The dot-com bubble is often reduced to “people invested in stupid websites, lost money, learned nothing.” That’s not wrong, but it misses the complexity of what actually happened.

Between roughly 1995 and March 2000, venture capital and public markets poured hundreds of billions of dollars into internet companies. Most of those companies failed spectacularly. Investors lost fortunes. The NASDAQ crashed 78% from peak to trough.

But the infrastructure built during the bubble - fiber optic networks, data centers, e-commerce platforms, web services - became the foundation for the modern internet economy. So was it a bubble or an investment in the future? Both, depending on where you look.

The Build-Up (1995-1999)

The bubble’s origins trace to the Netscape IPO in August 1995. Netscape was a 16-month-old company with $16.6 million in revenue. Its stock opened at $28 and hit $75 on the first day, valuing the company at $2.9 billion.

This was unprecedented. Traditional IPOs involved mature companies with proven business models. Netscape proved that internet companies could go public early and achieve massive valuations based on potential rather than profitability.

Investors developed a new logic: profit margins don’t matter, revenue growth doesn’t matter, all that matters is growth in users/eyeballs/traffic. Once you had scale, monetisation would follow.

This logic wasn’t entirely wrong. Amazon, founded 1994, lost money for years but was clearly building something valuable. The problem was applying this logic to every company with a .com domain.

By 1999, the market had gone completely irrational. Companies with minimal revenue and massive losses were achieving billion-dollar valuations. IPOs would double or triple on the first day of trading. Day traders were quitting their jobs to trade tech stocks full-time.

The NASDAQ composite index, heavy with tech stocks, went from 1,000 in 1995 to 5,048 in March 2000. That’s a 400% increase in five years.

The Insane Valuations

Some examples of peak dot-com valuations:

Pets.com: Online pet supply retailer. Spent $27 million on advertising (including a Super Bowl ad) in its first year. Revenue: $619,000. Market cap at IPO: $290 million. Liquidated in November 2000, nine months after going public.

Webvan: Online grocery delivery. Spent $1 billion building infrastructure before having proven the business model worked. Market cap peaked at $1.2 billion. Filed for bankruptcy in 2001.

eToys: Online toy retailer. Market cap hit $8 billion at its peak, higher than Toys “R” Us, a company with 1,500 physical stores and actual profits. eToys went bankrupt in 2001.

Kozmo.com: Promised to deliver anything in under an hour in major cities. Free delivery, no minimum order. Lost money on every transaction. Raised $280 million in funding. Shut down in 2001.

Boo.com: Fashion e-commerce site. Spent $135 million in 18 months. The website barely worked (too much JavaScript for most users’ connections). Collapsed in May 2000.

The pattern was consistent: raise huge amounts of capital, spend it on growth and marketing, worry about profitability “later,” run out of money before figuring out a sustainable business model.

The Metrics That Didn’t Matter

Traditional business metrics - revenue, profit, cash flow - were dismissed as “old economy thinking.”

New metrics emerged:

Eyeballs: How many people visited your site. Didn’t matter if they bought anything.

Page views: How many pages they looked at. Also didn’t matter if they bought anything.

Burn rate: How fast you spent investor money. High burn rates were sometimes seen as positive - it meant you were “aggressive” and “scaling quickly.”

First-mover advantage: The assumption that being first in a market segment guaranteed eventual dominance. Therefore, spending whatever it took to achieve market share made sense.

Some of these metrics weren’t wrong in principle. User growth does matter for platforms. First-mover advantage can be real. But they became disconnected from any path to profitability.

Why Smart People Believed

It’s easy to mock dot-com investors with hindsight. But the belief wasn’t entirely irrational.

The internet was genuinely revolutionary. E-commerce, online advertising, digital media - these were all valid business models that would eventually work.

Some companies from that era did succeed. Amazon, eBay, and Priceline survived and became massive. Google launched in 1998 and became one of the most valuable companies in history.

The problem was distinguishing which internet businesses would succeed and which were built on hype. In a rapidly changing market, traditional valuation methods seemed inadequate. Maybe “old economy thinking” really didn’t apply anymore.

Plus, everyone else was making money. Your neighbor was getting rich on tech stocks. Analysts were predicting the NASDAQ would hit 10,000. Venture capitalists were funding anything with a .com domain. Fear of missing out drove both investors and entrepreneurs.

The Crash (2000-2002)

The NASDAQ peaked at 5,048 on March 10, 2000. Within months, it was clear something was wrong.

Companies that had gone public were running out of cash. IPO markets dried up - companies that had planned to raise money through public offerings couldn’t. Private funding became scarce as VCs realized their portfolio companies weren’t going to succeed.

The crash happened in stages:

March-April 2000: NASDAQ drops 25% in a matter of weeks. Some companies start failing, but it’s not yet clear this is a sustained crash.

2000-2001: Steady decline. Companies run out of money and shut down or sell for pennies. Layoffs accelerate. The term “dot-com crash” enters common usage.

2001-2002: September 11 attacks compound the decline. The NASDAQ bottoms at 1,114 in October 2002 - a 78% drop from peak.

Trillions of dollars in market value vanished. Hundreds of companies shut down. Tens of thousands of workers lost jobs. The Bay Area office market collapsed. Webvan warehouses sat empty.

What Actually Failed

Most dot-com failures fell into categories:

Business model didn’t work: Kozmo delivering anything in under an hour for free. Webvan building massive infrastructure before proving demand. These weren’t viable businesses at any scale.

Too early: Some ideas were sound but the technology wasn’t ready. Streaming video companies failed in 1999 because bandwidth was too limited. YouTube succeeded in 2005 when broadband was common.

Too much competition: Having a good idea didn’t matter if 50 other companies had the same idea. Online pet supply retailers competed with each other, with Amazon, and with physical pet stores. None had sustainable advantages.

Poor execution: Boo.com’s website didn’t work. Numerous e-commerce sites had terrible user experiences, slow loading, and buggy checkout.

Fraud: Some companies were outright scams, exaggerating metrics or lying to investors. Many more weren’t fraudulent but were deeply dishonest about their prospects.

What Survived

Not everything failed. Companies that survived the crash:

Amazon: Lost money until 2003 but had a clear path to profitability. Built genuine competitive advantages in logistics and scale. Stock dropped 90% but recovered.

eBay: Profitable, dominant in online auctions, clear network effects. Survived the crash and thrived.

Google: Founded 1998, went public 2004. Avoided the bubble’s worst excesses by staying private and focusing on building a sustainable ad business.

PayPal: Payment processing that actually solved a problem. Acquired by eBay in 2002, later spun back out.

Salesforce: Enterprise software sold as a service. Launched 1999, profitable by 2001. One of the few enterprise-focused companies from that era to succeed.

The pattern: companies with actual revenue, clear paths to profitability, or genuine competitive advantages survived. Pure hype companies died.

The Infrastructure Legacy

Here’s the paradox: most dot-com companies failed, but the boom left valuable infrastructure.

Telecoms companies laid thousands of miles of fiber optic cable, anticipating massive internet traffic growth. They overbuilt spectacularly and many went bankrupt. But that dark fiber became the backbone of the modern internet.

Data centers were built to host dot-com companies. When those companies died, the data centers found new tenants.

E-commerce platforms, payment systems, hosting services, web development tools - much of the infrastructure built during the bubble became essential as the internet matured.

The boom also trained a generation of entrepreneurs, engineers, and investors. Many of them learned from their failures and built more successful companies in the 2000s.

So was the bubble wasteful? Yes - billions were spent on companies that never should have been funded. But the infrastructure and knowledge that survived had enormous value.

Lessons Learned (and Forgotten)

The dot-com crash supposedly taught investors to be more skeptical of unprofitable growth companies. That lasted about a decade.

By the 2010s, similar patterns emerged. Uber, WeWork, and dozens of other startups raised billions at massive valuations despite losing money on every transaction. The metrics shifted - it was about “total addressable market” and “unit economics” instead of “eyeballs” - but the fundamental logic was similar.

WeWork’s collapse in 2019 echoed dot-com failures. Massive valuation based on hype, charismatic founder, business model that didn’t make sense, eventual crash. The main difference was it stayed private longer.

The lesson seems to be that bubbles happen when money is cheap, new technologies emerge, and everyone convinces themselves “this time is different.”

Sometimes it is different - Amazon, Google, and Facebook did build massively valuable businesses from the internet. But most of the time, basic business fundamentals still matter.

What It Was Like

If you weren’t around for it, the atmosphere is hard to convey. The dotcom era felt like the future was arriving ahead of schedule.

Newspaper ads for web developers offered signing bonuses and stock options. 25-year-old product managers at failing startups acted like they were changing the world. Launch parties had open bars and expensive catering funded by venture capital.

The collapse was brutal. Layoffs happened en masse. Office space that had been impossible to find became vacant overnight. People who’d turned down traditional careers for dot-com equity found themselves unemployed with worthless stock options.

San Francisco and the Bay Area were hit hardest, but the effects rippled everywhere. The crash contributed to the 2001 recession. It damaged trust in the stock market. It made an entire generation of investors skeptical of tech companies, which ironically made them miss the next wave of genuinely successful internet companies.

Final Thoughts

The dot-com bubble was both irrational exuberance and a necessary phase of building internet infrastructure. Most of the money was wasted on companies that never should have existed. But some was invested in building the platforms and infrastructure that power the modern internet.

The companies that survived - Amazon, eBay, Google, PayPal - are now among the most valuable in the world. Their success validated the core insight of the bubble: the internet would transform commerce, media, and communication.

The crash taught hard lessons about business fundamentals, sustainable growth, and the danger of confusing a good idea with a viable business. But those lessons seem to need re-learning every decade or so.

The dot-com era left us with a faster internet, working e-commerce infrastructure, and proof that internet-native businesses could scale globally. It just took a few trillion dollars in losses to get there.