Southeast Asia Series B Valley of Death & VC Bet Market Timing - E533

Jeremy Au discussed three key insights on Southeast Asia venture capital. First, he highlighted the "Series B Valley of Death," where local funds support pre-seed to Series A, and global investors focus on later stages, but Series B and C funding gaps force startups into aggressive cost-cutting or revenue expansion. Second, he emphasized the importance of timing in venture bets, arguing that startups must be positioned in the right economic window—neither too futuristic nor too late—illustrating with examples like AI-driven call centers in the Philippines and Vietnam’s emerging luxury goods market. Third, he discussed how VCs balance two roles—judging founder potential and adding value—to maximize the likelihood of building unicorns - where startups adapt and learn quickly using the Lean Startup and OODA loops, with examples from sectors like HR tech, AI call centers, and consumer goods.

(01:06) Jeremy Au: What are Southeast Asia VCs looking at and what does Jeremy think is a good investment play?

(01:11) So let me just list it out in no particular order. First of all, there is understood in Southeast Asia to be a Series B Valley of Death right now. So there's a lot of capital by local VC funds that can do pre seed, seed in somewhat Series A, but Series B, there is a big shortage. There are American funds that are willing to do series D, series E, for example, but series B, series C is currently considered the value of death.

(01:34) So because these companies raise a lot of money very quickly from, and then after that they kind of hit this cliff where basically there isn't enough capital. And so they have to cut costs or increase revenues quite rapidly. that was not helped by the post COVID interest rate hike that also shoots off a lot of funding.

(01:50) So that's one. Two is, yes, a lot VCs are now looking at tech enabled plays rather than pure tech. And the key thing is it goes back to the time machine You know, (02:00) VCs can write a deal for a company today, and in 10 years, they become a IPO. But if you time travel 30 years from the future, I have this example.

(02:10) So imagine now you're like, I'm in America. I think America is really into robotics. I'm sure that Indonesia would really be a great place to install robots, right? And then you go back in time and you're like, wait a moment, the Indonesia cost of labor is so low as an alternative to robots.

(02:29) Whereas in America, the average cost of labor is about GDP per capita, right, per person, right? So there's a clear economic incentive to do so. Or maybe for the Chinese factories, they're saying, we have a certain amount of volume that we have to do, but we have a shrinking population for this working class, so unless I want to lose business, I better replace them with robots, right?

(02:47) So I'm just saying, even though China's at 15, 000 GDP per capita, they have a certain demand commitment they have to make, so they have to use robotics to increase productivity, right? If you time travel too far from the future, you're like, I'm going to Indonesia to build an advanced (03:00) humanoid robotics person.

(03:01) I'll be like, good luck to you, right? Because you're literally fighting like kind of like the price is too low. so what I'm trying to say here is that if you time travel 30 years back into the past, and then the VC makes an investment for 10 years, you're still too early, your company dies, right?

(03:15) It doesn't make enough money, right? So that's kind of a sweet spot, like you kind of like want to travel maybe like 12 years. From the future till today and they build a company and so you have to bet the right company. So I think what a lot of VCs have realized is that if you're too tech oriented that's way too far away in the future from where the reality and the gravity of the economy actually is, then you're just never being able, you know, the company would take 20 years or 30 years to get there rather than 10 years.

(03:41) So the trick that you have is like, how do I find the right idea that if I time travel machine 10 years in the future will be just the right timing. So tech enabled place feels a bit closer to reality. so now obviously the companies were founded almost 10 years ago, be the food delivery apps.

(03:53) They're still fighting it out. Obviously we have the data center companies. If you set up a data center company today, you're probably a bit by five years. I think (04:00) education tech, there was a big wave that was being built about three to four years ago because of COVID.

(04:03) So I think it's a bit late for education tech right now. So it depends on that piece, like what's the right play right now. So if I was a pre seed startup, I think a lot of them are looking at AI oriented startups. I mentioned a previous startup that basically buys call centers and replaces them with AI people.

(04:21) So call centers, there's a lot of call centers in the Philippines, right? The BPOs, et cetera. So there's a bunch of funding that's going there right now. I was talking to my friend recently and I was like, now's a good time to build a jewelry startup, you know because if you can build a Tiffany's for Vietnam, and if they can continue growing 7 percent to 8%, the GDP per capita is about same as Indonesia right now.

(04:40) If you grow 7 percent to 8%, you double your economy every 10 years. So I'm just saying, like, if Vietnam can maintain this growth rate for 20 years. They can grow 15, 000 to 30, 000, and 30, 000 technically can go to 60, 000, right? I'm just saying, obviously, it's a lot to believe. I don't know whether they can keep it up for the entire time. But I'm just saying like, yeah, now's a good time to build a jewelry startup, right?

(04:58) You know, because right now people are buying bags (05:00) and cosmetics in Vietnam, but people are not really buying a lot of jewelry, for example. So it was my friend and this Vietnamese lady and I were just brainstorming like, what's a good startup now to build? Because there are makeup startups that are doing well in Vietnam right now.

(05:13) There are starting to be bag start ups that are doing well, or fashion, affordable luxury start ups that are starting to do well. So we're just saying like, hey, the next one is gonna be jewellery, right? You know so that could be an example of the right time. but I don't know. I mean, if I knew for sure, then I'll be making a lot of bank really, really fast, right?

(05:30) so it's hard to tell. But I'm just saying the reason why is that tech enabled services, tech enabled start ups seem basically, If you're too late behind the curve, you won't make any money. But if you're too ahead of the economy of where it actually is, it's too much technology for the country, then, you won't exist as well.

(05:50) Yeah. Well, first of all, gartner doesn't do really independent analysis of the hype curve, right? So they talk to us and they look at the mentions. So they are actually a lagging (06:00) indicator of us. So we make money a bit like traders, right? By the time Gartner says that it's on the rise, it's probably the peak, you know, Like, it goes nowhere for a while.

(06:09) Similar to vapes were invented for many years. Then eventually there was a small hype cycle where it became hot. And then after that it dipped down because everybody hates vapes. And then suddenly now it's, but I think it's unstoppable. I think it'll become 100 percent market share.

(06:21) You know, if you think about it, you're like, is this a slow march, like, in 30 years time everyone's gonna vape instead of cigarettes, right? The cycle's like the hype cycle. So crypto, you know, the trick there for crypto is always like, sell before you get out of here.

(06:33) But when Gartner says, oh, we'll, we'll still have room to it's like Jim Cramer, they say, right? It's like, Jim Cramer is like the joke about Cramer is like, if he says buy, by the time he then it's time to sell, because this guy is like the last to know in the room, right? Gartner says it's hot, then it's probably like, okay, it's time to sell on the VC side.

(06:49) Unless you have a very long holds horizon that can fully deploy over 10 years, yeah, that's it. Yeah.

(06:56) So the question is, will you become a unicorn in 10 years? So I think a lot of VCs will say, okay, you know, (07:00) like, when I meet you, it's like, hmm, you know, professional, put together, can communicate well, all this stuff. You know, high integrity, blah, blah, blah.

(07:07) But the question I'm saying are you gonna become a unicorn in 10 years time, right? So this is like the time travel machine. This is where all the gamblers all come out all of a sudden, right? Because you're like, man, like, you know, how good am I at telling this?

(07:18) Another way of looking at this is, let's say you know Symantec has a company right now that has about a million dollars of revenue today, right? So the smaller version bet, do I believe that you can double the revenue by next year, two million dollars, and then can you double it again, you can make it to 4 million?

(07:34) I mean, some people are even more aggressive. Some people will say triple, right? So the jazz thing will be, can you make 1 million into 3 million by next year? And then can you make 9 million the following year? So, I normally use both of these metrics.

(07:46) The 10 year is like a different side of me. The short term is a little bit more. Logical, but a little bit more, you know, Excel oriented, right? So more business oriented. But I think you've got to keep those two things, because there are some companies that can do well in the short term, but cannot do well in the long term.

(07:58) And there's a lot of companies (08:00) that will take a long time and make no money for many years, but then they'll make it, right? So OpenAI had zero revenue for many years. if you are like, this is gonna make money, you'll be like, no, you make zero revenue next year. the founder will tell you, for the next five years, we'll make no money.

(08:15) And then at the end of it, we're gonna make billion dollars. Right? Then you'll be like, wow, this guy's a joke. Right? But it turns out he did it right? He spent many years in r and d and then eventually now he's out, and now he has $150 billion valuation, right? As of this week with Apple coming in, right?

(08:29) So I think we just have to be aware about that. When we talk about this, we need to understand what makes startups different from that. So we have something called Kaizen. So this is a good metaphor for looking at it. But Kaizen is basically, in Chinese it would be Kaishan, which is, you know, Japanese Kaizen which is the art of lean manufacturing, right?

(08:47) So it means change for the better. And it basically says you know, historically, automobiles were made kind of like by batch. So imagine all these cars, right, you know what I mean? Imagine this car is being made here, and then the whole car is just (09:00) made, everybody brings the materials to that car, and assembles the car on that spot, right?

(09:04) Now, obviously, Henry Ford made the assembly line, so that there was a line that was going by. And then Kaizen, which is, I think, most popularized by Toyota. So a lot of people like to drive Toyota cars because they're very reliable. But what they said was they created a bunch of practices that basically said, you know, first of all, we don't want to have a lot of inventory stock.

(09:22) We want to order it just in time to become more efficient. We allow every worker to stop the assembly line when they see an error in the automobile because it's gonna be less costly for us to stop the line and fix the problem immediately than it is to do a massive recall later. And so, right now, Boeing has, what, like, 70 Complaints by whistleblowers because everybody knew there manufacturing defects, and then they were getting overruled over and over again, and now they have class action lawsuits, they're being stopped, there's a moratorium on, it's a big happening right now, it's very expensive I was talking to this guy, he's an executive, and he's Super pissed because they bought a lot of Boeing.

(09:58) Imagine all your Boeing planes have to stop for (10:00) two weeks You're at the airline, you don't make that much money anyway. So if you stop your planes flying for two weeks, you basically lost all your profit for the year, right?

(10:08) And so it's super expensive and you're super pissed. But basically, this is, Kaizen. So it's like a new way of thinking for manufacturing. So a lot of start ups will call something called the Lean Startup, and this was obviously very hot 10 years ago. Now, I think it's become relatively mainstream, but may not necessarily be aware by everybody.

(10:25) But basically, Lean Startup Method, and it's a book that we recommend basically says that the most important thing that a start up needs to do is they need to learn faster than everybody else. This is actually quite similar to another. Framework called the OODA loop. O O D A, so it's Observe, Orient, Decide, Act.

(10:42) So it was basically saying that if I'm in a fighter jet pilot versus another fighter jet pilot, if he can observe, orient, decide, and act on this plane faster than he can turn that loop faster than me, I will lose to that pilot. So my job is how do I observe, know where he is, How do I (11:00) figure out what my position is versus that person?

(11:01) How do I make a decision about good or bad? And then make an action, I need to do that faster than the guy. So if I can do that loop faster than the other guy, my plane wins the dogfight. So that's the OODA loop or the Boyd's loop. So startup method is actually quite similar. Basically what he's saying is, if I can learn faster than your company, then I will eventually win.

(11:21) Basically, if I as a startup, if I can build something, and then I can, with a minimum viable product, something that gets me the learning, then I can measure, I can get those insights, you know, and then I get this data, and then as a result I'm able to learn and make a decision about what to do, then I can build a better version of that product.

(11:37) So it's kind of like a Toyota process, in the sense of being that Fast, lean machine, but don't get overblown. You don't build an annual strategic plan. You don't build an annual budget. You're just moving faster and faster and faster. So recently Paul Graham released this article. It became a hot meme It's like founder mode versus manager mode.

(11:56) And then Twitter blew up. And it's like, yeah, managers, all these (12:00) MBAs from Harvard, they're totally shit, they destroy a company, they're so slow. Founder mode is get things fast, don't care, get it done and then just go for it. So this is kind of like the learning loop that we have to go faster and faster.

Marker

(12:11) Jeremy Au: We'll definitely talk about failure patterns and how VCs can do value addition at each stage. I think that in general VCs will think about it in two ways. Judging versus adding value. So I think those would be two ways, right? So judging means I think this company can scale without me or with very little of my help.

(12:28) And so my job is to get a ticket on this. road. Doesn't make sense. And then the second side will be like, they can't get there, but I can help them get there. Right. And obviously it's a blend of both. so judging is just judging. Like, is the founder good? the founder Have a spirit of learning.

(12:45) Are they able to attract good people? Are they able to delegate? You know what I mean? Those are things that are quite key. And obviously, if they've done it before, I mean, the best case scenario is, you know, if Sam Altman exits OpenAI at a trillion dollars, and he leaves.

(12:57) He says, Hey, I believe I can create (13:00) a startup that I don't know, going to mass already being done by Elon Musk. But if Elon Musk was like, Hey, I want to be a new startup, but now it's going to go to another solar system, then, like everyone's gonna be like, yeah, no brainer.

(13:10) I can trust that. Now, obviously there's also a selection piece. Adding value piece is also important as well. VCs can really help so some VCs are very good at healthcare tech, some VCs are corporate VCs, some people, there's lots of ways that VCs can add value as well.

(13:23) So that's how you do it. Either judge it that they can already scale or you help them scale. Okay. So I think what startups often do as well is that they often are very much in a mood to do divergence and convergence. Basically what they do is they have a trigger point, then they say there's a problem, then diverge.

(13:39) I want to get as much opinion, let's brainstorm, let's get lots of different opinions, let me talk to a lot of people. And then after that, they empathize with the user, or the problem. Then they find the problem and they start converging on it. So they say, okay, I now understand the right problem.

(13:51) I understand how the problem should be formulated. The problem is not that America wants better quality rockets. The (14:00) problem is that America wants cheaper rockets.

(14:03) So they're going for bigger, larger, more expensive rockets, right? But basically Elon Musk said, I define the problem as the Cost to bring one kilogram into space, which requires us to have cheaper rockets. Expensive rockets, but they're reusable. So we define the problem differently, right?

(14:17) Does it make sense? lot of the Boeing and Lockheed Martin, they were being paid for on a cost plus model, right? So they're reimbursed. Like our favorite defense and government contracts or utility contracts. So as they were saying, Boeing was like, if I make a contract for a 500 million space ship, the government will pay me 500, 600 million basically, right?

(14:36) Cost plus 20%, right? Therefore they'll incentivize to build bigger rockets or spaceships. Versus SpaceX was like, I'm incentivized. to think about it from a payload perspective, because I'm trying to go for commercial payers and so forth. And of course, you have a problem statement that you diverge again, and then you come up with the right solution, and you ideate, you prototype, you test, and then you converge.

(14:56) So the rate of learning is really important. So for example (15:00) one of the big differences that SpaceX is versus other people, is that a lot of the Manufacturers, they have dispersed their manufacturing across as many states as possible. If you're making the fuel line in one place, and then you're making the pumps in another state and everything, you divide it all up, and if your funding comes from the government, then it's very hard for Congress or the state senators. to vote against you, to come together to lower your funding, right?

(15:24) Because your funding is coming from the Pentagon. And all of that was being done. But for SpaceX, they put all the manufacturing in Los Angeles, right? And the reason they put it in Los Angeles was because they said, Hey, first of all, there's a lot of good engineers that were formerly from the Army, the Navy, because of the California base.

(15:41) There's a lot of high tech people, engineers. But we're gonna have everybody be in the same place. There's no such thing as work from home. Everybody's gonna be in the same place because We need everybody to be learning how to make things that have never been made before, right? And so we want everybody in the same place and we're not gonna disperse it, right?

(15:57) So the rate of learning was faster because everybody was in the same place (16:00) and then they could work together, and it wasn't like, computer engineers sending a computer design to the product guys send it to the manufacturing. It's a bit similar to Toyota in the sense that everybody's all in the same place, right?

(16:10) In Toyota, something goes wrong on the exemplary line, everybody is expected to be there. The line manager is supposed to be there, the engineer is supposed to be there, everybody is supposed to be there, and everybody is supposed to solve the problem. At the same time, right? So there's that dynamic that you see there.

(16:24) So as a result, now we're looking at it, is that a lot of this will become a static canvas. It's called a lean canvas. And so basically it's a way of saying a business plan, right? But it's a little bit more lightweight because a business plan is normally oriented to make you sound good.

(16:38) If you write a memo, people will try to make it sound logical, in that sense. So a lean canvas is basically saying, if I look at a startup, and this is maybe the areas they want to cover, deal memos. What is the problem they're going after? What is the solution they're going after? What are the unique value propositions?

(16:53) What are they promising as a customer? What is the unfair advantage? , And then who are the custom segments? Who are the personas of the people they're (17:00) going after? Who are the early adopters in this persona? And then as a result, what are the key metrics they're delivering?

(17:05) What are the channels? How do you sell them? And what's the cost structure, the variable cost, the fixed cost? And what are the revenue streams? How do we make money from this? And to some extent, it's defined by the product, and it's defined by the market in a sense.

(17:17) An example, I'll just talk out loud, but for example, we are looking at. A company like Rippling. So Rippling, like Zenefits is an all in one HR system in the U. S. There are similar companies in Southeast Asia. So OmniHR, BrioHR there's a bunch of others. But this is so the definition could simple one would be the customer segment the problem.

(17:39) Okay, so the unified proposition is we are an all in one HR that handles payroll. Leave, tax, expenses, recruitment. So it's all in one software, right? And we are targeting HR officers, but primarily HR officers from smaller companies or fast moving companies rather than large companies. But it can't be too (18:00) small.

(18:00) They can't be 10 employees because that way you can do it yourself, but it needs a certain size. So these are segments we're going after. And as a result, the problems we're going after is they have a lot of fragmented systems. They don't talk to each other. It's unproductive. They do a lot of manual work, and therefore our solution, the top three features is all in one.

(18:14) Number two is digitized. Number three is interoperable with other extensions and automations, for example. And as a result, the key channels that we sell is we do a sales force, we do marketing. The key metrics we measure is productivity, outputs and so forth. Cost structure is software as a service.

(18:32) It's a classic AWS engineering cost and set up, but it's primarily fixed cost, not very little variable cost because it's digital software. And the revenue stream is we charge it per seat. So for every employee that's in your company, we charge three dollars, Singapore dollars. And then if you want to add another module, it's another dollar and so forth, right?

(18:50) So as a Singapore company have 100 employees and you're using 3 plus 1 equals 5 per user, the modules. (19:00) Therefore, 100 employees times 5 equals 500 per month, times 12 months is 6, 000. And then, if you install it all in one HR system, you'll probably never uninstall it, right?

(19:11) For the next 10 years at least. So 6, 000 of money that you make for one year times 10 years equals to 60, 000 of customer lifetime value. So this is a way for you to think about the lean canvas, a snapshot of that business plan. And what you normally notice is that for startups, what they're very focused on all the time, and it's quite different from a lot of other companies, is that they're focused on the number one problem.

(19:31) So they are focused on the number one problem for this person. There's always a new number one problem that emerges in anything, right? So I always tell people like, okay, if I look at Rilling and I say, Rilling, do you want an ambulance now? No, right? But if I cut off your hand now, then I'll say, do you want an ambulance?

(19:48) You'll be a yes, right? Because that'll be number one problem. The number two problem will be like sending Jeremy to jail, all right? After that, but the number one problem now is going to ambulance. You need to think about startups that are always laser focused. They don't care about the number two problem, the number three (20:00) problem, but they very much care what is the number one problem? Because if it's not your number one problem, you just delay your buying decision to later, right? Number one problem is really important and really important for you to understand who is having that number one problem? So you can't say oh, large companies need All in one HR.

(20:17) No. Who is it? Is it the CEO of a startup? Is it the HR officer of a medium sized startup? You get what I mean? Who exactly is having that number one problem? When are they having that problem? It's very important. Because it's, sometimes I was working with a technology company and they're doing cybersecurity, right?

(20:34) Basically what they said is, cybersecurity is the number one problem when, anybody wants to give a guess? When is cybersecurity the number one problem for you as a company? When there's a breach, exactly! Exactly! When there's a breach is exactly the number one moment when people are like, oh, time to buy cybersecurity, right?

(20:52) So that's the number one problem at that point. And that's reality, right? So there's a tricky part, which is that if someone has a data breach, it's probably easier to sell them a product at (21:00) that point in time, which is I'm doing data correction and recovery, for example, and then I sell you the protection stuff later, right?

(21:06) Does it make sense? But, so you have to think about when it happens, and you think about why it happens. So it's the number one problem, right? If you talk about, we joked about getting an A for math. If primary 1 to primary 3, the parents are pretty chill, they don't care, they probably won't buy this.

(21:19) But primary 4 starts to heat up, primary 5, primary 6, there's PSLE, and people start to heat up, right? And then people are like, okay, I want to spend more on education because of the PSLE. But then, if they're going to an integrated program, they don't need to go through this primary six exam, then parents don't care either.

(21:34) Then they spend more chill, right? Sorry, integrated programs for secondary school to junior college. Thank you! The areas look at me, it's what? And I was like, ah, shit, I got it wrong. Anyway, the point is that secondary school to JC. But then, from JC, you probably want to go to uni, you probably need better grades.

(21:46) I think you need to understand why people are having that as well. When you have cancer, the day before you got your cancer diagnosis, your number one problem was work, followed by my spouse is angry at me for not cleaning the dishes, followed by I (22:00) don't have time to hang out with my buddies.

(22:01) That's your number one, number two, number three problem. And then you have cancer, that's your number one problem, right? Everything else is and then everybody else will tell you, okay, your boss will tell you. Don't worry about work, go fight your cancer. Your wife say, okay, postpone this fight for now.

(22:15) Let's fix the cancer first, we deal with it later. And then your kids, your friends and buddies will be like, okay, I'm here to support you. Your number one problem is cancer, right? So your number one problem is cancer, then probably you want immunotherapy, you want any cancer, you want to see a doctor, you'll pay a lot of money to have the right solution, right?

(22:29) So that is a product that you want to have. If you're poor, for example and you are a business owner, And then you're facing poor, then you may want working capital to expand your small business as a company, right? For example, in Indonesia, in Vietnam, it's very hard to access. Capital loans, right?

(22:44) Because the banks are very slow, they're very high requirements. They require personal guarantees. It's a quite hard to access capital, right? And so startups that are able to spot the right small businesses there that the larger companies don't want to underwrite by actually are good bets because they're hardworking, there's a good revenue (23:00) growth, then they're able to land and they make money on that capital as well.

(23:03) So you always have to think about that piece as well. And obviously, you wake up in the morning, you look at yourself, and then you're like, oh shit, I got 20 wrinkles in my fucking face. And then you're like, okay, you know what, time to go to YouTube, look at longevity, aging, lasers, inside out, rejuvenation.