The dot-com boom, what actually happened and why it crashed

Field note. The fiber and infrastructure your server's traffic uses exists because of a 1990s bubble that built way more bandwidth than anyone needed at the time. We're still drawing down that surplus.

Between roughly 1995 and 2000, the U.S. stock market went on a historic run, fueled by a class of new internet companies. Most of those companies either died or were transformed by 2002. The boom and crash together are often called the dot-com era.

The story is more interesting than the simplified versions. It wasn't only mania, and the crash didn't kill the internet. This article walks through what actually happened.

The setup

In 1993, the internet was a research network. By 1995, the Web was mainstream-accessible. Netscape's IPO in August 1995 signaled to Wall Street that "internet companies" were investable.

Several factors converged:

Cheap money. U.S. interest rates were low. Investors had capital to deploy and were hunting for yield.

Tech euphoria. Personal computers had created PC giants (Microsoft, Intel). The same pattern was expected for the internet.

Genuine technological transformation. The internet really was changing things. Not all enthusiasm was misplaced.

Public market access. IPOs let early investors and employees realize gains quickly. The cycle of "raise money, build, IPO" became standardized.

By 1997, an "internet IPO" was a category. By 1998-1999, it was a frenzy. By 2000, valuations were untethered from any business reality.

The companies

The boom produced several categories of companies.

The legitimate survivors. Amazon (1994), eBay (1995), Google (1998), Yahoo (1995). These had real business models and (eventually) real revenue. They survived and became gigantic.

The legitimate also-rans. Lycos, Excite, InfoSpace, theglobe.com. Real businesses that lost the competition for dominance.

The famous failures. Pets.com, Webvan, eToys, Boo.com, Kozmo. Companies that raised hundreds of millions and burned it without finding a business model.

The shells. Companies that existed mainly to be IPO'd. No revenue, no clear product, no plan beyond cashing out.

The mix varied. At the peak (1999-2000), valuations even for category-defining companies were absurd. Amazon's stock peaked at $113 in December 1999, fell to under $6 by September 2001.

The metrics that mattered (then)

Traditional companies were valued on profits, or at least on a clear path to profits. Dot-coms were valued on:

Eyeballs. How many people visited the site.

Engagement. How long they stayed.

Burn rate. How much money the company was spending per month. Faster spending was a sign of "growth."

Pageviews per quarter. Even if the page views weren't monetized.

The phrase "get big fast" justified massive losses. The theory: the first company to dominate a category would have such network effects that it could later extract profits. The math sometimes worked (Amazon eventually monetized). More often it didn't.

By 2000, the disconnect was visible. Pets.com had famous Super Bowl ads, a sock-puppet mascot, growing brand recognition, and was hemorrhaging money. The market shrugged this off until it couldn't.

The peak

Spring 2000. The NASDAQ Composite Index (heavily tech-weighted) peaked at 5,048 on March 10, 2000.

For context:

• It had been at 1,000 in mid-1995.
• 2,000 in mid-1998.
• 4,000 in late 1999.
• 5,000 in March 2000.

A 5x increase in five years. Some individual stocks were up 10x or 100x.

The Federal Reserve had begun raising interest rates in mid-1999, trying to cool the speculative frenzy. By March 2000, the rates were taking effect.

The crash

From the March 10, 2000 peak, NASDAQ declined steadily:

• April 2000: 3,400 (already down 30%).
• October 2000: 3,300.
• April 2001: 1,700 (down 66% from peak).
• October 2002: 1,114 (down 78% from peak).

Two and a half years of decline. Many dot-coms went bankrupt. Many more were acquired at fire-sale prices. Tech employment dropped sharply in San Francisco, Silicon Valley, New York.

The S&P 500 fell about 50 percent over the same period (less than NASDAQ but still significant).

Notable failures by 2002:

• Pets.com. Liquidated November 2000, nine months after IPO.
• Webvan. Bankrupt July 2001.
• eToys. Bankrupt March 2001.
• Kozmo.com. Bankrupt April 2001.
• MarchFirst. Bankrupt April 2001.
• Excite@Home. Bankrupt October 2001.
• boo.com. Bankrupt May 2000.

Hundreds of others, less famous, also failed.

The 9/11 effect

A complicating event: September 11, 2001, in the middle of the decline. The terrorist attacks accelerated the recession and further depressed markets. Without 9/11, the dot-com recovery might have started a year or two earlier. As it was, the bottom wasn't reached until October 2002.

Why it crashed

A confluence of factors:

Rate hikes. The Fed raised rates from 4.75 percent to 6.5 percent through 1999-2000. Made speculative investments less attractive.

Y2K passing. The expected Y2K-related infrastructure spending didn't materialize. Tech companies that had been growing on Y2K sales saw revenue cliff-drop.

Lock-up expirations. Many IPO'd companies had 6 or 12-month lock-ups before insiders could sell. As these expired, insiders sold heavily, depressing prices.

Realizations of fundamental problems. Many dot-coms had no path to profitability. As cash ran out (running through "burn rate" with no replacement funding), they failed.

Self-reinforcing decline. As tech stocks fell, dot-coms couldn't raise new money. Without money, they failed. Failures depressed investor confidence further. Repeat.

Specific company collapses. When notable failures (Pets.com, Webvan) hit the news, the broader narrative shifted. "Dot-com" became negative.

The survivors

Companies that survived 2000-2002 included:

• Amazon. Stock down 95 percent at the bottom. Survived because Bezos was relentless about operational efficiency and they had genuine revenue. Now the world's largest e-commerce company.
• eBay. Profitable throughout. Less affected by the crash.
• Yahoo. Survived but eventually displaced by Google. Now part of Verizon.
• Google. IPO'd in 2004, post-crash. Built on lessons learned from dot-com excesses.
• PayPal. Spun off and acquired by eBay; later spun off again.
• Priceline. Survived; renamed Booking Holdings; now huge.

The pattern: companies with real business models, prudent management, and enough cash to survive the trough became gigantic. Companies without those died.

Infrastructure overbuild

A specific dot-com phenomenon worth noting: massive overbuild of network infrastructure.

Companies like Global Crossing, WorldCom, and Qwest invested billions in fiber-optic cables, anticipating exponentially-growing internet traffic. They overbuilt. The fiber was sometimes called "dark fiber" because nobody was lighting it.

When demand didn't materialize, these companies failed or restructured. But the fiber remained. In the 2000s, it was bought cheaply by surviving telcos and used to support the next decade's growth.

This is why broadband became reasonably affordable in the 2000s and 2010s: the dot-com era had paid for the cables.

Lessons

What the dot-com era taught:

Network effects are real but slow. First-mover advantage matters but isn't immediate dominance. Building the right product matters more than being first.

Burn rate isn't growth. Spending money is easy. Building a business is harder.

Eyeballs aren't customers. Traffic without revenue is a vanity metric.

Funding can stop. Companies need to be sustainable in their own right, not dependent on continuous new investment.

Markets are momentum-driven. Stock prices can decouple from fundamentals for years. They re-couple eventually, often painfully.

Some bubbles fund useful infrastructure. The dot-com era's overbuilt fiber, software-engineering talent, and consumer adoption of the internet enabled the next 20 years.

These lessons were learned by 2001-2002. They've been forgotten and re-learned several times since (2008's housing bubble, 2017's crypto, various tech booms).

The aftermath

Post-crash, the surviving tech sector evolved:

Web 2.0 (2003-2008). A new wave of more pragmatic internet companies. Google, Facebook, YouTube, Twitter. Better business models, more cautious valuations.

The mobile revolution. Starting 2007 with the iPhone. Reshaped everything.

Cloud computing. Amazon launched AWS in 2006. Enabled cheaper startup costs.

Modern tech giants. By 2015, the five largest companies by market cap were all internet companies (Apple, Microsoft, Alphabet/Google, Amazon, Facebook). The trend continues.

The dot-com era was partly a casino and partly the early innings of the trend that dominates the modern economy. The crash didn't reverse the trend; it just popped specific excesses.

A reflection

In hindsight, the dot-com era looks like a sane bet wrapped in a manic frenzy. The bet: the internet would transform commerce, communication, and society. Largely correct. The frenzy: any company with ".com" in its name was worth millions. Largely wrong.

Investors who held the most-resilient tech companies through the crash (Amazon, Google after its IPO) ended up vastly enriched over the next 20 years. Investors who bought Pets.com mostly didn't.

Distinguishing in real time was hard. With hindsight, the question "did this company have a plausible path to profitability?" was the critical filter.

Conclusion

The dot-com boom and bust was a real economic event with real consequences: billions of dollars lost, tens of thousands of jobs eliminated, several major companies destroyed.

It was also, less obviously, the founding period of the modern commercial internet. The infrastructure built, the talent trained, the consumer habits formed, the companies that survived: all of these defined the next two decades.

The story isn't "the internet was a fraud." The story is "the internet's transformative potential was real, and the market took a long, painful detour through speculation before settling on the reality."

If you're investing in any "transformative new technology" in 2026 (AI, crypto, biotech), the dot-com story is the most relevant historical analog. Both the optimism and the eventual painful re-pricing apply.

Coming up

Next: the mobile internet, the second great inflection. The 2007-2010 period when the iPhone changed everything.