Blitzscaling: Uber so với Lyft, Bẫy của người đi trước và Luật quyền lực VC - E539

“Network effects are actually one of the few things that are truly special about the digital world. And a part of it is because, for most digital stuff, the cost of replication is effectively 0, right? But if we talked about a USB thumb drive—if I use a thumb drive and you're also using a thumb drive—then actually both our experiences get better because we can swap files with each other, right? And so we see that a lot for ChatGPT as well. When you use ChatGPT—and I use ChatGPT—when you use ChatGPT, you are teaching the AI model how to respond to you. And because the AI is learning from you, my experience is getting better. So I want everybody to use ChatGPT because my experience will get better.” - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast

“If you told me, “Hey Jeremy, you pay one dollar and you get 15 dollars back,” or if you put one dollar into buying land in Singapore and you're going to get 15 dollars back eventually in the next ten years, I'll be like, “Oh my god, I need to go get as much land as possible because it's a great deal,” right? Whereas if you tell me, “Hey Jeremy, you put in one dollar to buy land today, but you only get one dollar and ten cents after ten years,” I'll be like, “Forget about it, I’ve got other things to do with my life.” So I think understanding unit economics is very important because that explains why you see the phenomenon of many startups that seem to go crazy, right?” - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast

“So there's a time allocation that you need to think about constantly. As a result, you need to be thoughtful about each funding round. Companies typically require more capital to grow every one to two years. For example, Simon pointed out that a company had scaled a bit—after putting in a million dollars, two years later, they were looking for five million more. At that point, you have to decide: Should I invest more money? Should I invest my pro rata, meaning the allocation I deserve? Can I double down? Can I ask for more because the company is performing exceptionally well? Or should I skip the deal? Most recently, in a fund, we evaluated a portfolio company and saw they were doing really well. I immediately called the team and said, “We need to make a decision.” After discussion, we concluded that we wouldn’t just keep our pro rata—we wanted to double down.” - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast

Jeremy Au explored the strategic dynamics of venture capital, network effects, and first-mover risks in technology markets. He illustrated how Uber outmaneuvered Lyft by aggressively raising capital to fuel network effects, while Airbnb achieved superior capital efficiency through globally pooled inventory. Using the case of Henn Tan, the Singaporean inventor of the USB thumb drive, he highlighted the vulnerabilities of weak patent enforcement and the unintended benefits of fast followers in driving mass adoption. He also examined how VCs allocate capital, using Instacart’s funding history to demonstrate the high returns of early investors versus the diminishing gains of late-stage backers. Through the lens of blitzscaling, he emphasized the importance of understanding unit economics—where strategic capital deployment can lead to market dominance, while miscalculations, as seen in the failed bike-sharing boom, can result in rapid collapse.

(01:04) So this is Trek. This guy invented thumb drive. You have never heard of this guy because he's not rich. He just came out of jail recently. So we'll talk about that story.

(01:14) this guy was the inventor of the USB thumb drive. And so basically what was happened was that he was part of a company. And there was at a time in the 1970s, 1980s, obviously. Singapore was like the Vietnam of Southeast Asia for electronics manufacturing. And they were basically creating these things called MP3 players, which was a very new type of technology, in the sense that, before there were CD players, before there obviously was the LaserDisc, and so forth.

(01:39) Obviously, they're making MP3 players. And his insight was that he wanted 

(01:45) to simplify it and not have, a headphone jack, not have the audio side. He just wanted to keep the flash memory of it, but make it compatible with many different computers.

(01:55) Because he noticed that he found it difficult to transfer files from place to place. And I (02:00) think, I don't even remember, but before this, some people were using Iomega Zip drives. I don't remember this very thick, chunky drive. Some people were using the floppy disks, those, the small ones. If you don't remember, then congratulations.

(02:10) you don't worry about longevity or anything about that. but for those, I think everyone should have seen a thumb drive. And so it was quite revolutionary to have that thumb drive, right? So this guy is called Henn Tan. He owns a company called Trek. And basically he created this company that, and so he was the first mover, because he was the first person in the world to simplify an MP3 player into a thumb drive.

(02:30) And he thought it was going to make it rich, and he started making quite a bit of money because it became very popular. Unfortunately he had two employees and perhaps there was a dispute or whatever it was, but those two employees moved back to Shenzhen, and then they set up another company that started selling thumb drives.

(02:45) Long story short is the Chinese company was granted the patents to manufacture in China. And then he had a patent for this in Singapore and a few Western countries. he's definitely a millionaire because he was able to license that patent with some Japanese companies and American (03:00) companies that wanted that.

(03:00) But unfortunately as we all know from the heyday of 1990s was that Shenzhen gave zero percentage care about this thing. And also the other was granted a court of law in China that they had patents. So as a result both companies grew in a sense, but most of that volume was captured by Shenzhen.

(03:17) And even that Chinese company struggled to enforce its patent claims against other companies in Shenzhen who were also not respecting their patent in that sense. Eventually, he went to jail because the company was doing well, and then he started falsifying accounts, and then Singapore sent him to jail.

(03:34) So that's one. The other company that is out there. I forgot the name, but I'll send it to you. But they are a publicly listed company in China right now. So this is a good example of it, right? And so going back to this example he was the first mover, but the Chinese company was a fast follower.

(03:49) There were network effects, obviously. So in the sense that once the thumb drive came to existence, everybody was like, we need to create thumb ports, right? And then everybody started using thumb drives. Because if you had a USB port Then everybody starts (04:00) to use that same protocol, right? So the product gets better and better the more people use that USB thumb drive, right? So I think some people even make a counter argument that if not for the Chinese not respecting his and making it effectively very cheap thumb drives, then he would never have succeeded to the whole extent anyway because Chinese oversupply of thumb drives basically created the default where everybody uses a thumb drive, right?

(04:24) That would be One version of the counter argument. Even though Henn Tan, for example, is at Trek his patents were there in Singapore, he did manufacturing in Singapore and eventually Shenzhen as well, right?

(04:33) The Economies of scale, you create lots of thumb drives. Obviously, he had IP patents that were granted and he was able to make some money. The Chinese company from his form piece also had IP and patents that they enforced against Chinese manufacturers. So they tried to do that but neither was able to fully enforce to the likelihood they wanted to and regulatory, he was unable to bring judge damages or, lawsuits against the Chinese entity, et cetera. And then, don't lose your employees to move to another country to (05:00) manufacture your exact same product, right? So I think we need to think about network effects as well.

(05:04) I think the reason why we want to think about network effects is that network effects are actually one of the few things that are truly special about the digital world. And a part of it is because most digital stuff the cost of replication is effectively zero, right? But if we talked about a USB thumb drive, if I use a thumb drive, and you're also using a thumb drive, then actually both our experiences get better because we can swap with each other those files, right?

(05:27) And so we see that a lot for ChatGPT as well. When you use ChatGPT you are teaching the AI model how to respond to you. And because the AI is learning from you, my experience is getting better. So I want everybody to use ChatGPT because my experience will get better because you are all training for my benefit. So network effects is really important. Because again, it is based on two things, right? One is zero marginal cost or low marginal cost to serve an additional customer. And other side is the ability to pull that synergies in that outweighs that, right? And so If you look at (06:00) Uber, this is a classic example, was that when Uber first came out, a lot of people said this would never be a billion dollar company.

(06:05) And the reason why was that when people looked at Uber, they said that taxi fleets were only making millions of dollars. Does it make sense? In a city, right? So New York, the taxi medallion industry was like, okay, taxi industry only makes about 10 to 20 million dollars per year. You'll never be a hundred million dollar, a billion dollar company because the total size of taxis as it, the behavior currently stands is very small.

(06:27) What Bill Gurley kind of did on this napkin, which has been recreated, but not exactly, but basically said, actually Uber actually creates a larger market size by existing. For example, but Uber brings on, let's just say, more demand to, because they're through marketing, et cetera. So we have more demand, you will have more drivers, because drivers want to come and serve those.

(06:47) Riders, right? Because the taxi fleet is fixed to 10,000 vehicles, for example. But now, if you have more riders, then you have, 100,000 drivers, because more drivers want to come in. But when you have more drivers, you have more geographic coverage, and you have more saturation, (07:00) right? So there are more drivers in that neighborhood, square kilometer, as a result.

(07:04) Because you have more density in that neighborhood, then your pickup is faster, right? So instead of waiting for 20 minutes for a taxi to come, now because there's so many more drivers now it only takes you two minutes for a pickup. And because you know that This is only going to take you two minutes for you to wait instead of 20 minutes.

(07:22) Then, some people are going to say, you know what, I'm not going to take the bus, I'm going to train, I might as well use the Uber. And as a result, more people will join this pool of driver riders, which will drive this flywheel effect. The other part that we talk about as well is that if you have more density saturation, then the drivers have less of downtime, because they're closer to the customer instead of driving.

(07:42) Because when they are driving 20 minutes, they're also losing 20 minutes worth of time, right? And so because they have less downtime, because there's more density of riders as well. So instead of spending a lot of time, a lot, historically, you go think back 10 years ago, a lot of taxi drivers were waiting in the airport, queue, right?

(07:57) Or they were just chilling because they didn't know where the riders (08:00) were. But now you have less downtime, and because you have less downtime, the driver tries to make about 30 an hour. Your ticket prices will go up, right? But if your whole hour is being used, then the prices will go lower, become cheaper, and because the prices are cheaper, more riders will drop, right?

(08:14) More will come in as a result. So there's this flywheel that drives over time. And so this is really important to know, because if you think about it, Lyft and Uber started roughly at about the same time, one on the West Coast, one on the East Coast, and you would say that Uber basically raised a lot more capital, they got very aggressive, and what you see here is that Uber was at about almost like 80% of the market share at this point from 2017.

(08:36) And then this gap started widening, to about a good chunk because of Travis Kalanick, the CEO, had some PR issues, so it was, but basically it was a media firestorm. So Lyft was able to out compete in that sense, but it was still always smaller than Uber. Obviously COVID crushed them both, but you see them come back again, and Uber is slowly starting to kill Lyft again, right?

(08:57) So no matter what Lyft does, as (09:00) effectively, they can't really beat Uber, right? And 74 percent of the market share. And Lyft has dropped from 38 percent market share to 26%. So no matter what Lyft does Lyft stock has plunged, dropped by 90 percent since it went public in 2019.

(09:13) So basically, the market doesn't feel like Lyft will survive. Why? Not because Lyft is a worse company. Lyft and Uber are effectively the same in terms of services, etc. But because Uber has network effects, has more density, by nature it's able to slowly choke out. Smaller competitors when you have a network effect, right?

(09:29) And if you think about it, there's also the reason why there's so many companies that create what they call competing ends, right? So for example, you look at but there are often alliances between manufacturers and software providers to create new standards.

(09:42) So for example, USB is the standard today, but there are competing standards, right? And USB1 because you have more network effects. So there's a alliance effect that often happens. And so what we want to see here is that, however, if we think about it, Airbnb was also founded around the same time.

(09:56) It's also the same generation of sharing economy, right? (10:00) Using a dead asset and creating more efficiency from the asset. 

(10:03) But Airbnb has created five times more value than Uber and Lyft. If you think about it, Airbnb was much more capital efficient because, if you look at this, they raised about a billion dollars and then they were able to get about 30, 40 billion dollar valuation, but you look at Uber, they raised 25 billion dollars, and they generated about 60, 70 billion in terms of market cap, and you look at Lyft.

(10:23) Obviously, and then they only raised about five billion dollars and they got to about, ten billion dollars of market cap. So I think, what I'm trying to say here is Uber raised a lot more money early on successfully. They grew larger, and as a result, a reward of a larger market cap, and they're slowly gaining share.

(10:37) Whereas Lyft is slowly deteriorating over time, but Lyft didn't raise enough money. 

(10:42) We'll talk about that because it created this concept called blitzscaling. if you have network effects, always raise a lot of money to create network effects that will slowly kill your opponents over time.

(10:50) But how it was interesting for Airbnb is that it's actually. More efficient than that. So the question you want to think about is, why is Airbnb, even though it has the same thing called network effects, more capital (11:00) efficient? one way to think about it, and this only emerged over the past five years in terms of understanding, is that first of all, even though we say that Uber has network effects and that was that flywheel that we had, it turns out there's also a geographic dispersion of that effect.

(11:11) what I mean by that is the simplest way to think about it is that if I am Uber in Singapore, and I'm number one in America, it has nothing to do with my ability to do anything in Jakarta, right? So Uber had to fight every city as it is. Uber having more drivers in America has nothing to do because you need Singaporean drivers for Singapore, you need Indonesian drivers for Indonesia.

(11:35) So first of all, the network effects do not cross borders because it's a car, right? That's one. But two, is that even within the same country, it doesn't matter, because if you are in Indonesia, and you are number one in Jakarta, but you are number one in Medan or whatever other city that's there, it doesn't really matter, because across cities, it doesn't matter in the same country.

(11:55) So you need to fight in every city. And if you think about it, even within cities, (12:00) within neighborhoods, there are also very weak clusters. So for example, if you and I looked at our phone and we just said, we want to go home and take a grab after this, we don't care if there's a car in woodlands in the north of Singapore because it's too far away.

(12:11) We only care about the density, the network effects of the riders and drivers within this one kilometer zone of us. So what I'm trying to say here is that even though there are network effects, It's actually quite neighborhood centric. It's very geographic dispersed. So that model that we saw that was conceived 15 years ago was too simplistic in that sense.

(12:29) So now we understand and we better understand that it's a geographic dispersion. But what's interesting about Airbnb is that there is no geographic dispersion of the network effect. So what we mean by that is that if you list home, not in Singapore because it's illegal, but if it is a home in, say Kuala Lumpur that you have there, anybody around the world can access the inventory, right?

(12:50) So you're traveling from Europe. So your inventory that's added anywhere in the world benefits every other traveler in the world. Does that make sense? Yeah, it's not as if your room that you live in KL is only (13:00) available to people who are searching within KL, right?

(13:02) In fact, it gets more value, right? Because you're from America, you're traveling to KL for a meeting, you want to live in it, blah, blah, blah. Airbnb, the network effects are much stronger as a result because there's a global pooling of that benefit. Does that make sense? And so I think it's that's why what I'm trying to say here, now we are saying this, is that network, the more people you add, the more network effects that you get.

(13:23) The more separated the people are, the less network effects you get. So you always so what you notice, and we talk about ChatGPT, we talk about Apple, you see everybody. So everybody's trying to squeeze everybody into a denser network, into this bubble to get more network effects. That's why everybody's trying to get you to, resign your terms of conditions all the time.

(13:41) So as a result when you think about it and you're thinking about this, your deal memo, et cetera, 

(13:45) you should be understanding that your company is a, what is the price, what is the margins, what is the lifetime that a customer has, and what is the upsell that you can have over time. So these are the four components when you think of revenue in terms of your deal memo, for example.

(13:58) So for example, if you're (14:00) doing a software as a service as a customer this is a real example of a company in Southeast Asia. What they're selling is that every company, they're charging about 1,000 per month. And the reason why is that this company is about 200 employees, and this company is charging them 5.

(14:13) per employee, if that makes sense. 

(14:15) So that comes up to 1,000 per month. And then because it's a software, there's very little cost, there's no marginal cost to it. It's all digital, so they have margins of about 75 percent And then they expect that, if you're using my software, you're gonna stay on my software for at least 10 years, maybe even longer.

(14:30) So you're never gonna uninstall the software at this company that you install. And then every year, we expect to sell 10 percent more. Next year, you spend about 10 percent more. Because on average, companies add more employees over time.

(14:43) They also can sell and cross out more and more deals over time. So they can raise prices over time. And so if you add this times this times this, that comes up to roughly, the customer lifetime value. A customer walking in is worth about 100,000 upfront. So if you think about it, when you sell the customer, you're only charging (15:00) them about 10,000 per year in the short term.

(15:03) But actually, if you look at them in the lifetime of that customer, it's actually 100,000 of lifetime value, maybe even higher, right? And so we also need to be thinking about the customer acquisition costs in your deal memos. So what we mean by that is for this company, when they get new customers, it's a function of marketing.

(15:21) Plus sales, plus onboarding, minus attrition. Okay? Let me explain this. So for example, for the same company, software as a company, so their cost was that they were doing about 5,000 of marketing. So they're spending maybe, let's say, 100,000 of marketing cost in a year. But obviously, in that year, there's about 20 customers will join.

(15:39) So 100,000 per year divided by 20 companies that joined this year, comes up to about 5,000 of marketing they spend, on a per company basis. Then, they also quantify how much time does it take to handshake the person, right? So to get them onboarded. Your sales rep times the number of hours they use, spending is about 1, 200.

(15:57) And then, they have customer support. (16:00) So they spend a lot of time making sure that this person is onboarded correctly. So that costs about 1, 100. It's a cheaper person, but they spend more hours getting set up, migration, etc. And the good news for them is that only 1 percent of the people will quit. From the funnel after they signed this contract, right?

(16:16) So as a result, if you think about it, is that this company is worth 100,000 as a pot of money, but then the cost to acquire them is only 7,000, right? So that means that the ratio of those two things, this divided by this, is roughly about 12 to 20 times, right?

(16:34) it's about 12 to 20 times more value versus the cost to acquire the customer, right? And of course, if you think about it carefully as well, because they only spent 7,000 to acquire this customer, but They collect about 10,000 in the first year.

(16:47) So it takes about 12 months to pay back on a accounting basis. But a lot of these companies will pay up front. So the cash payback is within three months. But the way I'm trying to say here is that if you look at this exact same company as a result,

(16:59) if the (17:00) company invest 1 into marketing or acquisition or whatever it is, they get 15 back, right?

(17:06) you put in 7,000, you get about 100,000, right? So you get, 15 times more money. So you invest 1, you get 15,000 back. So as a result, I think a lot of VCs start to say, wow. For every dollar that you're spending, you're getting 15 back of economic value.

(17:22) So this is a time to hit the gas, put a lot of money into the company. And so that's what Uber did that Lyft didn't do. and so that's the positive version, which is that if you understand that your unit economics of this business is really good, then this company needs to grow really fast.

(17:37) Because there's a once. in a lifetime opportunity to seize as much land as possible, right? Because if you told me, Hey, Jeremy, you put 1 into buying land in Singapore, and you're gonna get 15 back eventually in the next 10 years, I'll be like, oh my god, I need to go get as much land as possible because it's a great deal, right?

(17:54) Whereas if I say, you tell me, Hey, Jeremy, you put in 1 to buy land today, but you only get (18:00) 10 cents after 10 years, I'll be like, forget about it. I've got other things to do with my life, right? So I think understanding the unit economics is very important because that creates and while you see the phenomenon of many startups go, they seem to go crazy, right?

(18:12) So a lot of us, for example, we saw the bike sharing wars, right? So a lot of us saw the, remember the wheels, like there was bicycles everywhere. And then all these Chinese funded startups. And also American and local funded start ups started having all these bicycles. And then there was like so many fleas and bicycles everywhere.

(18:31) Now all of them got thrown away. But because all of them made a calculation before and probably made a wrong calculation, they thought they had positive economics. they thought they had the blitz scale. So they raised a lot of money to do that, right? So there's something we like to call negative blitz scaling.

(18:47) And what I mean by that is, Instead of having positive unit economics or strong unit economics, you have negative unit economics, and you throw more money into it, and then you lose a lot of money very fast. but the crux of it is that if you believe that this company has very strong (19:00) economics, the experiment has worked, you know that putting 1 in will give you 15 back.

(19:05) Then you as a VC should put in 10 million in order to get 150 million back, right? Now obviously there's a lot of can go wrong. You put in 10 million, the founder goes crazy, he doesn't know how to spend it, you waste the money, you don't get 150 million back. But at a point of investing, you should have that belief structure to be like, putting my money in now will give you a lot of value back. 

(19:26) First of all is that you as a VC will be doing two sets of activities after you do the investment that we talked about, writing the memo. First of all, either you'll be adding value, or you're doing judgment around your portfolio management. So adding value is like board work, strategy, compliance, financial reporting, helping them hire, giving them connections.

(19:43) Jeremy Au: The other way that you can do it is by managing your portfolio. So basically you're assessing them, you're re evaluating them after six months, after one year. So for example, recently it was the VC fund discussion, we were like, wow, this team after six months has really outperformed.

(19:56) And the guy who's voted no on the deal has been quite impressed by their growth (20:00) and so forth. We have also seen the opposite, where we're like, okay after one year, this team has disappointed us, not because the team sucks, but because the country is finding it difficult, and it turns out to be more difficult to make money in this country, and it's not because of the team, right?

(20:13) So you're making an evaluation, you're prioritizing. your home runs versus on track versus off track. So if it's home run, you wanna spend a lot of time because you're gonna become a unicorn. If they're on track, probably spend a little bit of time. And if they're off track, you gotta make a decision. Are you gonna save them or not? So obviously, you have to allocate your time, resources, and attention. So for example, you're a solo GP, you have a, 30 million fund. You have 20 companies that are looking for help. You only have, 40 hours, 50 hours, 60 hours a week. You don't have enough time to do sourcing, LP relations, and to do this thing.

(20:43) Adding value as well. And what you have to understand about VCs is that they are always scoring their portfolio. They're always thinking to themselves, who should I spend time with? I need to prioritize making sure that companies that can become home runs. And then we have to support some companies.

(20:57) Maybe I have to delegate it and get someone else to do my time (21:00) or I just ghost them and not help them anymore. If you look at this, for example this would be a good example of come, let's just say that this. is that has about, let's say, two funds. Okay. So it's an index approach. So they're doing 50 companies per fund.

(21:14) And right now they have 100 companies in their two funds they've done so far. And so if you look at that, let's just say that the portfolio 50 percent is a flat out loss, right? Then theoretically the 25 percent will be saved like is this, you get your money back but that's it, you put in one million dollars, you get one million dollars back, right?

(21:34) Then you small win is you put a million dollars, you probably get 10 percent back, right? Then a large win, maybe you get like 2x, 3x, you put a million dollars in, after 5 10 years you get Two or three million dollars. And then a unicorn is like you put in one million dollars, you get a hundred million dollars back, right?

(21:51) So imagine that we talked about this over again about the power law, right? imagine if you are managing these two funds, each of them has 50 funds each. You're saying, who should I spend time with? Do I spend (22:00) time with the unicorns? Do I spend time with the large winds trying to make them into unicorns?

(22:04) Do I make a decision to help the small winds to make them become large winds? Does it make sense? So people always have to think about their time. And so theoretically, of course, small save is you get 50 percent on dollar, small win you get three times on a dollar, large win is 15 times on a dollar, and unicorn is you get 50 times on a dollar, right?

(22:19) So that's how you should be thinking about it. But of course, the paradox that we always say in VC is that, your best winners, they're very good at crushing it. And of course, what we know is that when you're a winner, everybody wants to help you, right? who's gonna say no to Travis Kalanick if he wants, he's doing well and he's on top, everyone wants to help him.

(22:33) So you don't need to help him that much because everybody wants to help him. Paradoxically, the companies are doing the worst. They're struggling very hard, they're gonna want the most of your help. Does it make sense? so you need to spend a lot of time those losers. But the question is that how do you proportion your time, should you put your time in this cluster so there's a time allocation that you need to think about all the time. And so what you need to be thoughtful about as a result is that, every round, what we talked about is that companies about (23:00) every year to two years, they will require more capital to grow. So Simon was like, hey, this company was a bit scaled.

(23:05) So yeah, you put in a million dollars already. Now two years later, they're looking for 5 million. So you're making a decision. Should I invest more money? Should I invest my pro rata? That means I should put in my allocation that I deserve. Can I double down? Can I ask for more? Because I think they're doing really so well, right? Or should I just skip the deal, Most recently, when we were in a fund, we had a portfolio company that we looked at, and then we talked to the company, and I was like, this company is doing so well and I immediately called the team, I said, we've got to make a decision, and then we said, okay, we're not going to keep our pro rata, while we justify, we want to double down.

(23:38) First, we told them that we want to expand our allocation, We were told that the round is already oversubscribed, so it's no more. Who told me to spend ten hours to think about a deal? I should have done it in one hour. Does that make sense? You always have to make that decision all the time, which is do I put in more money in the company?

(23:53) A lot more money? A bit more money, or I'll just say no more money for you for this next round. Because the company is saying I don't want to scale. (24:00) But of course, you have to be thinking about it. So this is a good example of Instacart. And so you look at Instacart in this company. So they receive.

(24:10) Series. A, B, C, D, E, F, then GHI, then the IPO, yeah, for Instacart. Okay. So you can see who the investors are. So obviously the equity price, the price went up, right? So 35, 35, 9, 30 4,000, 6,000, 10,000, 15,000, 39,000. And so this is the. equity price of the value of each share that they had at that point of time, the value.

(24:33) And so your investors, Cosla, Canon, Sequoia, Anderson Horowitz, Kleiner Perkins, Sequoia they came in again. So each of this is a discrete funding event, right? Because it's about every year, every two years, they're making that funding event, right? DST Global, and then Sequoia came in for the Series I, see that?

(24:47) So every check, you're making a decision, okay? If you put in that first check, you have gotten the equivalent of the compounded return of 55 percent. Over this time period, the compounded annual return of S& P is about 13 percent. So you have (25:00) incremental outperformance of 42 percent.

(25:02) See that? So if you were to first check into Instacart, you would have made 50 percent upside up tick every year, which is crazy, right? So if you go to DBS, it's like, what, 1 percent? If you're lucky, right? So you're losing money versus S& P, which is 30 percent, huh, during this time frame, right? so what is interesting is that you see this decision.

(25:18) Sequoia came in for the second time. none of these people wanted to come in again because they were like, we don't think it works. But this second check did really well because they came in at the same price. But, there's one year less from the IPO.

(25:30) So they made even more return, 62 percent. They made a higher return. And then Anderson Horowitz came in for this, right? And then they made 29 percent, right? Year on year, right? And then Kleiner Perkins came in for a new time. They only made 10 percent at the CBC, right? Versus S& P, 12%. And then, Sequoia was like, Oh, you know what?

(25:49) But now is the time for me to double down. So Sequoia came back and Y Combinator came back. You see that? They both came back. But then, the check Only had 8 percent returns, right? And then you go back here. So see, D Tech Global came in (26:00) 2018 and 9,000, then D doubled down again, right?

(26:03) For $14,000, right? But this check didn't make any money. And this check lost 14%.

(26:10) So you could say that Y Combinator and Kanan like cross like Mr. Boat. they saw the company and they made a decision. Obviously, they're not dumb people. They're smart people. They saw the risk, they made a calculated risk and they said, we don't want to do it.

(26:21) But of course, fast forward 10 years now, I'm like, oh, shucks. I should have, made that double down investment, right?