How a Bubble Burst Built a 16-Year Roblox Bet

In 1999 and 2000, every company in Altos Ventures' portfolio went to zero. Despite all the markups, despite all the IPOs, despite looking smart during the bubble—the eventual number was zero.

This devastating lesson transformed Altos Ventures from a typical Silicon Valley feed-to-bigger-VCs model into something radically different. Founded in 1996, Altos is now a global generalist tech investor that thinks in decades, not years, partnering with founders who share their vision of building durable businesses over the long term.

In this interview, Partner Tae Yoon reveals the **three core principles learned from legendary Stanford professor Jack McDonald (known as the West Coast Warren Buffett) that rebuilt Altos from the ground up, their patient capital approach that led to a 16-year Roblox journey, and why they seek 'hedgehog' founders over brilliant but distracted 'foxes.'**

Tae Yoon: My name is Tae. I'm a partner at Altos Ventures. So Altos is a venture capital firm that was started here in Silicon Valley in 1996. As a firm, we are global generalist tech investors, so it's a very wide and flexible mandate. We really focus more on the orientation and alignment with the founders who we work with.

I actually think we are more like life cycle investors. You can think of us as early stage investors typically leading Series A's and after that initial investment we take a very long term approach to company building and investing literally thinking in decades rather than years. We love partnering with founders who also have a similar mindset of building these durable businesses over a long period of time and we have a lot of flexibility and capacity to further support those businesses through additional investment.

As time goes on, there are many cases in our history where we would start very small and early with a particular company and then end up investing hundreds of millions of dollars in addition to the initial check along the entire journey of the company.

Tae Yoon: So in 1996 when the firm was started, we had one singular LP and the idea was to invest in seed stage, very early stage startups, and then feed them to the very well known Silicon Valley VCs to come in and mark them up and lead the Series A. So Ho and Han were classmates from Stanford GSB. They started Altos back in 1996.

I think in the early days when we thought we were good, we were fairly smart enough to spot lots of interesting business ideas, and we realized that as long as we got the founders of these companies to make a bit of progress on management hiring and potentially offer themselves up to be very open to getting a CEO into their companies, as long as we got them prepared and we put that package together, we took it to the larger venture funds, they would mark up the price and invest at a high valuation, they put the new CEO in, then the bankers start showing up and the companies were ready to go public. It was literally like that during the bubble days of the internet. And so we obviously thought that was the formula for being a very good venture capitalist.

That worked until it didn't. In '99, 2000, many of these companies go public, crazy high valuations, everyone looks really smart and then literally all of our companies went to zero. And I think that gave the founders insight that wow like the way we were doing it was just wrong. There is no real value creation. Who cares about markups when the eventual number is 0? So I think they went back to the drawing board and were trying to get back to the fundamentals like why do we invest and who should we invest behind and how do you ultimately create value and build these durable businesses and I think that's when the Warren Buffett Charlie Munger thesis comes into play with the help of Professor Jack McDonald.

Tae Yoon: I actually met Ho and Han back in 2015 when I was at Capital G, Google's growth equity fund. One of the partners who I worked closely with, Gene France, he and Ho and Han were all classmates from Stanford GSB. So through that relationship we looked at a few of Altos's companies, including Toss and Roblox. Ultimately couldn't get there on either of them, which was a big mistake. And then I just always felt like they were really good people and so we always stayed in touch.

And then when I went to Stanford for my own business school experience, I took a class called the Investments class. It was taught for 50 years by this famous professor Jack McDonald. He was regarded as the West Coast Warren Buffett. I was sitting in the classroom and Ho and Anthony just walked in and I was like, wow, what are they doing here? And it turned out that Jack was a very important figure and advisor to Altos when Altos got started.

Tae Yoon: There are 3 core principles that are really important elements of what Jack taught. So the first thing is around the power law. So in any asset class there are only a small number of companies or assets that drive all of the returns. If you look at the public markets, you would look at companies like Apple, Google, Meta, and so forth that have driven almost all of the returns of the S&P 500 over the last several years. This is even more amplified and more true in the world of venture capital where only a small number of companies in every vintage drive all of the returns.

The second one is around compounding. There's this really simple and classic example of should you take $1 million of cash upfront or should you take a penny that doubles every single day until the end of the month? And of course the answer is you should take the penny because it turns out that even in February when you have 28 days, that penny ends up becoming more than a million dollars at the end of the month. If you're lucky enough and you are in a leap year and you have 29 days in that February that becomes $2.5 million plus and then if you're super lucky and you're in a month like January with 31 days it's over $10 million at the end of the month.

And so we believe that a lot of the value that gets created for a startup and a company happens in those latter years when the base gets much bigger and the value keeps accruing in a durable way. So that's why we try to be extremely patient. I think Roblox is probably the most well known example of a company where we first partnered with Dave and the team 16 years ago when we led a round with a $1.5 million dollar investment and up to the pre-IPO round we invested another hundreds of millions of dollars into the company and so we were of course the largest shareholder when the company went public and in some sense we view this still as an early stage company. There is so much that the company is building and a lot of potential to further grow the business from here.

In some sense we believe in this notion of the infinite game and that's the game that we want to play which has no time bound because the compounding of value of these special assets happen in years 10, 15, and 20, not in years 1, 2 and 3. Selfishly we want to generate wonderful returns and we should be super patient.

And then the last idea is around this notion of a hedgehog. There's this Jim Collins analogy from the book Good to Great about hedgehog founders versus fox founders. We have a blog post on our website as well, and hedgehog founders are people who are maniacally focused on trying to solve one problem and trying to do that really, really, really well. Foxes are extremely smart, very cunning, but they often have a lot of things going on. They have a lot of projects they're quite distracted and for us we gravitate towards the hedgehog founders, people who feel like this is their life's work and something that they want to endure through the ups and downs, and we've experienced that those types of founders are the ones who can endure the difficulties of a startup journey. No startup journey is linear and up to the right all the time. There are a lot of challenges that come and being that hedgehog founder, we feel like is an important ingredient to building something great over a long period of time.

Tae Yoon: I think even within our team we have slightly different views on what a hedgehog founder means and how do you identify someone as a hedgehog versus a fox. In some sense it's a spectrum, so it's very difficult to say well you're a fox and you're a hedgehog. It's a spectrum of how hedgehogy are you versus how foxy are you.

I find some of the best founders and oftentimes they are these hedgehogs to have very strong opinions but loosely held, so the strong opinions come from either an inherent connection to the industry or the problem that they're trying to solve and that was the genesis of why they started the company to solve their own problems so there's one bucket of founders who come in that category but then there are other founders who also become extremely entrenched into an industry or a problem through a lot of studying and a lot of conversations and more of a top down approach to trying to figure out how do you solve this problem that I want to solve.

In either case the best founders seem to have at least a strong point of view on what the world should look like and I think that's very important because you have to be a visionary to change the world or bring to the market a service that did not exist before. At the same time, what's really interesting about a lot of these founders is that these opinions are loosely held, so they're strong, but they're willing to change their views if they see something new or if they learn something new and that part is also really important in our view because you have to have the mental plasticity to figure out whether you're right or wrong about something. And have the humbleness to admit when you're wrong that you have to course correct and so this is where the continuous learning mindset comes into play and the best founders and the best learners tend to have that characteristic.

Tae Yoon: I've been working with and partnering with a number of vertical market software companies. One company is called Push Press, which is a gym management software company. We first led the series A 2.5 years ago, what they're trying to do is to provide the best software to independent gym owners across the US and across the world.

The founders were former CrossFit gym owners. They had a pain point of finding a software tool that worked for their own gym. So the founder, who had a technical background, just built a software tool for his own gym that ended up becoming Push Press. Now they have about 100 people at the company of which the majority were former gym owners or former fitness professionals. They live and breathe the category, the industry that they serve, and it shows in the way they approach customer support and the way they talk to their customers. But that's been a really fun journey so far.

Tae Yoon: A lot of founders would come and say we are looking for the next 10 year type of partner. We're very serious about finding the right partner. It's not about valuation. Oh and by the way we need a term sheet by tomorrow. That's very hard for us because how do you get married after the first date and we view the partnership as a marriage and so we really appreciate having the time to get to know founders and get to know each other and I think that is actually really important to building a very long term and fruitful relationship because let's say you enter a relationship without having that alignment well in the first board meeting or that second date it's going to be very apparent that this is not a good fit and what's the point of that if the goal was to build a generational company for 10+ years.

Tae Yoon: I have one definition or one analogy that I personally really like. I was at my son's preschool and there was a teacher parent conference where the head of the school was giving a speech to the parents and she made this analogy of 'you parents know your child the best, we teachers know children,' so if we partner well together it can be an extremely fruitful partnership and a symbiotic relationship and I was listening to that and I was like hm that's kind of like my job as an investor.

There is no way that we as investors and venture capitalists would know a singular company more than a founder or the team that's working hard at it but we do see a lot of companies and we end up working with a lot of companies over the years. And that pattern recognition or the sheer amount of data that you can aggregate of seeing the good ones, the OK ones, the bad ones, the full gamut of company building allows us to hopefully provide guidance when it's needed for these businesses to avoid mistakes or accelerate their path to greatness and so in some sense maybe I was thinking to myself a venture capitalist is like a preschool teacher.

Where you're trying to be a helpful steward but you should also be humble enough to know that there's so much more to that singular child or company than what you probably see so you have to respect that and work with the parent or the founder to understand it as much as you can and then with that context and the partnership you can provide the insights and the support to hopefully accelerate that path.

What we are trying to do is of course to not be average and hopefully add value and maybe the way to add value to a company's trajectory is increasing their odds of success. Again the success is still unclear whether you can succeed or not as a company, as a founder, or even as investors. There's a lot of inputs that go into that eventual outcome which hopefully is great, but there's luck and there's a lot of other circumstances that have to go your way for that to happen so then what is it that we can do? Well, we can definitely control the inputs to getting there and hopefully the partnership with a venture capitalist or an investor is a strong partnership that allows the thinking to keep expanding and for the VC to help steward and guide the company in a direction that increases the odds of success. So I really like that framing of the role as well.