Dot-coms What Have We Learned? Dot-coms soared. Dot-coms crashed. Somewhere in between we uncovered 12 truths about how the Net really changes business.
By Jerry Useem Reporter Associate Grainger David

(FORTUNE Magazine) – Long, long ago, back when VC stood for Vietcong, an ASP was an ill-humored snake, and Andre Agassi had hair, a new and exciting thing came into being. It grew to be known by many names--the Information Superhighway, the World Wide Web, the Internet, Great Supreme Digital Being--and its promise spread its wings the way Promise is spread on a really large piece of toast. This new thing was so deeply wonderful that it would boost planetary wealth and alter the business landscape with all the impact of an asteroid. Startups rise. Giants fall. Hope survives.

What a cruel, cruel illusion it all turned out to be. Stoked by the false promise of office foosball and a lot of irrational exhibitionism, the dot-com phenomenon proved to be shot through with phoniness--an apparition within a hologram wrapped inside two specters of a mirage, with some tulip mania to boot. As for why anyone ever thought doing business on the Web was a good idea, search us. More like Amazon.bomb!

There you have it: the gloom-and-doom story line that these days passes for conventional wisdom. Like the hysteria that preceded it, it sheds little light on anything.

Here's what we do know: The dot-com era is over. The Internet era, by contrast, is just getting started. This we also know: Dot-coms were an experiment--a nexus of hypotheses, really--that tested radical new ways of starting companies, managing them, and investing in them. James Collins, co-author of the bestselling book Built to Last, likens it to tinkering with new forms of engineering. "There are some timeless truths like E = mc2 and F = ma, and if you're going to build a company, you can't ignore those," he says. "It's like trying to go against the speed of light. But what's changing is our engineering--the way we apply physics." Some of the dot-coms' innovations, such as the concept that a business needn't earn a profit, proved simply to be bad engineering. "Bridges fell down," says Collins.

Yet other innovations seem to be here to stay, which is the reason today's conventional wisdom is so, well, conventional. As Paul Deninger, CEO of high-tech investment bank Broadview, puts it, "it is true that the Internet will change everything. It is not true that everything will change." That's why it seems appropriate to take this moment in Internet history--the end of the beginning, some are aptly calling it--to reexamine some of the assumptions that got us so exercised about dot-coms to begin with. For as every Internet type worth his or her casual wear will concede, this period has been nothing if not a learning experience. "What B2C means now is Back to College," says Bruce Richardson of AMR Research in Boston. Agrees Fred Hoar, chairman of high-tech PR firm Miller/Shandwick Technologies: "We've gone to school with this traumatic experience."

So what have we learned? Herewith, a dozen lessons from the dot-com crash.

THE INTERNET ISN'T AS "DISRUPTIVE" AS WE THOUGHT. Early this year Josh Harris, founder of the streaming-media company, stood before a CBS camera on 60 Minutes and declared with certitude, "Our business is to take you guys out. I'm in a race to take CBS out of business.... That's why we're going to make the big bucks."

At the time his bravado didn't seem entirely overwrought. The Internet, it was presumed, was a classic "disruptive technology"--a term popularized by Harvard Business School professor Clayton Christensen in his book The Innovator's Dilemma--that would favor new entrants like and send many old-line companies the way of the dodo and Digital Equipment Corp. It hasn't worked out that way. CBS is still on the air. folded last month.

To be sure, a handful of Internet businesses truly have changed the rules. Few would argue that Napster has done anything less to the music industry. E-brokerages and online sales of airline tickets have roiled their industries. But the vast majority of dot-coms--especially those that sell something physical, like big sacks of pet food--ended up looking decidedly unrevolutionary. "A lot of people assumed everything was--this is a bad word, but--Internetable," says financier Deninger. "They said, 'Has anyone come up with Okay, then let's do it.' " (An Italian company actually owns that Web address.) At the trend's apogee, venture capitalists bought up hundreds of domain names in the hopes of assembling a business around each, and eagerly funded such paradigm-shatterers as

Today it's clearer that not all that glitters is gold. "One filter we like to use is, Is this a business that could have existed before the Internet existed?" says Steve Jurvetson, a venture capitalist at Draper Fisher Jurvetson. "If so, it's probably not worth it." Even Clayton Christensen has had to refine his thinking. He now considers the Internet not a disruptive technology per se but "an infrastructure that can enable disruptive business models." He says, "I've been learning along with everyone else."

IF IT DOESN'T MAKE CENTS, IT DOESN'T MAKE SENSE. Other companies proved not so much disruptive as simply destructive--mainly to capital. While some dot-coms could at least project profitability at some level of sales (albeit a scarily high one; eToys puts its breakeven between $750 million and $900 million in annual revenues, for instance), others seemed to have no plan for making money, ever. "It was astounding to see how few pure dot-com e-tailers had a grasp of their gross margins," says Jim Breyer, managing partner at venture capital firm Accel Partners. "These companies were driven by the premise that at some top-line number, scale would produce profits. But in 99% of the cases, the analysis suggested that many of these companies would never become profitable.... It was a bridge to nowhere." To wit, a recent report from McKinsey & Co. concludes that many e-tailers' putatively "fixed" costs become unfixed as the companies expand, with expenses growing apace with sales.

So why were so many e-tailers funded? Because venture capitalists were flush (see the following story). "When there's a ton of water, a lot of things float and look like boats," says David Roux, co-founder of high-tech buyout firm Silver Lake Partners. "When there's a ton of money, a lot of things float and look like companies."

TIME FAVORS INCUMBENTS. Having failed to overturn the industrial order, the Internet is now looking suspiciously like an enabler of that order. The power of the so-called clicks-and-mortar approach, whereby a corporation integrates its Website with existing fulfillment, logistics, and marketing, looks to be quite real. Noting that 93% of stockbrokers say they consider the Net an enabler, not a threat, e-consultancy Mainspring concludes, "The threat of channel conflict for brokers has been vastly overstated." The new catch phrase: "channel confluence."

Many corporations, moreover, are now betting the Net will lower their costs faster than it will lower their prices. This is certainly true for banks, for whom an online customer is roughly a third cheaper than an ATM customer, and much cheaper than one who uses a teller. Companies such as Cisco Systems have shown the value of digitizing internal processes. And at GE, CEO Jack Welch's internal slogan "" (coined to convey that GE should attack its own businesses before dot-coms did) has morphed into the sunnier "" It may be one of capitalism's small ironies that the dot-coms' once-terrifying stock valuations will serve mostly to goad big companies into benefiting from the Internet.

Internet-only "pure plays," once engineered through spinoffs (such as or through the financial gimmick known as a tracking stock (DLJdirect or Disney's, now look downright naive., for example, began independently but is now merging back into its parent, industrial supply giant W.W. Grainger. "It became obvious to us that we needed interdependence," explains President James Ryan. Meanwhile, dot-coms are rushing to find places in the physical world. E*Trade is opening outposts in SuperTarget stores, while Amazon has allied with Toys "R" Us and other e-tailers are introducing print catalogs., which started life as an allergy-product e-tailer, quit the Web entirely after it opened five physical stores that proved profitable. And in an unusual bricks-to-clicks-and-back-again move, --itself the virtual progeny of Bank One--is tapping into a network of 30,000 ATMs. "Online penetration is going to be small at best. So once you have an infrastructure with fixed costs, you'd better damn well push as much revenue through it as possible," explains Julian Chu, an e-strategist at Mainspring. "In six to nine months," he predicts, "there will be no major retail pure plays left."

As for who will win this rush to the middle--the virtual companies seeking physicality or the physical companies seeking virtualness--we merely ask, Which do you think is easier, learning e-business or learning how to build a multibillion-dollar corporation? "If the Internet didn't change the whole world overnight, obviously time favors the incumbents," says tech investor Roger McNamee of Integral Capital Partners. Agrees VC Bill Unger of the Mayfield Fund: "eToys was an early mover. But Toys 'R' Us gets to screw it up a bunch of times, and eToys doesn't." Of the thousands of dot-coms launched, analysts say as many as 95% could fail.

MAKING A MARKET IS HARDER THAN IT LOOKS. One set of companies that did look very real was the business-to-business, or B2B, exchanges that cropped up in 1998 and 1999 and were promptly awarded mungo market capitalizations. They were supposed to turn the fusty world of industrial procurement into one big, transparent e-marketplace. Yet their stocks have developed the same funereal pallor as their business-to-consumer (B2C) brethren. What happened?

In retrospect, say analysts, most B2B efforts betrayed a pronounced cluelessness about how industrial buying actually works. Start with the supposition that purchasing managers would be thrilled to take bids online from dozens if not hundreds of suppliers, each vying to be the lowest bidder. This has proven true in some cases, such as spot-buying of commodities like excess rolled steel. But for the bulk of spending, corporations have long been moving in precisely the opposite direction, establishing deep relationships with a few favored suppliers in a "total cost" approach. Under this approach, price is but one of a host of criteria, which include quality, cycle time, services, geography, and baseball tickets. "What have buyers been doing for years?" asks Edward McCabe, a B2B analyst at Merrill Lynch. "Whittling down their preferred-vendor list."

Another blithe assumption was that the B2B exchanges could use their hub positions to effectively seize control of an industry's commanding heights. That may happen in highly fragmented industries. But in concentrated ones such as autos and aerospace, according to Broadview's Paul Deninger, "the guys with purchasing power are saying, 'No, actually, I've got the power, thanks very much, and I'm not giving it to you." Determined to prevent dot-com interlopers from coming between them and their customers, corporations responded with a slew of in-house efforts and industry consortiums that have earned the collective moniker "The Empire Strikes Back." (A bonus lesson here: The real threat of the Internet to corporations isn't "disintermediation," as widely proclaimed, but the very opposite--a host of intermediaries.) General Electric CIO Gary Reiner began boasting that he paid two interns just $30,000 to set up an auction site.

Many of the manufacturers' B2B efforts, it turns out, are not really e-marketplaces at all, but merely cheap ways of interacting with suppliers--somewhat less world-altering than the original B2B vision. Again, the Net looks enabling to the old order, not disruptive: "A lot of this is really about automating existing processes," says McCabe.

Add in the fact that some categories have no fewer than 14 exchanges competing for the same dollars, and the result has been a devastating lack of liquidity for most independent marketplaces; hence the deflated stock prices of such pioneers as VerticalNet and Internet Capital Group. The only sure winners seem to be so-called arms dealers such as Ariba and CommerceOne, which create the software for all this market making. That explains why the exchange company formerly known as Chemdex has refashioned itself as an infrastructure builder named Ventro. Exchanges that can't pull off similar sex changes may face extinction.

THERE IS NO SUCH THING AS "INTERNET TIME." Lunging headlong into uncharted territory, the B2B exchanges were, of course, merely taking to heart another new-age shibboleth: that industrial transformation would henceforth occur at the superfast pace known as "Internet time." Yet as tech investor McNamee observes, this concept had problems too. As he sees it, technological change happens in three phases: the creation of infrastructure, the arrival of enabling technologies, and, finally, companies built on the previous two. "A lot of folks jumped the gun," he says. "They tried to skip the first two phases." While relatively simple businesses like eBay could prosper, streaming-media companies found an unready audience, and B2B companies ran into serious technical speed bumps. "The idea that it was all going to get done in two or three years was complete crap," says Joe Costello, CEO of software company Think3.

Certainly, the dot-com explosion has changed our perception of what fast is in business. But "Internet time" now seems less an expression of commercial reality than the confection of people with vested interests: venture capitalists who wanted a quick exit from investments; consultants who wanted to spook clients into buying advice; ordinary investors who wanted to be romanced by it all; and executives who felt pressured to live up to the hype the financial bubble had created. "Venture capitalists, the stock market, day traders--they colluded to create a rush that was not called for," says Broadview's Paul Deninger. Says McNamee: "People used 'Internet time' as a justification for lack of discipline. They got looser and looser in their behavior and got more and more rewarded for it.... In retrospect, 'Internet time' will prove to have been a hormonal thing."

"BRANDING" IS NOT A STRATEGY. The concept of Internet time also spawned a highly abnormal approach to company-building. Since time immemorial, most businesses had grown organically, using operating profits from early customers (think Wal-Mart's first store in Bentonville, Ark.) to fund expansion and ad campaigns. But the Internet craze--invariably described either as a "land grab" or a "gold rush"--turned the process on its head. First build a "brand," the thinking went. Then get eyeballs. Then turn them into paying customers. Then figure out how to make a business out of it.

If there is one element of the dot-com experiment that can be declared a categorical failure, it's this attempted creation of the instant company. Except in a handful of cases ( being the most notable), huge up-front ad expenditures failed to translate into big sales, leading to customer acquisition costs as much as four times higher than those of offline competitors, according to McKinsey & Co. "Lesson numero uno is that the cost of customer acquisition is prohibitively high," says McNamee. "It suggests the Internet is an extraordinary way to communicate with known customers but a terrible way to attract new customers."

It may also be an admonition not to put the marketing cart ahead of the horse. "The fundamental mistake the dot-coms made is they assumed that brand is a thing in itself, separate from the product, and that it created value: 'Just get me known, and all this other stuff will fall into place,' " says Regis McKenna, the dean of Silicon Valley marketers. (This quest for attention has taken on an increasingly desperate air: SocialNet's CEO is offering a date with herself; has put sloganed rubber pads in urinals; recent e-mails from declare "Sock Puppet mania!")

The instant-company approach rested in part on voguish concepts like first-mover advantage and the "network effect," whereby the first company to build a major brand would "lock in" customers. But the eyeballs that Websites managed to attract didn't turn out to be very loyal: One market study reports, for instance, that music e-tailer CDnow has 83% name recognition among online shoppers yet just a 17% loyalty rating. Brand familiarity does not necessarily breed loyalty, it concludes.

Why anyone thought the Internet--where ease of entry attracts an immediate flock of imitators--would promote customer loyalty is hard to account for. As McKenna notes, the billions of dollars poured into venture capital every year are essentially a bet that customers will be disloyal to existing offerings. "The first mover and the network effect clearly win over and over in software," says CEO Dan Nordstrom. "But people took those metaphors and tried to apply them to e-commerce," where they didn't necessarily apply. Says Amar Bhide, a Columbia University professor who studies startups: "A lot of what people think of as first movers are really first succeeders."

Many companies are finally capitulating to this reality. "One lesson I've learned is it's not about brand building," says Kenneth Kurtzman, CEO of, a luxury-goods site that ditched its TV ads in favor of more targeted marketing online. "It's about brand experience," or how customers feel when they encounter his product, which involves getting 1,000 little things right. Venture capitalist Steve Jurvetson now has a litmus test for which companies to invest in. "If they're just saying, 'Oh, it's my brand,' that is not the right answer," he says. "Brand is not a strategy."

ENTREPRENEURSHIP CANNOT BE SYSTEMATIZED. Or at least not as easily as the Internet incubators would have us believe. What began as a promising and sensible idea--that a cluster of startups could share services and a core of entrepreneurial expertise--rapidly devolved into a pandemic of me-tooism, with FORTUNE 500 companies and 23-year-olds alike scrambling to set up incubators; one estimate puts the total at 350. Though the jury is still out on whether the best of the lot will create real value, the unsustainable stock prices of many of their offspring have many calling these idea factories "incinerators." "The idea that you can institutionalize the creation of entrepreneurial ventures is absolute bunk," says venture capitalist Bill Unger. Agrees Randy Komisar, author of the bestselling Silicon Valley tale The Monk and the Riddle and a "virtual CEO" who has helped godfather such companies as WebTV and TiVo: "Incubators are the Hula-Hoops of the new-millennium economy."

INVESTORS ARE NOT YOUR CUSTOMERS. Drive down Highway 101 from San Francisco to Palo Alto with the radio on, and you're apt to hear two or three dozen commercials for dot-coms. "Most of that advertising is not value creation," says Regis McKenna. "It's aimed at investors, to keep the market value up." (A joke around Silicon Valley these days is that the chief of marketing should report to the chief financial officer.) That may not be true in every case, but it does reflect the dot-coms' persistent confusion about who their customer is: the paying consumer or the capital markets.

Inside the bubble, many dot-com entrepreneurs were less interested in creating companies than in creating securities that would make them very rich. Quips David Roux of Silver Lake: "The company was something you had to put up with to create the security." Sometimes investor and customer quite literally became one and the same--as when startups awarded large chunks of pre-IPO stock to their important corporate clients.

As long as the bubble continued to inflate, this approach paid off handsomely. Unprofitable eyeballs could be made profitable simply by selling them to someone else, creating, in effect, a chain of ever-greater fools, beginning with the entrepreneur and ending with the investing public. Executives joked privately about seeking "dumb money."

But then investors said no mas. For Sun Microsystems co-founder Bill Joy, that point came in March, when the stock market valued Palm--then almost wholly owned by 3Com Corp.--more highly than 3Com itself, a deeply kooky situation that suggested that the stock market had teetered into irrationality. "I knew things were crazy, I was saying they were crazy, but this really was crazy!" says Joy. "I didn't want to pick stocks anymore. I didn't know what was high or what was low anymore." So he simply sold off a percentage of his portfolio. The following month a lot of people followed suit.

Many welcomed the resulting April meltdown as a chance to refocus on an overlooked reality: Companies are built by slavishly winning over a real base of customers. But others aren't so sure the hysteria is over. "I'm not real sanguine about what we've learned," says Randy Komisar. "The alarm bells went off, but we pushed the snooze button." He notes that venture capitalists have essentially picked up the party and moved it elsewhere, leaving old dot-coms for dead while occupying a series of newer bubbles in such areas as wireless and optical networking.

It does seem hard to keep a good mania down. The young founder of a prominent Silicon Valley company stopped by FORTUNE's offices recently, explaining how his company had just "hit a home run." Asked what that might be, he said it had just closed a round of financing at a high valuation.

THE INTERNET STILL CHANGES EVERYTHING. It would be tempting to conclude that the Internet economy was simply a figment of our financial imaginations, a chimerical conjuring of Jeff Bezos and friends, and that now things will return to normal. But as Komisar argues, "the last couple years have fundamentally changed what 'normal' is." Says Roger McNamee: "The risk now is that people will draw negative conclusions every bit as outlandish and inappropriate as the ones they drew on the positive side. For instance, right now Wall Street knows with the conviction of Mother Teresa that you can't make money on B2C. And guess what? That is for sure wrong." He adds: "Anyone who thinks the Net is not transformational is dreaming."

Because here's the thing: While the Internet may not have disrupted the old industrial order, it has disrupted the old way of doing business--particularly the relationship between customers and corporations. Consider the power that companies must now put at customers' disposal, like it or not: customization tools, "choiceboards," decision-making data. So equipped, customers are exerting ever more control over the economy's productive capacity, turning--in the words of Mercer consultants Adrian Slywotzky and David Morrison--from "product takers" into "product makers."

Will stand-alone Internet companies have a part in this picture? Of course. One of the Net's demonstrable effects is the reduction of what economists call transaction costs, or the cost of interacting with other businesses. As these come down, companies find it more economical to farm out functions once performed internally. The result is a host of small, narrowly focused companies like MyVirtualModel, a Canadian firm that provides a customizable 3-D mannequin for shopping sites like Lands' End. Indeed, every time a consumer makes a purchase from the Lands' End Website, she passes through a number of such largely invisible intermediaries.

MyVirtualModel doesn't have ".com" in its name, but spelling isn't the only way it differs from the original crop of dot-coms. For one thing, it's trying to make money not by upending but by supporting the retail infrastructure. It need not build its brand with Super Bowl ads. And it has--gasp!--a paying customer in Lands' End. Or consider the company iChoose, which helps online shoppers do instant price comparisons and collects its revenues from retailers. As Nicholas Carr, an executive editor at Harvard Business Review, has noted, iChoose is essentially charging companies for the right to sell their wares at cut-rate prices--a model that manages to be enabling and disruptive all at once.

Meanwhile, the protean, evanescent quality that characterizes many dot-coms doesn't seem likely to go away. Some startups will exist solely to inject innovations into the corporate bloodstream, functioning as the R&D departments of their eventual acquirers. Even those that do stick around will give up on seeking the once-coveted "sustainable competitive advantage" in favor of what some call "renewable competitive advantage" or "leverageable advantage"--using one temporary position of strength to hopscotch into another. Silicon Valley, that is, will continue to turn out not just technological innovations but managerial innovations too.

THE INTERNET CHANGES YOUR JOB. The dot-com boom spawned an immoderate number of "death of" and "end of" scenarios: the death of the salesman, the end of the travel agent, and so on. So far these folks are still with us. But as many basic transactions migrate online, big changes are in the offing. It's already clear that stockbrokers can no longer make a living simply conducting trades; human resources professionals cannot make a living processing resumes; travel agents cannot make a living booking airline reservations. Talk to people in those professions, and you'll hear some variation on how they're moving into more "strategic" or "consultative" roles. (Another bonus lesson: The Internet turns everyone into a consultant.) Again, this shift is both enabling and disruptive. On the one hand, many workers relish the end of order-taking drudgery. But it's also not clear just how many travel/financial/HR "consultants" the world needs. Warns John Sviokla, vice chairman of e-consultancy Diamond Technology Partners: "Salespeople will have to step up and add even more value."

The dot-coms, meanwhile, have clearly ushered in a more entrepreneurial approach to careers. Early last year E*Trade CEO Christos Cotsakos recalled to this reporter how one twentysomething job applicant had asked him point-blank, "How do I know that you're smart enough not to screw this up?" "That stunned me," Cotsakos said. "Back in the early '70s, when I was a young manager at Federal Express, you'd be thrown out for asking that! But in this case I had to say, 'You know, you really don't know that.' " The candidate, Cotsakos stressed, was not unusual. "They all demand to meet with the CEO first and hear my vision.... It's not an interview. It's a conversation about business strategy, and they'll decide whether or not to work here based on what you say."

THE DISTINCTION BETWEEN INTERNET COMPANIES AND NON-INTERNET COMPANIES IS FADING FAST. About a year ago Intel Chairman Andy Grove declared that within several years, "there won't be any Internet companies. All companies will be Internet companies, or they will be dead." One wonders if we're already nearing that point. Up to now we've been captivated by the story line of the online attackers vs. the offline defenders. But as the Internet becomes as deeply ensconced in business as, say, the telephone, the protagonists blur. Which brings us to the final lesson....

THE REAL WEALTH CREATION IS YET TO COME. In the end, the biggest obstacle to building an Internet-based economy may be our eagerness to build it. "We are too impatient to write history," says Grove. In other words, since it hasn't happened in "Internet time," we conclude it won't happen at all.

The past can be instructive at such moments. Consider, for example, the personal computer industry in 1985. Back then, an initial mania had turned into a bust, Apple was laying off workers, and some commentators were writing off PCs as a fad. Yet the serious wealth creation was still to come--most of it from companies no one had yet heard of, like Gateway and Dell. Similarly, says David Roux, "the big winners on the Internet probably aren't even in business yet."

Most of today's dot-coms will prove mere prologue. "A lot of paper wealth was created around concepts that didn't even have viable business models," says Roger McNamee. "It didn't last very long, and now that most of that paper profit has gone off to paper-profit heaven, we're going to settle down to the business of building the Internet industry." This will involve more trial and error, and it can't be accomplished with press releases. "It's work," says Grove. "Very unglamorous work.... The heavy lifting is still ahead of us." Yet as even Phil Kaplan, creator of the notorious Website, concedes, "[The Internet] is going to change the world. Just not in the way some companies wanted it to."

The dot-com revolution is dead. Long live the Internet revolution.