarded, weighing his comments carefully and revealing only what he wants you to know about himself. At lunch, he remarks that he rose at 5:15 that morning to run in Central Park. With some prodding, he adds that he runs about four miles, four times a week, then ends the casual personal conversation with a wave of his hand. He prefers to speak about members of his management team and frequently refers questions to them, hinting with his self-deprecating wit that he'll get the answer wrong anyway.

Seidenberg's desire to fly below the radar may partly explain his success in dealmaking. "He's masterful at the behind-the-scenes negotiations," notes an analyst who didn't want to be named. One negotiating weapon is his remarkably low, calming voice. Publishing magnate Mort Zuckerman once said that Seidenberg possessed "one of the most soothing and persuasive voices. 'Sure, whatever you want, Ivan. You want my wallet? You want my child? You want my wife?' "

That soothing voice can also scathe. In conference calls with Wall Street analysts, Seidenberg makes it perfectly clear through his tone or brusque manner when he thinks a question is, well, dumb. A mutual fund investor who didn't want to be named recalls an episode at a conference last year when Seidenberg chided another investor for asking whether Verizon might acquire long-distance company Sprint. "Real condescendingly, he responded that he didn't know why he even bothered to answer these types of questions," recalls the witness. "This was an investor who owned something like six million shares in the company. I always wondered what he did with them the following Monday."

Still, Seidenberg is willing to set aside his ego for a time to get deals done. In 1996, NYNEX and Bell Atlantic announced their $23 billion merger under the Bell Atlantic name. Two years later, another megamerger, with GTE. By the time that deal closed in 2000, the company had a new name: Verizon.

What set pundits agog was that Seidenberg took the back seat in both deals. The first time, Bell Atlantic CEO Ray Smith landed the top job; the second time, Seidenberg shared a co-CEO job with GTE's Chuck Lee. "Taking the second-banana position a second time was shocking," says John Malone, founder of consulting firm Eastern Management Group. "But today, it looks like a Harvard Business School model on how to do a merger. It was brilliant." Seidenberg's acceptance of the No. 2 slot made the deals happen quickly. And in both cases, he made sure that the other CEOs had fixed retirement dates when the reins would pass to him. He notes dryly, "Sharing responsibility for a three-, four-, or five-year period in the history of the world was not a big deal."

Another example of Seidenberg's unrelenting focus on getting the deal done occurred in 1998, after he lost a heated two-week bidding war for San Francisco's AirTouch Wireless. Determined to create a nationwide cellular footprint, Seidenberg called up the winning bidder, CEO Chris Gent of Vodafone, and pro-posed a partnership. The company that emerged, renamed Verizon Wireless, gave Bell Atlantic 55% ownership and operating control. Vodafone got 45% of the company and the option, each June, of selling its stake back to its partner. (Vodafone isn't interested in selling.)

By the time Seidenberg became sole CEO of Verizon in spring 2002 (he got the chairman title in late 2003), the Internet frenzy had cooled. Verizon's stock price had plunged; it fell from an all-time high of $70 in late 1999 to $27 in the summer of 2002. But Verizon was one of the lucky ones in the telecom rout. It survived, but only to face a new enemy: cable.

"The competition used to be AT&T, WorldCom, and Sprint," explains Paul Lacouture to a group of about 100 Verizon employees stuffed into a small conference room on a rainy afternoon in Baltimore. His sleeves rolled up, the head of Verizon's network gazes around the room. "The enemy has changed. Now cable is the enemy. They're the bad guys," he says, poking his finger in the air at the invisible threat. Members of the audience start to murmur and nod in agreement. "We've got to go out and beat cable," he continues, his voice rising. "Because if we don't, they're going right after our core business: voice."

For Lacouture and others at Verizon, the biggest enemy is cable giant Comcast. "The power of [Comcast CEO] Brian Roberts and the capital that company can put to bear, to me, is really concerning," admits Doreen Toben, Verizon's energetic CFO. Cable companies spent an estimated $75 billion in recent years upgrading their infrastructure to offer customers discounted bundled packages of local voice, high-speed Internet connections, and video. They could grab a quarter of the local voice market over the next decade as they deploy new voice over Internet protocol (VOIP) technology, estimates John Hodulik, a UBS telecom analyst. Indeed, last year Verizon's revenues from local voice fell 3%, to $28.3 billion, and operating income fell 20%, to $7.2 billion.

Seidenberg and others at Verizon deeply regret that they didn't follow through on earlier attempts to launch broadband. (In 1993, Seidenberg had announced that NYNEX would spend about $12 billion bringing fiber to the home; in 1994, he proclaimed of another venture, "We will offer consumers the next generation of on-demand programming and interactive services, including shopping, games, sports, news, information, and education." The efforts went nowhere.) "Just think of where we'd be if we had started building fiber-to-the-premise ten years ago," Seidenberg says wistfully. He doesn't want to wait another five years to allow the cable companies--or other predators--to widen their lead.

While Verizon works to get a broadband network in place, Seidenberg is hedging his bets. Last November the company partnered with digital broadcast satellite provider DirecTV; by midyear most Verizon customers will be able to get DirecTV bundled with their voice and Internet services, all on one Verizon bill. After Verizon installs fiber, it hopes to piggyback on DirecTV's programming agreements with cable channels to offer content to its subscribers by early next year.

Analysts agree that Verizon has to make a move, but many wonder whether fiber to the home is the right technology. Many say there's no reason to take fiber from the street (fiber-to-the-curb, in telecom lingo) all the way to the house. Verizon estimates it will cost an average of $1,000 to $1,200 to bring fiber to each home. About 45% of that cost is incurred from the street to the house, which may be a few yards or several hundred yards and could also require rewiring inside the building. "Nobody in their right mind wants to take fiber from the curb to the home," says Scott Cleland, a telecom analyst at research firm Precursor. "It's a huge cost with zero benefit. You simply don't need that much speed attached to the house."

What do you do instead? Some telcos--particularly in Canada--have juiced up their high-speed Internet technology (called digital subscriber line, or DSL) and video compression to offer TV and movies-on-demand to customers via ordinary copper lines. Some pundits say Verizon ought to wait a couple of years for emerging wireless technologies, which may be capable of handling video. The telcos that are looking into fiber to the home, including BellSouth and SBC, are taking a wait-and-see approach.

A big reason that Seidenberg is going with fiber-to-the-home is regulation, not technology. Last year the Federal Communications Commission indicated that it will not force the Bells to give access to competitors on fiber networks that run all the way to the home. That may not be the case with networks that stop at the curb. "That makes no sense to us at all," says Bill Smith, the chief technology officer for BellSouth, which has run fiber to the curb of one million homes in its area. "Fiber to the curb is the functional equivalent of fiber to the home, and in many cases it's much more economical to deploy." Bottom line: There is plenty of regulatory uncertainty, and that's a big risk.

How long would it take Verizon (which, by the way, is the most highly leveraged of any Bell but Qwest, with $45 billion in debt) to make money on this venture? Stock analyst Kalla has constructed several return-on-investment models. If Verizon is shooting for an optimistic return-on-capital rate of 12%, it will take ten years to pay back the $1,000 investment per home, assuming the company gets consumers to sign up for at least one new service--like video--at $50 per month. Under a more pessimistic scenario (a 6% return on capital), it will still take five years, Kalla estimates. Historically, the Bells have undertaken projects only if the wait for payback is less than four years.

This is all deja vu to Verizon's executives. "When we told analysts we were going to spend $4 billion a year in capex [capital expenditures] for our wireless group, they were merciless," recalls CFO Toben. She leans back in her chair in her sunny 39th-floor office in Verizon's headquarters, dramatically rolling her eyes. "They all pointed out that Sprint was only spending $2 billion a year and AT&T wasn't spending as much. Well, not a one of them bothers me anymore about the capital budget there. The quality of the network was the strategic advantage, and it needed the investment. It's the same thing with fiber-to-the-premise."

She's right in saying that Verizon's wireless gamble is paying off. "By far and away, Verizon Wireless is the quality leader in the wireless market," says Roger Entner of research firm Yankee Group. Wireless now accounts for 33% of Verizon's total revenues. Last year, operating income for the group climbed 12%, to $4.1 billion. Even with continuing consolidation in wireless, most analysts say that Verizon's huge quality advantage will keep it on top for years as new wireless broadband and data services are introduced.

Seidenberg is still spending heavily on the wireless network. This year alone he plans to pour another $5 billion into juicing it up. Half a billion of that will go to EVDO, a product the company launched last year that will give laptop computers DSL-like speed wirelessly, as well as improving gaming and picture quality on cellphones. "This wireless broadband service is a very disruptive technology that will let us cross over into new markets," says Verizon Wireless chief marketing officer John Stratton. For example, suppose consumers could download music onto their phones, he says. "Look at the success Apple has had with the iPod. The music industry is an $11 billion business that we think is ripe for the picking."

When asked to describe what Verizon will look like in 2010, Seidenberg momentarily looks shocked. Then he refuses to answer. "Do you think Wal-Mart had a ten-year plan? I doubt Wal-Mart had a department of strategic planning in Bentonville, Ark., that was designing its current model. But my guess is the airlines probably had strategic planners, and the world changed around them." The communications industry changes so fast, Seidenberg says, that Verizon is focusing its energy on figuring out the market, the customer, and the technology over the next couple of years--and then trying to be as nimble as possible.

That strategy certainly applies to the fiber rollout. Seidenberg notes that Verizon isn't "calling out the Army Corps of Engineers" to upgrade the entire network immediately. "We're going to give this a couple of years and test our case. If we're not right, we'll make adjustments," he says, as the cafeteria slowly empties. "But we're not going to be waylaid before we get started. And if it's the right strategy, we'll deploy it a lot faster than you think."

FORTUNE.COM SUBSCRIBERS ONLY Read more about Verizon online, including its complete FORTUNE 500 stats.

FEEDBACK jcreswell@fortunemail.com