The Dot-com Bubble: Irrational Exuberance and the First Internet Crash

Zusammenfassung

This article traces the dot-com bubble — from the Netscape IPO of 1995 that announced the commercial internet’s arrival, through the years of investor euphoria in which companies with no revenue were valued at billions of dollars, to the Nasdaq crash of 2000-2001 that destroyed $5 trillion in market value. It is a story about the collision between genuine technological transformation and the financial mania that attached itself to it — and about why the crash, devastating as it was, did not stop the internet.

The Starting Gun: The Netscape IPO

On August 9, 1995, Netscape Communications went public. The company was sixteen months old. It had no profits. Its lead product — the Navigator web browser — was given away free to most users. By any conventional valuation metric, it was worth a modest amount.

The stock was offered at $28 — doubled at the last minute from a planned $14 — opened its first trade at $71, and closed at $58.25. Netscape was worth $2.9 billion by end of day. Jim Clark, its co-founder, became worth $660 million. Marc Andreessen, twenty-four years old, was worth $58 million.

The message to the investment community was unambiguous: the old rules did not apply to the internet. Companies did not need profits. They did not need revenue. They needed a URL and a sufficiently compelling story about the future, and investors would make their founders wealthy.

Alan Greenspan, Chairman of the Federal Reserve, gave the moment its name in a speech on December 5, 1996: “How do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions?”

Nobody stopped.

The Rules of the New Economy

The intellectual scaffolding for the bubble was provided by a set of ideas that were individually defensible and collectively dangerous.

Metcalfe’s Law — named for Ethernet inventor Robert Metcalfe — held that the value of a network was proportional to the square of the number of its users. Applied to internet businesses, this meant that user growth was the paramount metric: every new user made the network more valuable, justifying investment in growth at the expense of profitability. “Get big fast” was the strategic imperative.

The winner-take-all logic of network effects argued that internet markets would be dominated by single companies — the first to achieve critical mass would lock competitors out. Therefore, establishing market position quickly was worth any short-term cost.

The “eyeball economy” held that companies that attracted large audiences — measured in page views or registered users — could monetize those audiences later. Advertising revenue would follow audience size. Profitability was a second-order concern.

Each of these ideas was true in some contexts. Metcalfe’s Law does describe network effects. Some internet markets are winner-take-all. Audiences can be monetized. The error was applying them universally — assuming that any internet business with growth metrics was therefore valuable, regardless of whether the growth could ever be converted to profit.

The Companies That Embodied the Mania

The venture capital and IPO markets of 1998–2000 funded experiments at a scale and velocity that produced a remarkable set of failures.

Pets.com, founded in 1998, raised $82.5 million in a February 2000 IPO, and liquidated in November 2000 — nine months after going public. It had spent roughly $1.2 million on a single Super Bowl advertisement — part of more than $35 million in advertising in one year — for a business that shipped 40-pound bags of dog food at a loss. The cost of shipping exceeded the revenue on every order.

Webvan raised $375 million in a 1999 IPO for a grocery delivery business that required building automated warehouses in multiple cities simultaneously. The warehouses cost more than projected; customer acquisition was slower than projected; the company burned through its capital and filed for bankruptcy in 2001. The grocery delivery model it attempted has since been revived — but by companies that grew more slowly and proved the unit economics first.

Kozmo.com promised one-hour delivery of books, DVDs, food, and other items from urban warehouses. It operated in multiple cities, delivered items for free, and burned approximately $280 million in capital before closing in 2001.

The common pattern was not fraud but miscalculation: the assumption that customer acquisition at any cost was justified by future revenue that never materialized, combined with the pressure to spend capital quickly to demonstrate growth before competitors did.

The Bubble Was Not the Internet

The dot-com crash destroyed $5 trillion in market value and ended thousands of companies. It did not destroy the internet. While dot-com stocks fell 78% from peak to trough, internet usage continued to grow throughout the crash. The number of internet users doubled between 1999 and 2002. Amazon, Google, and eBay survived and grew. The infrastructure built during the boom — fiber optic cables, data centers, software tools — remained in place and became more valuable as the surviving companies scaled.

The crash was financial, not technological. What collapsed was the belief that any internet company was worth billions; what survived was the internet itself.

What Survived and Why

Three companies defined the difference between the companies that survived the crash and those that did not.

Amazon nearly died. Its stock fell from $107 in 1999 to $7 in 2001. It survived because Jeff Bezos had built actual infrastructure — warehouses, logistics, software systems — rather than spending on marketing. The business was real, the losses were investments, and Bezos had convinced investors to accept losses in exchange for scale. Amazon turned its first full-year profit in 2003.

Google was not yet public during the crash. Founded in 1998, it grew quietly, did not IPO until 2004, and was profitable before it went public. Its business model — charging per click on relevant search advertisements — was directly tied to user intent, not to speculative eyeball counts. It had discovered a form of internet advertising that worked.

eBay survived because it had a genuine network effect: each additional seller made the platform more valuable to buyers, and vice versa. Unlike Pets.com, eBay did not handle physical goods; it was a marketplace that extracted value from transactions it did not have to fulfill.

Dead End: The Business Model Without a Business

The dot-com era’s fundamental error was mistaking the excitement of a new technology for the automatic creation of new economic value.

The Revenue Model That Wasn’t

Many dot-com companies had no revenue model at all — they planned to “figure out monetization later” once they had achieved sufficient user scale. This was not universally wrong: Twitter operated for years without significant revenue while establishing its network. But Twitter had network effects that made users genuinely more likely to stay. Companies that had neither revenue nor genuine lock-in had no reason to survive once their capital ran out.

The crash taught investors, temporarily, to ask a question they had stopped asking during the boom: How does this business make money? The lesson faded; it was asked again during the 2021 SPAC and crypto mania, with similar results. Each generation of technology investors rediscovers, at some cost, that a large addressable market and a compelling story are not substitutes for a functioning business model.

For the technologies the bubble helped fund, see The Connected World and The Browser Wars. For the search company that emerged from the wreckage, see The Search Engine Wars.