Bubble Memory: A CEO’s Reflections on the Dotcom Boom and Bust
Steve Case’s lessons for founders building in the AI gold rush.
Friends,
Perhaps no one experienced the dotcom boom and bust quite as viscerally as Steve Case. The founder of Revolution began the decade as an executive at a small software company called Quantum Link, and ended it as the CEO of a behemoth: AOL.
It is hard to overstate just how important a company AOL was at the start of the millennium. Unless you were already of working age and minded towards tech and business, there’s a good chance you haven’t calibrated its size appropriately. AOL was a giant. Between its IPO in 1992 and the start of the new millennium, its stock appreciated approximately 80,000%, reaching a peak of $222 billion in December 1999. By that point, it had eclipsed many of the business world’s biggest companies, surpassing Disney, IBM, McDonald’s, and Berkshire Hathaway. At the time, AOL was larger than Boeing and General Motors combined.
It was also wildly significant and influential, responsible for opening the internet up to a new generation of users and making it fun to use. The AOL of that era pioneered many of the user experiences and behaviors that make up the modern web, from screen names to instant messaging.
In January 2000, AOL looked set for another decade of dominance. On the tenth of that month, it announced a planned merger with Time Warner. Remarkably, AOL was the senior partner, receiving 55% of the combined $350 billion entity. Again, the scale of the deal is hard to comprehend unless you lived it. At the headline valuation, AOL Time Warner became the fourth largest company on the planet – a figure greater than “the output of Russia or the Netherlands.”
What came after is the subject of much study and coverage. The market crashed, the merger failed, and not long after, Case stepped down as AOL CEO. By 2005, Case publicly argued in favor of the two companies splitting, and at the end of the decade, that came to fruition. Though there were strong strategic merits to an AOL-Time Warner union, it proved to be a marriage that suited neither party.
It seems as if we are in yet another torrid cycle and potentially one of greater magnitude than that which Case surfed. A recent analyst report argued that the current AI bubble is 17x greater than its dotcom corollary, and 4x larger than the 2008 financial crisis. (We can hope it will be more productive than the financial crash, at least.)
A bubble is a strange environment in which to live. It is an even stranger one in which to build a business. How do you keep your feet on the ground when the world beneath is rushing past at a thousand miles an hour? How can you hold fast to reality when wide-eyed investors, gasping for breath, paint ever more fantastical futures in front of you and shove your pockets full of cash? Where is the boundary? Is there really such a thing as a horizon?
We believe that “this time is different,” and there is truth in that. We do not live in a perfectly iterated game. The board has changed, fresh players have stepped onto it, and new moves are always being invented. But it is more naive to pretend the past has nothing to teach us. In fundamental ways, today’s founders face the same set of challenges as their late 90s forerunners. How do you build a team at warp speed? Is it possible to scale a culture when you’re growing exponentially? How can you make smart acquisitions and avoid the most common missteps? What role should partnerships and business development play when you’re forging a new category?
As I thought about these questions, I realized that perhaps no one is better suited to speak from experience than Steve Case himself. And so, I asked him to reflect on his journey at AOL, the similarities and differences between that era and ours, and the lessons founders should heed. We also discuss what it was like to compete head-to-head against Bill Gates, the strategic necessity that drove the Time Warner deal, and the similarities between Jeff Bezos and Steve Jobs.
Brought to you by Mercury
What gets measured gets managed.
And when you’re juggling multiple portfolio companies, having a crystal clear measure of cash burn is invaluable. That’s why you’ll want to keep Mercury’s cash burn rate calculator in your toolkit — and pass it along to the founders you’re backing.
In seconds, you can see how many months of runway remain at a startup’s current burn. That kind of clarity can help guide make-or-break decisions around spend, headcount, and future funding rounds.
Mercury does banking* differently. They don’t stop at banking and financial software to help businesses move money. They help businesses stay two steps ahead with the help of free tools and resources like this.
*Mercury is a financial technology company, not a bank. Banking services provided through Choice Financial Group, Column N.A., and Evolve Bank & Trust; Members FDIC.
An Interview with Steve Case
To start, I’d love to connect the two eras we’re talking about and the lessons we might take from them. How much does the current AI wave remind you of the dotcom bubble?
It has some similarities, but some differences, too.
Like the internet, AI has been many years in the making. When you look at AI, some of the core technologies can be traced back to something like 70 years ago. This is not a new idea. It just took that time for it to become a mainstream consumer phenomenon. Obviously, over the last ten to twenty years, a lot of companies have integrated AI. Most of the apps we use on our phones rely on AI to some extent. The difference is that nobody called it AI. But it was driving recommendation engines and so on; it was just more of a back-end, invisible thing. It only became an in-your-face visible technology with the success of ChatGPT.
That timing dynamic reminds me of the dotcom era. The gold rush mentality does too. People feel that this is the next big thing, and they don’t want to miss out, which leads to investors pouring a lot of money into companies. A few of those will be wildly successful, many will be OK, and a good number will also probably end up hitting the wall and going out of business, which is what happened with the internet.
Clearly, AI as a core technology and platform is going to have a comparable or greater impact to the internet. It has the ability to transform industries and society in very fundamental ways, but there will be a mix of successes and failures.
A final point I’d make is about government and policy. What most people don’t appreciate about the internet is that the United States wouldn’t have been the leader it was with respect to that technology were it not for a series of government decisions: first, you have the government’s decision to fund the Defense Department’s ARPANET, which basically became the internet. Next, there was the Justice Department’s decision in 1984 to break up AT&T into seven regional operating companies, which drove communication costs down and made the internet much more affordable. Third, there was the decision by the FCC to mandate open access for the telephone networks so that they had to let AOL and others plug into them. And, finally, in the early 90s, Congress passed a law opening up the internet to consumers and businesses. Up until that point, it had been restricted to government agencies and educational institutions.
It’s easy to forget these things, but if they hadn’t happened, the internet would not have evolved the way it did. In the case of AI, it’s really important that we figure out what that policy overlay should be. How do you make sure that while the technology is being developed, you don’t over-regulate and stifle innovation? Europe has historically made the mistake of focusing too much on what might go wrong and, as a result, stifles what might go right.
At the same time, though, because AI is so ubiquitous already (versus the internet, which grew slowly before mass adoption), some engagement, some policy guardrails are necessary now.
What might some common-sense guidelines look like?
One of the lessons from the internet is that you need to mandate open access. When it comes to AI, I think we need to make sure the big platform companies are essentially required to have open access. We don’t want to create a system that just allows the big companies to dominate; we have to force them to open their networks and keep them open for new companies to participate and innovate.
You had the rare experience of building what was considered the most significant company of its time during a historic bull market. It remains to be seen to what extent AI is a bubble, but I suspect the lessons you learned building AOL during the dotcom boom are deeply relevant for today’s founders.
As a jumping-off point, how did you manage your mental game during that period? AOL went from a relatively small company to one of the most important on the planet in a short period of time. The company must have been changing noticeably almost every day.
Part of my job was to be a shock absorber for the rest of the company. As the CEO, you have to even out the highs and lows for people. When things are going great, remind them about the risk. I used to call that “delegating paranoia.” When things are down in the dumps, remind them of why you have a bright future. There were plenty of moments like that at AOL – we’d be on the cover of a magazine with the headline saying “AOL is going to go under for XYZ reason.” It’s your role to provide that balance so that people can focus on the overall mission.
It probably helped that it took more than a decade for us to get real traction. I joke that AOL was a ten-year-in-the-making overnight success. The time it took us probably gave all of us, including me, more perspective and humility when all of a sudden it was this go-go-go momentum. It helped me not buy into all the positives or all the negatives.
Beyond that, you have to surround yourself with good advisors, including board members. They were helpful in making sure we were taking a step back at the right times and not just focusing on the day-to-day so much that we missed something. You have to learn to look at the company not just as a snapshot but as a movie that’s playing out.
I imagine that one of the biggest challenges of scaling that fast is finding enough great people who understand what you’re building, and creating the right structure to get the best out of them. Although AI founders today can probably get more leverage on people than during the 1990s, great talent is still essential.
What was your approach to recruiting and organizing talent, and how did it evolve with the business?
Well, it definitely evolved. In the early days, when we were just a couple dozen people, it was about surrounding ourselves with true believers. This was before people knew what the internet was, so there weren’t that many people interested in it. We were looking for people who wanted to be a part of the journey and could add value to it.
When headcount got into the hundreds, we started to be a bit more precise in terms of the roles we needed. You start to run searches for roles and look in a more structured, disciplined way.
When we got into the thousands, and things really started accelerating, I realized that as the CEO, I couldn’t be involved in everything. And so I needed to shift my mindset to spend the majority of my time on recruiting. It might not have been 50% of my actual hours, but it was more than 50% of my attention.
A lot of my job in those days was not just promoting AOL but evangelizing the internet. I’d travel a lot and speak at a lot of conferences – Allen&Co in Sun Valley and others. I always had my radar on: Who’s going to be there? Are there people who might be partners in a way that works for us and works for them? Is there anyone I’m really impressed with that I can convince to leave what they’re doing and join us?
Every meeting I was in, every conference I was attending – whatever it was that I was doing – looking for great people was always on my agenda. With the right people and the right focus, anything is possible.
You mentioned that as the company scaled, you felt you couldn’t do everything anymore. The pendulum in tech seems to have swung towards CEOs feeling like they should micro-manage more and have their hands on every part of the business. Why did that feel like the wrong fit for you?