(FORTUNE Magazine) – "I'm glad you're here," Vivek Ranadivé, the CEO of Tibco Software, says to me as he bounds up the stairs of his Palo Alto headquarters. "You're going to watch me sell to Marc." Marc is Marc Benioff, the CEO of, a five-year-old company, half the size of Tibco, that offers a web-based program companies use to track and analyze all of their sales and support functions. Ranadivé ushers us into a conference room, where he has a team ready with a presentation on why needs to rewrite itself using Tibco's recently acquired application development software. Almost as soon as the demonstration starts, though, it's clear that Ranadivé wishes I weren't there. He's not selling; he's getting sold.

"This is much bigger than you think it is," says Benioff, cutting off the presentation. "This is a tremendous opportunity for Tibco. But you need to build a five- to ten-year plan to really take this seriously." Benioff leans forward. "If you do that, you'll take your company to another level. This is your greatest opportunity to go from doing $500 million worth of revenue to getting into the $1 billion to $2 billion club. Vivek, this is your opportunity. And we will help you."

Benioff leaps to a whiteboard, grabs a marker, and sketches the details of how the system works, how an "ecosystem" of companies selling add-ons to his software is growing around it. Tibco shouldn't try to hawk a product to him, Benioff says; it should transform itself into a partner, selling to the 12,500 companies using He talks about Tibco's joining a worldwide road show, about creating "webinars" together, about how Tibco can thrive by grabbing onto the rocket. Then he slyly suggests that Tibco could pilot its new tool in-house--all it would need to do would be to expand its use of

"Okay, we'll do that," says Ranadivé quietly. His bravado gone, he sinks into his chair, holding his head up with his hand. Benioff beams. He has just laid the groundwork to add another 100 users to the 400 at Tibco who are currently subscribed--and he wasn't even trying.

There is no salesman in Silicon Valley quite like Marc Benioff. Benioff overwhelms a room when he enters it, standing 6-foot-5, sporting a slightly deflated Bob's Big Boy haircut, and usually wearing clothes that make him look even bigger: a perfectly tailored suit one day, a bulky letterman's jacket bearing the Great Seal of the U.S. (in honor of co-chairing a federal tech advisory committee) the next. He delivers his pitch by shifting between dry humor and quiet, near religious fervor. Benioff is an evangelist not just for his product but for the way his product is delivered, a process called "software on demand" or "software as a service." The idea is that instead of buying software from a big vendor like Microsoft or SAP and installing it on-site, companies simply pay a monthly fee and access the software over the web.

It hardly seems radical. But in Benioff's world this is nothing less than the democratization of software: allowing anyone with an Internet connection, a browser, and an average of $75 a month to get the same kind of productivity-enhancing technology available to the biggest companies. isn't about just offering salespeople new tools, it's about bringing programming to the people.

"We feel an obligation to prove the model," says Benioff. "If we don't do it, who will? If not now, when? This is something the industry needs, the economy needs, and the community needs. I feel that obligation every day."

Benioff has made an art of drawing attention to his company and ideas. He has hired protesters to demonstrate outside of archrival Siebel Systems' events--and even hired fake news crews to "cover" them. The B-52's performed at's launch party (at the time, the company had no customers). David Bowie sang at a fundraiser sponsored by the company, and Arnold Schwarzenegger helped kick off an update of the service. Benioff sends out countless trinkets to journalists--T-shirts, footballs, chocolates (all with's Ghostbusters-style logo: the word "software" surrounded by a circle with a line through it).

Customers and potential customers get calls, e-mails, free brainstorming sessions, and invitations to functions--a party, perhaps, or a fundraiser for the Tibet House, where Benioff was on the board of directors. Spirituality is a personal quest for most people, but Benioff sees it as intricately tied to his promotion of software on demand. A student of Buddhism, he once distributed a poster showing the Dalai Lama sitting cross- legged below the inscription THERE IS NO SOFTWARE ON THE PATH TO ENLIGHTENMENT. It didn't play well in Tibet.

For years, this larger-than-life attitude has caused people in tech and sales circles to dismiss Benioff as a showman, though one with a handy product for small businesses. Now, after the company's late June public offering, the detractors have quieted a bit. His company, while still small, is the largest in a smoking-hot field. And it's getting larger. In the past 12 months Cisco, Staples, SunTrust Banks, AMD, and Nextel have signed on for thousands of subscriptions each to And the small to medium-sized firms that were early believers in are still coming knocking: The company is averaging 7,500 net new subscribers a month and is expected to finish the fiscal year (which ends in January) with 220,000 users. That should translate into an 80% jump in annual revenues, to $174 million, and a hike in net income, to $6 million. Investors are exuberant: The stock traded recently at $18--a feverish 138 times expected 2006 earnings--making Benioff's 28% stake worth more than $500 million.

But Benioff's sales skills cut another way as well: Big competitors haven't just grown tired of being upstaged by him, they've started to believe him. Suddenly on-demand has become the battle cry of the traditional software industry--and it's clear that's influence is much larger than its size. The company's biggest competitor in so-called customer relationship management software (CRM), Siebel, is increasingly pushing its Siebel OnDemand and has revamped its sales and support groups to go after's base of small and medium-sized businesses. IBM, a company that in the past two years has all but replaced its iconic "Think" slogan with "On demand," recently announced plans to convert some of its Lotus products into services available over the web, starting with Lotus web conferencing. "This is going to be much more than a small but important area for us," says Mike Riegel, the director of on-demand strategy for IBM Global Services. Even SAP, the German giant that's known for its six- figure, incredibly complicated business-management software, is gingerly entering Benioff's world. This fall it said it's offering an on-demand version of SAP for customers in four narrow markets, with more to come.

This is disruption in action. Spending on on-demand software is expected to grow at a compounded 25% a year to $9 billion by 2008, vs. just 5% to 6% for the rest of the industry, according to tech researcher IDC. Says Jason Maynard, a senior analyst at Merrill Lynch: "The software industry is going to be radically different in the next couple of years; it's going to be wild."

The question is whether--as the industry catches on and copycats multiply--Benioff can actually dominate the vision of the future that he so convincingly paints. "We want to lead people out of Egypt and into the Promised Land," he jokes. But then, Moses never actually made it in. So Benioff is working hard to transform both himself and his company from irreverent outsider to respected--and entrenched--vendor. In the past year he's targeted the biggest companies in business, hired seasoned executives from places like IBM, Oracle, Siebel, and Sun, and encouraged third parties to build expansion applications that would take too much time and too many resources for to do itself.

Benioff has even started offloading the job of proselytizing to his partners and customers, who seem to relish the opportunity to Be Like Marc. "We have fervor and we have become evangelists," says Mike Rosenfelt, a senior executive at e-mail- disaster-recovery firm MessageOne, which not only uses but sells its own services on demand. "Marc has done a great job promoting the whole No Software thing. This is the future."

It may seem strange, all this excitement over software. Or in the tortured logic of the on-demand crowd, no software. It helps to understand the pain wrought by software vendors over the past few years.

Think of buying corporate software as akin to purchasing a fix-me-up mansion. First a company spends millions for the product --an accounting system, say, or a program to track inventory. Then the real spending starts. The company needs to buy hardware to run the software and bring in consultants like Accenture to customize the product and enable it to talk to other internal systems. For access to upgrades and 24-hour support, the software vendor also levies an 18% to 20% annual "maintenance fee" on the contract price. By the time the system is in place--and it can take years for that to happen--companies pay $6 to $8 in extras for every $1 they spend on the software, says Merrill's Maynard.

Software rented on-demand removes much of the pain. Customers sign up for the software and pay a monthly fee for each employee who needs access. For new or small companies, setup time can be a matter of minutes; larger companies might take months transferring old data or connecting to old systems. On-demand companies host all the data and software on their own hardware at centralized locations. That allows them to achieve economies of scale, as well as to perform bug fixes or major upgrades on all customers at one time, meaning every subscriber is automatically working on the latest available version. Plus, a company division can go on-demand without even checking with the IT staff: Since the service is web-based, all that's needed is a browser. No haggling over compatibility, no need for capital expenditure requests.

Finally, while a traditional software seller's job is largely done when the contract is inked, on-demand players get paid by getting the most people in a company on their software--and keeping them there month after month. That means doing whatever it takes to make customers happy, which helps explain Benioff's constant handholding and vision-painting for his clients--and why sales and marketing costs, at $25 million last quarter, consume 50% of's revenues. (SAP, by contrast, spends just 20% on sales and marketing.) "If a trial account calls me and complains, I'll deploy a whole team for that person," Benioff says.

For all its benefits, on-demand software isn't perfect. Slow or nonexistent Internet connections can mean a lack of access to key data; on-demand systems aren't always as customizable as big software; and the data--instead of being locked up behind a client's firewall--rest on a faraway server. Benioff allows companies to do audits of his systems, which have met the approval of security- obsessed clients like data-protection company SunGard. Yet worries still remain. Take ADP, the payroll-processing company with $7.8 billion a year in revenue. It moved all its salespeople onto this summer (choosing it over Siebel) and expects to have 5,000 users by the end of June--Benioff's goal is to have all 42,000 ADP employees on the service. Yet Mike Capone, the vice president of corporate information services, says that he's not sure just how far he's willing to go. "Something that is the lifeblood of your company you have to be careful with," says Capone. " is doing a great job, but they have a little bit more to prove before I'm going to bet my entire customer service and support for something like payroll."

Still, in a time of shrunken IT budgets, the lure of paying a monthly subscription fee instead of a million dollars up front is hard to resist. Which is why Benioff is going hard after users today and worrying about costs later.

"We need to lead the industry in market share at all costs," says Benioff. "This is one of the mantras that has guided us for the past five years, and I think that this still applies today." He has gathered his top 25 lieutenants for a weekend at CordeValle, a golf resort outside of San Jose, to draft an attack plan for the next two years. Benioff is dressed in jeans and a tucked-in, long-sleeved T-shirt bearing his software-busters logo. Only recently turned 40, Benioff confidently lays out his vision to people who, in many cases, far exceed him in experience and age. There's co-president Pat Sueltz--the former head of services at Sun Microsystems and before that a 20-year IBM veteran. Next to her sits the other co- president, Jim Steele, who worked for 23 years at IBM, heading up sales for the Western U.S.; down the table is CFO Steve Cakebread, a veteran of Autodesk, Silicon Graphics, and Hewlett-Packard.

The goal for the next two years is bolder than ever. Benioff plans to win market share not just by stealing accounts from competitors and signing up new ones but also by snagging more and more subscribers from within current shops. There's no reason, Benioff says, that the data customers keep on can't just keep growing until becomes their de facto on-demand operating system. "If [customers] do project management they're going to use us; if they want a legal compliance system on-demand, they're going to use us," he says. "I want us to be the standard way customers share their information. Big goal, big dream, big vision."

That, say friends and co-workers, pretty much sums him up. Benioff grew up in Hillsborough, Calif., in the heart of Silicon Valley. His parents owned a women's clothing store, but Benioff was drawn to tech. He spent his free time in high school crafting programs, first for the TRS-80 computer and later for the Atari 800. His first program: How to Juggle. "You've probably heard of it," he says dryly. "It became the industry standard for the juggling community." He sold it for $75 and never left the software business. He churned out enough games to pay for a Toyota Supra and put himself through the University of Southern California.

In college he worked as both a programmer and salesman for Apple. If that company taught him the benefit of an obsessive fan base, his next job taught him everything else. At 21, Benioff joined Oracle and quickly rose through the ranks: at 22 he was named Rookie of the Year; at 25 he became the company's youngest vice president; later he ran marketing. Over the years, he developed a close relationship with CEO Larry Ellison.

It was clear why: Benioff was Ellison 2.0. "Marc would get people believing in his vision and supporting whatever project he wanted to do," says Ray Lane, former Oracle president and now a partner at venture capital firm Kleiner Perkins. "That's Larry." At global management meetings, Lane would have Benioff and other development managers come to discuss their projects. "These other guys would put the crowd to sleep," says Lane. "When Marc talked, these people would jump to their feet and say, 'This is what I want to sell!' "

Benioff became close to another senior executive at Oracle as well: Tom Siebel, who quit and started Siebel in 1993 with an early investment from, among others, Marc Benioff. In 1996, Benioff started to get restless himself. Rich but feeling lost, he took a sabbatical with Ellison's blessing and traveled through Asia and India, becoming interested in Eastern religion. When he returned to Oracle six months later he was convinced that he needed to give back to the community--and start something on his own. Over the next few years a buzz developed around software on demand (companies providing it were called "application service providers"). Ellison thought the idea had promise and talked about it with employees, including Evan Goldberg--who went on to found competitor NetSuite --and, of course, Benioff. In 1999, Benioff scaled back his work at Oracle, and with encouragement--and $2 million--from Ellison, he started planning his company.

Benioff's first idea was to build it in partnership with Tom Siebel, whose company was selling CRM software, but the kind installed at companies. Siebel wasn't looking for a partner; he proposed that Benioff join as an employee. Recalls Benioff: "I said I'm already working for one billionaire S.O.B. I don't want to work for another one." Benioff, who had sold most of his $50,000 investment in Siebel for more than $25 million, took $6 million of that, combined it with Ellison's stake, hired three programmers, and in March 1999 started That same week Siebel an- nounced its own online offering, a sales portal called

Benioff quickly set himself apart from his peers. He bucked the dot-com trend, throwing his kickoff party in February 2000, a month before dot-coms started keeling over. And he made giving back central to the company: He put 1% of stock into a philanthropic foundation; strongly encourages employees to spend 1% of their work time doing charitable service; and donates 1% of the company's profits to charity--mostly in the form of granting nonprofits free access to

He also started to distance himself from his mentor. In 2000, Benioff discovered that Oracle had started its own on-demand CRM application. He insisted Ellison leave his board. When Ellison refused, Benioff aired his problems to the Wall Street Journal, putting pressure on Ellison--and also building some media buzz for the young company. Eventually the two struck an agreement: Ellison left the board but kept his stock. Today he owns about four million shares, which he and other insiders can begin selling in December.

Benioff says he and Ellison are friendly these days. The same, though, can't be said of his relationship with Tom Siebel. Benioff, always sarcastic, seems to reserve a special mocking tone just for his company: "How many large customers do they say they have on Siebel OnDemand?" he asks, his voice getting higher. "Oh, they don't mention it? How many subscribers do they say they have? Oh, they don't mention that either? What about strategic wins over us?" Soon he's diagnosing problems with the company, which at an expected $1.3 billion in revenue this year is over seven times the size of "Siebel's failed entry into our market over the past year has been the biggest catalyst to our growth so far. Their product is terrible. And they're not serious about it. They don't have the religion. Where's the fervor? Where's the excitement? Where's the user adoption?"

Mike Lawrie, the CEO of Siebel, is not about to take Benioff's bait. But in the best tradition of IBM--Lawrie's employer for 26 years--he will offer a genteel putdown: "Listen, I think there's certain skill sets that are required to get businesses going, and I think those skill sets change over time as companies mature and as the business model shifts," he says. "So what's right today or tomorrow may not be right five years from now. But listen, the guy had a vision and an idea, stuck with it, and has done pretty well. So let's give him his credit."

By that he means: let's bury him. In May, when Tom Siebel stepped aside and handed the CEO reins to Lawrie, something needed to be done. The company had lost over $45 billion in market value over the past four years as it dealt with a slowdown in IT budgets, damage to its brand by stories of failed implementations, and a failure to adjust to the on-demand boom. After shutting down in 2001, Siebel spent two years watching steal its show. Finally, in October 2003 it bought Upshot, a smaller competitor, combined it with its own R&D and at the end of the year launched Siebel OnDemand. "Tom ignored Marc for a long time," says former Oracle president Lane. "He thought, 'This little $100 million company, I'm going to do away with it.' He was in denial up until the end." (Tom Siebel wouldn't comment.)

Lawrie is under no such illusions and has brought in one of Siebel's first employees, Bruce Cleveland, to be his Benioff beater. Cleveland's approach is to take the wind out of the hype. If Benioff is a preacher, Cleveland is a scientist, arguing that on-demand isn't heaven-sent; it's just a different way of offering software. "The basic concept is one product, two deployment versions," he says. "When people want a CRM system, they don't think, 'I need a hosted CRM application' or 'I need an on-premises application,' they think, 'I have a business problem.' "

So Siebel is offering both and says the two are basically interchangeable. Pulling off that strategy won't be easy. Cleveland has worked to make the on-demand version as functional as the on-premises version, something he says will take at least another year to complete. He has had to change the mindset of its sales team; under Tom Siebel, it made much more selling on-premises software than OnDemand. And he's set Siebel apart from by crafting industry- specific versions of OnDemand instead of's approach of leaving it up to users to customize themselves.

One of Cleveland's best--and worst--defenses against, though, is Siebel's giant base. When Ingersoll-Rand, a $10 billion in revenues Siebel customer, discovered that its Kryptonite locks could be opened with a ballpoint pen, it ordered a recall and needed to get an instant system built to help track complaints. The company went with what it knew, bringing in Siebel OnDemand. But there are plenty of other cases where Siebel users are just desperate to get away from the company. At a recent Siebel conference in the basement of the Four Seasons Hotel in New York City, one customer who had come hoping for help walked out fuming. "It's been a disaster," said Tina DiMare, marketing services manager at legal supplies company Blumberg Excelsior. "We purchased Siebel four years ago and haven't seen any return on our investment. I've never even spoken to Siebel." Would she consider using Siebel On-Demand? Not a chance, she said.

That's the sort of answer that gives Benioff confidence that Siebel--or any other traditional software vendor that has burned customers in the past--will never be able to stop him.

In a drive through Silicon Valley in Benioff's Porsche SUV,'s vice president of European operations, a seven-year Siebel vet named Phill Robinson, explains the difference between his employers. "I had to relearn marketing from the basics," he says. "At Siebel it was target marketing: You figured out the ten companies you wanted to sell a major transaction to. Here, the idea is that one guy likes it and tells ten of his friends. What I'm doing is trying to stoke up the viral effect."

"To build up the good karma," corrects Benioff. "And at Siebel, you were just building up the negative karma. It wasn't your fault."

We're on our way to meet with a customer, Nuance, which, like Tibco, is a user that's now trying to sell Benioff. This time it's a suggestion that Benioff buy Nuance's voice recognition software. But when we sit in the office, Benioff starts hammering away at Chuck Berger, Nuance's CEO, about his personal involvement with When Berger says that he doesn't touch it, Benioff says he'll send a group down to get him personally set up. Even better, he'll have co-president Steele walk Berger through it himself.

As for Berger's goal of selling to, Benioff persuades him instead to build a voice recognition product that will tie into Salesforce, then sell the product as--wait for it--an add-on, on-demand service for users. If it works out, Nuance will be not just selling its own products but Salesforce's as well. And if Nuance users like it, they'll tell their co-workers, and so on.

"We don't have a couple hundred salespeople; we have a couple hundred thousand salespeople," he says later. "They're called our customers. And it would be really cool if they were all out there chatting it up, telling their friends, because it takes a really long time to go out and visit all of the customers." It's hard to escape the feeling that Benioff's not going to try.