(FORTUNE Magazine) – Richard Frank is only 20 minutes into the first meeting of the day when he loses his temper. That comes as a surprise to everyone in the room because Frank is not a man given to fits of pique. A 30-year veteran of the entertainment industry and onetime head of Disney's film studios, Frank is naturally soft in his voice, his clothes, even his posture--someone easier to conjure up tasting wines on the patio of his Napa Valley estate than screaming about statistics in a meeting room.

Yet it happens that Frank--whose distaste for the nuts and bolts of technology is so strong that it's almost a point of pride--now has the misfortune to find himself running an Internet company, Food.com, a Web portal for everything from ordering pizza to learning how to roll sushi. Thus it is that he is holed up with his chief operating officer and his chief financial officer doing precisely what he likes least, which is to pore over columns of numbers, crinkle his brow, and try to understand why the hell it is that however much content his team adds to the site, the average visitor still looks at only a pathetic 2.8 Web pages before going somewhere else. Frank says he has a "gut feeling" that the numbers from the measurement service he hired are wrong, but he can't prove it. In fact, his own research head has just told him, with a certain amount of hemming and hawing, that the service's figures match up exactly with Food.com's own electronic logs.

Frank asks for more and more detail, as if hoping to find a hole that will let him trash the numbers. He goes over one bit of technical arcana after another and asks what the measurement company has to say in its defense. The COO responds that his calls and e-mails of complaint have gone unanswered. And that is when Frank loses it.

"Here's what I want you to do!" Frank spits out. "Draft a press release. Have it say that Food.com is dropping [the measuring service], and here's why, and we want everybody to know about it. Then send [the service] a draft of the press release and see what they say."

Everyone shifts uncomfortably in his seat, contemplating the silence that is likely to meet a press release of this sort if it ever sees daylight. Is there any chance it would make even the trade magazines? No. Alas, Food.com is not Disney, and the little-known measurement company is not Nielsen. To a man used to fighting his battles in the press, this must feel like an awfully small canvas. After a minute even Frank seems to give up. He looks with kindness at the research head and gently asks her, in a sadly Dilbert-esque moment, to analyze the numbers again.

Welcome to the scary world of the big-company executive who has gone dot-com. Frank has become a character in one of the myths of the Internet Age: that a startup can transform itself into a real company--the sort that sells things to customers and hires a staff of thousands and books millions in earnings and pays regular dividends--by hiring a CEO from an old-economy company, someone who knows how to preside over such sacraments. Old-world CEOs manage people and processes. They command and they control. A dot-com inoculated with their managerial DNA will become a different, better organism.

Would that it were so simple. While success stories of managerial transplants from the old economy do indeed exist, the adaptation is a hard one for both the graft and the host. Forget about the absence of chauffeured limousines and armies of administrative assistants in the dot-com world. More nettlesome is the tension between the fast pace and spontaneity that spawns and sustains startups and the mastery of formal managerial techniques that won the incoming CEOs their last big job in the old economy. Now these seasoned executives are in a world where they must rethink all their assumptions about not just little things like when to send out press releases but all the big things--strategy, technology, management style--that go into running a company.

Frank is one of a small cadre of executives recruited from the upper reaches of the FORTUNE 500 to run fledgling dot-coms. In the peculiar argot of the Net, they have become known as "adult supervision," grownups tasked with the care, feeding, education, and (not least of all) funding of Internet companies presumed to have outgrown the "visionaries" who got them up and running. Sometimes they are brought in, as was Frank, by venture capitalists to turn around a floundering company. At other times they come--as was famously the case with Jim Barksdale at Netscape--to apply rigorous business thinking to a company that has a killer technology and no definite plan to turn it into cash. In still other cases the experienced CEO plays a role akin to that of an escort at a debutante ball, doffing his cap to investors and shepherding a promising dot-com to the public markets; that was the part played by George Shaheen, who left Andersen Consulting for Net supermarket Webvan just weeks before its initial public offering.

In its first blush the liaison can bring riches to all parties involved. To investors and customers the arrival of a hotshot FORTUNE 500 CEO is the signal event that validates a dot-com, proving that it's the real thing and will probably stay around. A prominent recruit adds an instant sheen of respectability that can help startups forge critical alliances with big established companies. At Food.com, Frank raised $80 million from a group of powerful strategic partners, including McDonald's. For the incoming executives, the lure is the prospect of staggering wealth. Meg Whitman, the former FTD CEO and Hasbro division head who is now CEO of eBay, has amassed more than $1.4 billion in eBay stock since she joined the online auctioneer in 1998.

Though melding different cultures is no easy feat, some dot-coms have struck gold after bringing in new talent. The trick is finding a CEO whose background and temperament mesh nicely with new-economy enterprise. Whitman's success at eBay shows how well things can go when the fit is right. Her experience at a major consumer company has clearly been a plus. Says Robertson Stephens analyst Lauren Levitan: "Being the steward of the eBay brand is something that Whitman has done very admirably. A brand is fragile; [eBay needed someone] who could nurture it." Whitman has brought that experience to bear while embracing the startup culture of a Net company. Her bio on eBay's Website is telling. She lists her most important contribution as "develop[ing] the work ethic and culture of eBay as a fun, open, and trusting environment." Over the past two years Whitman has been having plenty of fun. She has handily defied the pundits and expanded eBay's franchise in the face of competition from Net giants Amazon and Yahoo, building quarterly revenues from $9 million to $97 million and turning a profit of $11.6 million in the last period.

Bringing in a CEO can be an emotional challenge for a dot-com because the recruit is often the first person other than the founder to run the company. But if the handoff comes at the right time, a startup can begin to capture more than mind-share. Such was the experience of Net computer retailer Cyberian Outpost. "The biggest mistake that founders make is that they run their companies for way too long," says Darryl Peck, Cyberian Outpost's founder and chairman. "I thought there had to be somebody out there more qualified than me to run this as it grows into a billion-dollar company." Peck says that while he is himself "obsessed" with numbers and measurement and could always see which parts of the company's operations needed to be improved, new CEO Robert Bowman (the former head of conglomerate ITT) is a lot better at implementing the necessary changes. Since Bowman came in nine months ago Cyberian Outpost's flagging quarterly revenues have revived, nearly doubling to $65 million, while marketing expenses (the bugaboo of Internet retailers) have stayed almost flat--no mean feat in a desperately competitive business.

Most dot-coms, though, don't have nearly as happy an experience with their FORTUNE 500 recruits. In fact, the short history of the Internet is already filled with examples of big-company executives who never adjusted to their new Net gigs. NBC's network president, Neil Braun, headed an online entertainment company, iCast, for less than a year before being deposed in a clash with its chairman and chief financial backer, David Wetherell. After William Harris left his post as CEO of Intuit, the No. 2 consumer software maker after Microsoft, he landed at online bank X.com. Running an ambitious Internet banking venture seemed like a natural step for a man whose experience straddled finance and technology, but Harris lasted only six months before getting fired by the board of directors. According to insiders, he created lots of structure and held endless meetings, virtually none of which yielded any resolutions. Decision-making slowed to a crawl. Harris would not comment for this article.

Many of the old-economy recruits run aground because of a stubborn refusal to relinquish old-economy notions of command and control, which can stifle growth at smaller companies and deflect attention from fast-changing markets. Says Paul Bracken, a professor at the Yale School of Management and a frequent consultant to startups: "The dot-coms are structureless. They can show you their structure on a sheet of paper, but it doesn't really mean anything. They're used to coming in on, say, a Wednesday morning and changing the structure entirely." That's not necessarily a good thing, but cumbersome new procedures can give a young dot-com hardening of the arteries.

For instance, look at the problems that Shaheen has run into as CEO of Webvan, the online supermarket. With weekly operations meetings and biweekly management committee meetings, Shaheen has given Webvan exactly the kind of management infrastructure that dot-coms are supposed to lack. He makes the process of management sound straightforward, offering the kind of perspective that once helped the big corporations he consulted for run...well, run like big corporations. But has this made Webvan's corporate decision-making better? Maybe not.

In July, shortly before Webvan was to announce its quarterly earnings, Shaheen exuded confidence. "Analysts develop their models," he calmly said, "and then they sit back and wait to see if you deliver." But just a few days later Webvan missed its numbers, announcing earnings that fell just short of analysts' projections.

What went wrong? Thomas Weisel analyst John Weiss says Webvan has been unable to control its distribution costs (according to Weiss, they're already running at a level Webvan didn't expect to reach until its revenues had tripled). He thinks that Webvan could reel in costs by making relatively simple changes, like promising deliveries in a one-hour window instead of half an hour. Shaheen, however, thinks of the half-hour delivery window as one of Webvan's "basic tenets," the kind of thing that he hasn't been willing to touch. Instead, he has chosen to focus on complex issues of logistics; asked to name an issue that has engendered debate in the company, Shaheen recounted a discussion of the size of Webvan's trucks.

This narrow focus on metrics and tactics over strategy has been costly. Webvan's already battered stock fell 20% in response to the poor earnings. The company's reaction has been to deny that there is any problem. A spokesman insists that Webvan's quarterly results were in line with its own expectations and the guidance it gave analysts.

In the worst cases of dot-com meltdown, old-economy assets can be major liabilities. Consider David Dorman, the onetime head of Pacific Bell who took on the task of running Pointcast. Having arrived in 1997 at what seemed to be a leading company of the Internet Age, Dorman quickly found out that he had a huge problem on his hands. Pointcast's software, which was supposed to stream news and entertainment to users over the Net, routinely caused computers to crash. Dorman, as well as others at Pointcast, saw that the company needed new software and a new strategy.

Unfortunately Dorman came up with the ultimate big-company solution. Drawing on his powerful telecom connections, he opened negotiations with a consortium of telcos that hoped to adopt Pointcast's technology in a portal for their high-speed Internet-access services. After seven months of negotiations--months that Pointcast couldn't afford--the deal died. Having pinned all hope on its CEO's contacts, Pointcast was left nearly out of cash and, in Dorman's own words, "out of runway." Dorman left to head Concert, a joint venture of AT&T and British Telecom. The remnants of Pointcast were sold for a reported $7 million, a tenth of what investors had poured into the company.

As they negotiate the byways of the strange new world they're joining, migrants from the old economy seeking a model should look to Yahoo CEO Tim Koogle, arguably the most successful head of an Internet company in Silicon Valley. Before taking over Yahoo, Koogle ran a midsized technology company--not the high-profile big-brand experience dot-coms so eagerly hunt for these days. Koogle argues that because of the fast pace of Internet commerce, corporate management techniques are morphing swiftly and dramatically. "A generational change is under way," he says. "We're moving away from command-and-control to distributed decision-making."

Koogle has created a management structure that lets him share power and responsibility with his staff without losing control. At Yahoo--a company that almost everyone agrees functions quite well--there is no such thing as a weekly management meeting. Koogle has traded it in for a more wide-ranging and yet equally rigorous analysis of the company's operations. Every week 30 top managers receive an e-mail from finance summarizing all the key measures of Yahoo's operations and the major executive decisions of the period. But the managers don't send a weekly report to Koogle. Instead they carry on a running dialogue with him over e-mail in response to the weekly numbers. This free-flowing exchange of data and ideas keeps Yahoo nimble. Senior managers rarely find themselves in the position of waiting to receive dictates from the CEO. Koogle says he "zooms in and out" of many meetings to understand what's going on in Yahoo's divisions, but never second-guesses a manager's operational decisions. Yahoo's executives feel more accountable because they aren't forced to push the ultimate responsibility upward to the CEO. In Koogle's words, they "own the results."

In running Yahoo, Koogle draws from his experience but never seems confined by it. Before joining the company in 1995, Koogle ran Intermec, a Seattle maker of handheld industrial computers. It was working at Intermec, which has a network of overseas subsidiaries, that convinced him Yahoo should go global and establish international beachheads early in its life. He has also successfully negotiated the delicate task of taking the rudder from founders David Filo and Jerry Yang without making them useless. Both have remained very much involved in the business. Recently Yang, always Yahoo's most popular public face, has been spearheading the company's drive into Asia.

As the dot-com casualties pile up, the furious search for seasoned management is likely to intensify. Most FORTUNE 500 recruits, however, aren't likely to prove as skillful as Koogle in culling from their past. All CEOs reason by analogy to circumstances they have encountered before, so the hardest struggle is unlearning old-economy lessons that no longer apply.

Even for the most eager converts to the Net, it can be touch-and-go. Nearly two years after joining Food.com, Richard Frank is still struggling to figure out how to apply his Disney experience at his new job. A man who has spent his career in media, he is most comfortable when he's talking about creative content. The pity of it, as Frank himself repeatedly tells his bankers, is that the essence of Food.com's business is not about creative content. It is about databases of recipes, about alliances with companies like Kraft, about polls and surveys. And about ordering food--that, ultimately, is where Food.com hopes to make most of its money. In discussions with bankers Frank studiously avoids talking about content, a goner in today's investment marketplace. He points to a deal to buy the food-delivery service Takeout Taxi. He carefully goes over the details of Food.com's blockbuster deal with Papa John's pizza, which is using Food.com to put its thousands of franchisees online--and giving Food.com a cut of the take.

And yet, as the day ends and he has put the discussions of technology and the meetings with his bankers behind him, he forgets himself. His last meeting is with the site's two senior editors, and here he is in his element, as comfortable as a producer giving his notes on a screenplay. Somebody proposes a partnership with Zagat Survey, the publisher of restaurant guides. The chief operating officer says that Zagat wants too much money for a partnership. Frank smiles. All thoughts of food delivery and Papa John's and McDonald's go out the window, and he remembers his days at Disney and the national launch of Siskel & Ebert, the syndicated movie-review program. "Rex Reed and the New York Times were the only authorities on the subject," recounts Frank. "We made Siskel & Ebert an authority. We can do our own guide to America's best restaurants. We can make Food.com the authority, not Zagat."

So the whole group, led by Frank the media guy and vineyard-owning bon vivant, is soon off on a discussion of restaurant guides. Quickly the whiteboard fills up with a schematic of not only a U.S. guide but also guides to restaurants abroad. Frank is amused: Why not a food joke of the day, he proposes. The editors duly notes it down. For the first time today, Frank is really psyched, making up stuff that will bring people to the Website, business model be damned. And just for a minute Frank seems to forget that he is running an Internet company at all, forgets that he has personally poured millions of dollars into it, forgets all the big questions about strategy that he is supposed to be dealing with. "I can write the dining guide for Napa," Frank blurts out. "I'm not kidding about that."

FEEDBACK: mgimein@fortunemail.com