After the Bubble

  |  Command Line Heroes Team   科技史

Command Line Heroes • • After the Bubble | Command Line Heroes

After the Bubble | Command Line Heroes

About the episode

The Y2K bug generated a lot of fear, but all that hype fizzled when the new millennium didn’t start with a digital apocalypse. It turns out that fear was just aimed at the wrong catastrophe. While plenty were riding high on the rise of the internet beyond the Y2K scare, another disaster had been brewing since 1995—and would bring them back down. But the dot-com bubble wasn’t the end. The internet was here to stay.

Not long after the turn of the millennium, the dot-com economy collapsed. Peter Relan points to the flawed business plans that fueled the dot-com bubble, and how many entrepreneurs and investors underestimated the complexity of building a business on the internet. Ernie Smith tells the story of Pets.com, and how a similar idea a decade later had a much better chance of succeeding. Gennaro Coufano reveals the element of luck that saved Amazon from going under—and how it evolved in the aftermath. Julia Furlan reflects on the changes the dot-com bubble brought, and what’s left to consider. And Brian McCullough describes how the dot-com bubble paved the way for a more resilient digital economy.

Command Line Heroes Team Red Hat original show

订阅

Subscribe here:

Listen on Apple Podcasts Listen on Spotify Subscribe via RSS Feed

脚本

It's the final seconds of 1999. Alongside the partying, there's anxiety, catastrophe-scale anxiety, and that's because nobody knows whether the world's computers are going to survive the end of the millennium. Will a Y2K bug plunge our digital lives back to the Stone Age? Hmm, I guess not. Our tech didn't crash when clocks ticked over to the year 2000. But here's the thing: As we soared happily into the future, there was a tech catastrophe waiting around the corner. It just wasn't the one we expected. I'm Saron Yitbarek and this is Command Line Heroes, an original podcast from Red Hat. All season we've been telling the story of one of the most extraordinary years in tech history—1995. It was the year that fueled e-commerce, and web design, and search engines. 1995 was the year our online lives truly took shape and the dot-com bubble began dominating the stock market. For our season finale, we're looking at the end of that wild period of invention and growth. But 1995 didn't end after 12 months. From the tech world's perspective, the true end to 1995 arrived on March 10 of the year 2000. That was the moment the NASDAQ reached its all-time high, fueled by a fever of excitement about the world wide web. And then, well, here's the thing about bubbles. They burst and when the dot-com bubble burst, three-quarters of NASDAQ's value soon evaporated. The value of all those dot-com startups was thrown into question, and a new frontier of web-based technology had to be radically reimagined. Had it all just been a mirage, or was there something real behind the dream of 1995? This is the story of that fateful moment when one future came to a crashing end and a new future arrived on the horizon. By '97, the business plans that were being created were starting to be, shall we say, increasingly less "deep". Peter Relan is a technologist who runs an incubator in Silicon Valley. And he was there for the late 90s rollercoaster and saw startups hitch themselves to the dot-com craze. You really did see a lot of people who were non-engineers coming to Silicon Valley with a couple of pages of an idea and presenting it to the venture capitalists and getting funding. Once your new venture was funded, you'd go hire a CTO, or engineering manager, and get them to bring your idea to life. Business plans could be minimal, scrawled on a napkin, and that had some pretty huge consequences. It meant companies were being formed without reserves of cash, without any worries about overhead or profits, without any real plan for how to succeed in anything other than a bubble. Relan worked in one of the late 90s startups that actually had more than a half-baked idea behind it. In 1998, he joined an online grocery delivery service called Webvan, as their senior VP of Technology. Webvan's plan was to cut out the grocery store and bring your food to your front door at the click of a few buttons. Today, that sounds like a decent proposition. But back in 1998, it strained the limits of what could be done. Like most 90s startups, there were also a few holes in their business plan. The business plan included a service that was really good, high-quality products and brands, delivery at no extra charge, no questions asked upon returns, self-selected delivery slots down to a 30-minute window of your choice. You can't just do all of that and not have a premium pricing model. And the pricing model was supposed to be Safeway pricing. I think that was a combination that just was not going to pan out. But the other problem was timing. A shiny new startup in the late 90s had just a few years of grace to build up the kind of cash reserves that would help them survive the bubble bursting, and think how hard it would be to save much cash with the kind of overhead early web companies had to carry. Webvan's warehouse in the Bay area was the size of 18 supermarkets. So that's not cheap. But even more important, it was a digital company operating at a time when everything had to be built from scratch and maintained by you and you alone. So you built everything from the ground up, including automation of the warehouse, and the delivery hub and spoke model for sending the trucks and vans out. You built your own e-commerce engine, you built your own payments engine, you built your own logistics engine, you built your own customer support system. So we basically had to create the entire end-to-end experience without any third party infrastructure that you could leverage. That's the crucial difference between the startups of the late 90s and the startups of today. Back then, infrastructure was something you had to build yourself. And it was expensive. And time consuming. Today, all that infrastructure is waiting for you in the cloud, ready to rent out by the month. Today, if you wanted to create Webvan, you could probably get it done with 20 people in 6 months, using existing services from Amazon, and Google, and Stripe, and all the other infrastructure players that exist today. Which brings us to the most famous casualty of the dot-com era. If Webvan sounds like an idea that was ahead of its time, wait until you hear about Pets.com. The startup that would forever symbolize dot-com excess and poor planning. Ernie Smith is the editor of Tedium, a newsletter about the dull side of the internet, and he's done a deep dive into the Pets.com story. Pets.com was basically trying to be the Amazon of pet supplies. The idea was that people would go online, they would order dog food, cat litter, whatever they needed for their pets, and it would be delivered to their door. Again, this sounds like a perfectly reasonable business model today. But in 1999, the infrastructure and consumer behavior just weren't there yet. The problem was that pet supplies are heavy and bulky and cheap. So you're dealing with items that cost very little but are expensive to ship. And in 1999, people weren't used to ordering everything online. They were still going to pet stores, they were still going to grocery stores to buy pet food. Plus, Pets.com made some spectacularly bad business decisions. They spent $1.2 million on a single Super Bowl ad, featuring their mascot, the sock puppet. The sock puppet became famous, but famous doesn't necessarily translate to sales. And when you're burning through cash at the rate that Pets.com was, you need sales, not just brand recognition. Pets.com went from IPO to complete liquidation in just 268 days. But here's the thing: the idea wasn't wrong, it was just early. If you look at companies like Chewy, which is basically doing exactly what Pets.com tried to do, they're incredibly successful. The difference is timing and execution. Chewy launched in 2011, when people were much more comfortable with e-commerce, when shipping infrastructure was much better, when the unit economics actually made sense. Which raises an interesting question: Were the dot-com failures really failures, or were they just premature successes? Let me tell you about one company that managed to surf the dot-com wave without wiping out: Amazon. Today, it's hard to imagine Amazon ever being in danger of going under. But during the dot-com crash, that's exactly what almost happened. Gennaro Coufano is the founder of FourWeekMBA, and he's studied how Amazon survived when so many other dot-coms didn't. Amazon was in real trouble during the dot-com crash. Their stock price fell from over $100 to under $6. They were burning through cash, they were being written off by analysts, and there were serious questions about whether they would survive. So what saved Amazon? It was partly good management, partly good timing, but also partly just luck. The key was that Amazon had managed to raise a lot of money right before the crash. In 1999, they raised $1.25 billion in debt and convertible bonds. That gave them enough cash to survive the downturn and keep growing while their competitors were going out of business. But Amazon also made some smart strategic decisions during the crash. Instead of just trying to survive, they used the downturn as an opportunity to expand and improve their operations. While other companies were cutting costs and laying off employees, Amazon was investing in new warehouses, new technology, new product categories. They basically used the crash as an opportunity to build the infrastructure that would make them dominant in the post-crash world. Amazon also benefited from the fact that they had a real business model. Unlike many dot-coms that were losing money on every sale, Amazon was actually making money on each transaction, even if they weren't profitable overall. Amazon understood unit economics. They knew that as they scaled, their costs would go down and their profits would go up. Many dot-coms were hoping that scale would somehow magically make their business model work, but they didn't have the fundamentals in place. The dot-com crash was a brutal reality check for the tech industry. But it was also a necessary correction that separated the companies with real business models from those that were just riding the hype. Now, I want to consider what the dot-com era means for us today. What lessons should we take from that period of boom and bust? Julia Furlan is an independent podcasting consultant and professor at The New School who thinks about technology and society. I think the dot-com era is a really important reminder that technology isn't neutral. The way we develop and deploy technology has real consequences for real people. Furlan argues that the "move fast and break things" mentality that characterized the dot-com era had some problematic consequences that we're still dealing with today. When companies prioritize growth over everything else, when they prioritize disruption over stability, that can have real negative effects on workers, on communities, on society as a whole. Take the gig economy, for example. Companies like Uber and Airbnb have used technology to "disrupt" traditional industries like transportation and hospitality. But that disruption has also created new forms of economic insecurity for workers. The rhetoric of disruption often masks the fact that what we're really talking about is transferring costs and risks from companies to individuals. When you turn taxi drivers into independent contractors, you're not just changing a business model—you're changing people's lives. This is where Furlan thinks we need to learn from the dot-com era. The focus on growth at all costs, on moving fast and breaking things, created a lot of value for investors and entrepreneurs. But it also created a lot of problems for everyone else. I think we need to be more thoughtful about the technology we're building and the impact it has on society. We need to think about sustainability, about equity, about the long-term consequences of our choices. But Furlan also acknowledges that the dot-com era created a lot of genuine innovation and progress. The question is how we can keep the good parts while avoiding the bad parts. Technology has incredible potential to make people's lives better. We've seen that with everything from search engines to social media to e-commerce. But we need to be more intentional about how we develop and deploy that technology. This brings us to the question of what we should make of the dot-com legacy today. Was it a cautionary tale about the dangers of hype and speculation? Or was it a necessary phase in the development of our digital economy? I think it was both. The dot-com bubble was definitely a speculative frenzy that got out of hand. But it was also a period of incredible innovation and infrastructure building that laid the foundation for everything we have today. Brian McCullough is the host of the podcast "Techmeme Ride Home" and has written extensively about the history of the internet. He sees the dot-com era as a necessary growing phase for the tech industry. Every new technology goes through a hype cycle. You have early adoption, then you have a period of inflated expectations, then you have a crash, and then you have gradual, sustainable growth. The internet is no different. McCullough compares the dot-com bubble to other historical bubbles, like the railway bubble of the 19th century. In each case, the bubble eventually burst, but the infrastructure that was built during the bubble remained and became the foundation for future growth. When the railway bubble burst, a lot of investors lost money. But the railways themselves didn't disappear. They became the backbone of the industrial economy. The same thing happened with the internet. All those fiber optic cables that were laid during the dot-com boom didn't vanish when the bubble burst. All those data centers that were built didn't disappear. They became the infrastructure that companies like Google and Facebook would later use to build their empires. The companies that succeeded after the crash—Google, Facebook, Amazon, Apple—they all benefited from the infrastructure that was built during the bubble. They got to focus on their products and services because someone else had already built the pipes. This is why McCullough argues that the dot-com pioneers, even the ones who failed, deserve credit for what they accomplished. Companies like Pets.com and Webvan were basically doing the heavy lifting for the companies that came after them. They figured out what worked and what didn't work. They built the infrastructure and proved the concepts. The companies that came later just had to execute better. But there's another side to this story. While the dot-com crash cleared out a lot of bad ideas and unsustainable business models, it also had real human costs. Thousands of people lost their jobs, their savings, their dreams. I think when we talk about disruption, oftentimes it's considered this great positive thing, but I think that I would caution everyone to really consider what it means to disrupt and to take that very seriously. After our Netscape IPO moment, after 1995, it was time to ask, not just how the web could become a money machine, but how it could do some good. I think people should think about their companies and their technologies in the larger ecosystem of the world, because I think that hasn't always been the case in Silicon Valley, and that hasn't always been the case in business, and we can do better. Furlan is calling for something we discussed in our last season finale. Check out the end of season 6 for an episode about venture capital and how it can support sustainable companies instead of just money-making unicorns. The path from 1995 to 2000, from dangerous heyday to a more stable industry, isn't really an anomaly. In the middle of the 19th century when a brand new thing called the railway arrived, people raced to invest in that tech, too. There was a railway bubble and just like the dot-com bubble, it ended up bursting. People lost a lot of money. But Brian McCullough, host of the podcast "Techmeme Ride Home", reminded us that burst bubbles aren't just cautionary tales. They're also the start of something new. When the dot-com bubble burst, there was all of this investment in infrastructure, laying fiber and laying the groundwork, for what we would now call cloud computing and having the infrastructure in terms of actual server farms and datacenters and things like that. The tracks weren't ripped up after the railway bubble burst, and the fiber optic cables didn't disappear when the dot-com bubble burst either. In fact, in both cases, we inherited the landscape they prepared for us. The groundwork was all laid in the dot-com bubble. And then all of the companies that we're familiar with today—the Googles, the Facebooks, Instagrams, what have you, because there was this glut of infrastructure—it was cheap for all of them to come to fruition and essentially create the internet world that we're living in. By the time those post-bubble companies came around, they had the luxury of focusing on their products alone. The infrastructure and the technology stack were just waiting for them. Today, there are entire libraries of open source and free software that you can plug and play. If you need to have an e-commerce backend, there's all sorts of plug and play to do that. Not only in the 90s, did you have to code that all yourself, you had to hard code your own website. McCullough points out the work those early companies did played a vital role in our technology's evolution. They were the first on the battlefield, the ones who threw themselves at the unknown. You had to have some people lead the way, and the leaders often are the ones that get the arrows in their back for their troubles, but you have to have generations rise, and then another generation come after them with slightly different ideas, and another generation come after that. I think that's just the nature of capitalism. In fact, it's the nature of evolution. And from 1995 to the bursting bubble, to the world we live in today, that dramatic evolution brought us every fantastic opportunity we have. If you wanted to spin up a little online company today, you could do that from your sofa, but that's only possible because back in 1995, some dreamers started dreaming. We get to make use of all the infrastructure, all the platforms, all the tech that they had to wrestle into being. So yes, there were some bold business decisions made during the bubble. And yes, some of those 1990 startups behaved like teenagers who stole the keys to mom and dad's car. But you know, players like Pets.com and Webvan were also pioneers. They were throwing ideas at the wall, trying to imagine for the first time, what the web might one day become. And out of those huge risks they were taking during the dot-com bubble, came the infrastructure that command line heroes rely on today. All season long we've seen how the world we take for granted was built by coders and engineers who had an audacious vision of what the future could be. In a way, we're all still dreaming the dream that came out of that brilliant, outlandish, game-changing year, 1995. You can learn more about how 1995 changed our online lives by checking out all the bonus material we've collected over at redhat.com/commandlineheroes. And that does it for season 7, but stay tuned because season 8 is already in the works. Until then, I'm Saron Yitbarek, and this is Command Line Heroes. Keep on coding.

More about 科技史

要构想未来，我们应该向历史探寻，找到那些能激发我们灵感、带来震撼的创新故事。

博客文章

采用“快速失败”方法来开发 RHEL 升级自动化

2025 年 12 月 6 日
  |  
Bob Mader
Automation and management
博客文章

红帽智能分析现已更名为红帽 Lightspeed：加速实现 AI 赋能的管理

2025 年 12 月 6 日
  |  
Brent Midwood
Automation and management

About the show

Command Line Heroes

During its run from 2018 to 2022, Command Line Heroes shared the epic true stories of developers, programmers, hackers, geeks, and open source rebels, and how they revolutionized the technology landscape. Relive our journey through tech history, and use #CommandLinePod to share your favorite episodes.