Why Startups Fail: 6 Disaster Patterns and How To Succeed by Harvard MBA Professor Tom Eisenmann - E397

· Failure,Podcast Episodes English,Singapore,VC and Angels,Start-up


“I'm talking about this for the founders and operators, and people who want to be founders. Being very thoughtful about why executive failure is happening, when you open up the newspaper and you see something is failing, you have to think that that is one point of view. What is the deeper point of view? What is the insider's point of view? What is the point of view of the executive team with the benefit of hindsight, experience, and the knowledge of the counterfactual reality? What are those lessons that we have?” - Jeremy Au

“The truth is a lot of the success has been handicapped because failure is happening because of avoidable mistakes. Avoidable mistakes happen because we don't want to look at failure directly in the eye. This is where we enter that feedback bubble that founders can have because founders often face disapproval from family and friends who ask them why they’re doing this crazy thing. They get a lot of criticism that’s not well-founded. On the other side, they can also have people who are in their camp supporting them. There are very few people or advisors around the founders who are supportive of them. They want the company to succeed. They are spending the time, and they have the experience needed to fully understand the business. If founders don’t have these, then it kind of falls apart.” - Jeremy Au

“When we describe startup failure, we're looking at this from the angle of being thoughtful about all the different parts, that it is due to multiple factors and not due to a single factor. We have to be clear about what to attribute it to, which is both the environment and the individual actions of the team.” - Jeremy Au

Jeremy Au discussed how founders, VCs and executives need to understand predictable failure patterns in order to avoid disaster and succeed, based on Harvard MBA Professor Tom Eisenmann's research and book. He highlighted how the odds are worse than what people think with 90+% of startups failing, with only 4% of companies achieving 10x to 50x+ returns on capital and sweat equity. He goes over why it's difficult to evaluate failure objectively due to single cause fallacy, fundamental attribution error, and time-lag in understanding the truth. He explains the six types of startup failures with multiple case studies: 1. "Good Idea, Bad Bedfellows" (being killed by teaming issues, Quincy $1M) 2. "False Start" (misjudged product-market fit, Triangulate $2M) 3. "False Positives" (early enthusiasm doesn't translate to broad market appeal, Baroo $4M) 4. "Speed Trap" (rapid growth outpacing company building, Fab.com $336M) 5. "Help Wanted" (vertical downdraft or strategic missteps in a growing company, biotech 1990s and Dot & Bo $19M) and 6. "Cascading Miracles" (multiple high-risk logic and milestones chain, Iridium, Segway, Webvan vs. FedEx, SpaceX and Tesla).

Join the Singapore Growth & AI Summit with Sean Ellis!

BRAVE has partnered with Causality and Sean Ellis to bring you discounted tickets for this Singapore stop on Sean Ellis’ World Tour. Sean’s book, Hacking Growth, has sold over 750,000 copies and he’s sharing the latest insights on enabling breakout growth. Use code seanbrave7 today.

(01:40) Jeremy Au:

Hey, everyone! I wanted to take the opportunity to review a book that I really enjoyed. The book is called "Why Startups Fail" by Professor Tom Eisenmann. He's a great professor at Harvard MBA, and it was a really interesting experience to learn from him because he was doing research on this topic that nobody really wants to talk about, which is about startup failure.

The interesting thing about startup failure is that people really talk about the topic rather than talking specifically about startup failure. What do I mean by that? Well, obviously we all know the heroic images that we see on magazines, right? So founders who seem to have a perfect business, everything's firing all cylinders, and if we do kind of go deeper into it, we often hear the stories about the almost failures, right, about how they went through some really tough times, but how they were able to pull a rabbit out of the hat and they were able to avert disaster, and as a result, their eventual success looks even more amazing.

We also talk about the other side of the coin, which is we talk about the catastrophic failure. So we often see fraud like Theranos, or for example, we see companies that were blitzscaling like WeWork. And then we talk about their failures, either ranging from conflict of interest to fiduciary duty, to flat out that they were not smart and they were not doing well as a result. So we're talking about the failure of these very large companies.

What Professor Tom Eisenmann did, which I found interesting, was that he really went to talk about why startup failures happen. What are the clusters or reasons why it happens? Because from his perspective, he was looking at a whole number of Harvard MBA students who were supposed to be successful, and from his perspective, was that these people who honestly had everything going for them in terms of perhaps in terms of intellect, in terms of social capital, in terms of ability to raise their initial capital, they were able to succeed initially, but then he started seeing these patterns for why they started to fail.

And so from his perspective, as somebody who has been a professor in entrepreneurship, who has been very indexed on the success factors or the almost success factors, he wanted to see the other side of it, which is why do people fail? It's an interesting study for me personally. I think it's an interesting approach because when we think about solving any problem or improving any system, obviously we think about the best case scenarios. And so, we want to talk about a system that provides success. But often very much we can improve the system by improving the ways that it falls apart, because as we all know is that, the best part about Oreos is that it's always the same Oreo every single time, and if I open up a packet of Oreos, it doesn't go bad. There's no cockroach inside. There's no catastrophic failure on the Oreo's side. So for me, a successful Oreo is an Oreo that comes out every time and delivers the same experience every single time.

In other words, if you have a great hotel and it gives the best experience to a few folks out of many folks, that's amazing, but if many people are going through a failure state through the hotel experience, then guess what? You're going to get many one-star reviews and a few five-star reviews and that is not a very good hotel. What I appreciate about Professor Tom Eisenmann sharing as a result is that he's really looking at something called the vocation, the profession, and the professionalization of entrepreneurship, not just in terms of the mechanics, the patterns of just the founding, but also the growing and the scaling of the business. And I think these are really good lessons, not just for myself, but hopefully for someone else out there who's thinking through the process, not just as perhaps a VC, but also as a founder looking to build your own company, or as an operator who's working hand-in-hand with the executive team to avoid these failure patterns.

So how big is a problem of startup failure? Well, there's a lot of numbers about the word startup failure and the truth of the matter, and we've discussed this in previous episodes, is that failure is a default case. What I mean by that is that startup success is really saying that we have built a company that people want a product for, and it is a new product. The consumers don't really necessarily know that they want today, but is a harbinger, is a potent product. So the question is how common is startup failure? How big is this problem? Well, if you open up and look at Forbes 30 under 30 and so forth and Fortune, well, the truth of the matter is that it looks like it's 100% success rate because all the success stories, the survivorship bias is that only the successful founders, the ones who managed to pull a rabbit out of the hat all the time are the ones who make it onto the press. But of course, when you look at, you know, The Straits Times or Channel News Asia or Wall Street Journal, New York Times, then it almost feels like, okay, you know, it's all doom and gloom, right? Because it's all fraud, it's all failure, it's all explosions. And then there's a few companies who no longer really seem to be startups or seem to be doing well.

(06:01) Jeremy Au:

So let's define failure. Professor Tom Eisenmann defines failure as when early investors put money into the company, fail to get the money back and fail to get a return of the company. Now, obviously this seems to be a bit technical from people's perspective. What if you build a startup and it's chugging along and it takes 20 or 30 years, yeah, well, there's some level of return that's there, and so it's not necessarily a failure, but it's not necessarily a success, but definitely we've seen other companies where they've burned a lot of capital, hundreds of millions of dollars of capital, and then they go public at tens of millions of dollars or a hundred million dollars valuation. So as a result, the early investors made no money, even though the company became viable. So taking the lens of an early investor to put that financing in is also, I think a good way to benchmark for the founder's perspective, because the founder is actually the earliest investor in a startup because they're putting in their own time and sweat and honestly, a ton of pain. And so they are putting in sweat equity. Which is I would say way harder than the financial equity of the early investors will put in, but they're very much similar because the eventual financial reward for all of that early sweat equity or seed capital, that return has to be taught and managed, and honestly if it's a high return, then it was worth it.

You work your ass off for 10 years, 20 years, and you got there, or there wasn't a return, in which case it wasn't worth it. So I think this way of pegging to the early economic returns for the early investors is really important because it gives us a decent definition. That means obviously that if you are a startup that was a smart bet, but it never paid off, well, it will still be considered a failure because even though you executed well, you kind of knew what's coming, you knew what the future was, and you built it 20 years early, in the context of the venture, in the context of the individual founder who had built that company, it would be still a failure.

A good example would be Jibo. Jibo was the world's first social robot. It was a beautiful tool that would be by bedside and it could interact with you. It could chat with you, could talk about the weather. It was a very nice and conversational experience that people were supposed to not just find them as a helpful assistant, but also be someone that could become attached to. This company was founded in 2013. It raised 73 million USD, and it closed down in 2018. During this process, they had a beautiful minimum viable product. People love the concept and the prototypes, but what they learned was that during the engineering process was that the hardware, as you imagine, and the middleware that was associated with it, would be much more expensive than what they had thought it was going to be. And so the robot ended up being more expensive than what they thought they could provide a product for.

For example, the light sensors they had were primarily industrial, commercial usage, but in the context of the home, these sensors were not good enough to detect what people were interacting with and the requirements of the social interaction. So they had to create new middleware, but it also had to be thoughtful about the hardware that needed to be actually there. And that increased the cost. And then, as you can imagine, if it's expensive for a consumer social robot, then consumers are just not going to buy because they're price sensitive.

Other things happened. The CEO was diagnosed with leukemia. The CTO, who was a co-founder had to step in as the interim CEO. The CEO eventually got better and then came back, but during this timeframe, the Amazon Echo came out and no one expected it at that time because at that time, Amazon was primarily, as you can imagine, shopping and commerce company. And they had to suddenly come up with this Amazon Echo and they're pricing it very cheaply at $100, $200, so much cheaper than the Jibo robot. And it was just voice. There was no sensor in terms of the display. There was no light sensor, so they were very focused on just the voice assistant of it.

And as you imagine at the time, for those who remember, during the timeframe, that also triggered the launch of all of the various voice assistants, as you imagine. So Siri came back, Google Home came up, and the whole smart home ecosystem really was birthed in that timeframe. And so it's interesting to see so many homes now have these smart voice assistants today. And at the time, the truth was Jibo saw that curve happening and they inspired because they were started in 2013, years before people were using the social robots. And so in many ways, Jibo was the inspiration for me out of this and they were ahead of the curve. But at the end of the day, there was still a failure because the $73 million ended up into a $0 return for the early investors.

And now we know that these social robots at home have been supercharged by the invention of ChatGPT in 2022, 2023, and now we're starting to see the further, you know, spread of this idea, which just goes to show that Jibo was a failure from an economic perspective, but very much was a pioneer and honestly saw the future years and years in advance.

So the reason why we're talking about this as a result is talking about how common this economic failure is. And this definition of this economic failure is relatively straightforward. So the truth of the matter is how common is this? Correlation Ventures analyzed 22000 financing. So this is the set of all financings that were done by VCs. That means if you didn't get VC, you're not part of this data set, which is important as we come into it. But the long story short of it was that only 4% of the companies were able to generate 10%. The truth of the matter was that only 4% of the companies were able to achieve 10x returns to 50x or higher returns. So that meant that about 96% were really not great outcomes. Bringing that down further, 65 percent generated zero to one X. So that means that if you put money into the company, you lost money. You lost the sum. About 25% achieve one to five X. And about 5% to 6% achieve 5 to 10X so this is a good way to think about that spread, which is that 65% just lost all the capital. And then there wasn't great returns of capital for another 25% effectively. And then only about 4 to 5% percent really generated that outcome.

(11:37) Jeremy Au:

Again, what I'm trying to say here is that this is the set of all known financings. And what that means at some level is that there is obviously a huge number of startups that honestly were not captured in this data set. They never receive any financing or they receive financing that was not an institutional round financing. And so the death rate is not 65%, I would say, but closer to 95% if you really think about it, if you include all the various folks are out saying, I want to be a founder and so so forth. What I mean by that is that in general, the highest death rate is the pre-seed quantum into the institutional seed round. So that's one gate that's a huge death rate. And then there's another death rate between the seed to the Series A. And then after that, the Series A to Series B, then supposedly from the Series B to your eventual exit, the death rate is much smaller, but some way I often think about it is that, there's effectively an 80% death rate, I would say between pre-seed, which is, idea pre-product-market fit to the seed stage, and at the seed stage, there's about effectively a 50% flat out death rate from the seed to the Series A.

And I think about it again as another 50% death rate between the series A to series B. And then I think the death rate kind of like drops a little bit from the series B onwards round by round, but there's still a significant amount of risk. And again, I'm saying this because it's a heuristic, but also it varies, right? Some of this data I just gave you was for the US market. For emerging markets, it's way tougher. It's way harder. A lot of the death rates are clustered early stage because financing is tough in the early stage, but also the quality of founders and the quality of the business opportunities are so more limited in emerging markets. And then, there's obviously a death valley that we know for growth stage capital for emerging markets because there isn't a large amount of growth capital. So again, take it as it is. These are the heuristics I try to think about in terms of how I think about the numbers and obviously that varies, again, by institutional funds versus non-institutional, pre-product-market fit versus with product-market fit, US versus other emerging markets. And within emerging markets, as you imagine, there's a big difference between Singapore, Vietnam, Philippines, Indonesia, Malaysia, Cambodia, Thailand. There's a whole set of conversations about the individual valleys of death and the death rates and the definitions of the failure rates as a result in each of these countries.

But what I'm trying to say here is that it's high. At minimum, if you're like a glass half full kind of person, your 65% death rate in America once you've achieved an institutional round of capital, which is, if you're high, if you think about it and honestly, it's probably closer to 90, 95%. So you just have to understand that this is a high risk venture. And as a result, we kind of recognize at some deep level that being a founder is very tough. And we honestly lionize and respect people who build businesses because we know it's tough. And of course, I think the tricky part is when it looks easy or people feel like it's been done using a shortcut, but I think at a deep level, the reason why we see founders as heroes is because we understand at some level that's hard, but here, I'm trying to do and try to quantify how hard it is.

(14:22) Jeremy Au:

So Professor Tom Eisenmann does a great job talking about the three aspects that makes it hard to analyze failure. It talks about the first, which is about a single cost fallacy. Secondly, is about attribution error. And third, of course, is just the struggle of figuring out the truth.

The first thing that we think about is that startups often fail due to multiple issues. And of course, when I say this, it sounds super obvious because when you and I think about it, when something goes wrong, the challenger shuttle explodes. It's not because of one thing, but because of multiple failures in the safety process, the environment, decision-making, and bad luck. So all of those things happen because multiple things happen. Multiple players made a decision. All of those things chained up together to cause a problem. And when we think about diseases, for example, we also think about it the same way because if we hear that somebody has cancer, then we think about it in terms of multiple different ways. What were the different factors that came? Was it because of inflammation? Because of diet? Because of sleep? Is it because of stress? Is it because of the radiation? There's so many different reasons that happen, but at some level, we know that it's an interplay of multiple factors at once. And so what we can't do, for example, is like, well, if you ate bananas, then you wouldn't have cancer, and everyone's gonna laugh at you because it's not due to a single cause. And so, solving it with a single way is not sufficient for us to truly understand what the situation is. The truth of the matter, though, of course, is that when it comes to startup failure and the newspapers that we have, then we just want to say one thing, which is, the founder sucks, or the environment sucks. So whatever it is, we only talk about one thing. We don't want to talk about the multiple reasons why a company fails. As a result, we really want to oversimplify both the good and the bad outcomes into a single cause.

That second part is the fundamental attribution error, which is that when we observe others, we tend to blame bad luck for what's fundamental about them, the people, the team, the execution. And then we tend to downplay the situational stuff, which is the environment, the luck factors. Of course, when it's the other way around, when we are the ones who do something wrong, then we tend to say, oh, it's the environment, the industry was bad, and that we downplay what we ourselves were responsible for. And of course, it goes all the way around, which is that when things are happening in a positive way, then we say, well, it was all thanks to my individual skills and the environment has nothing to do with it. And when somebody else has something positive, then they're going to talk about how it was their execution and their leadership and nothing to do with the environment. And so as a result, it makes it hard to figure out the truth because everybody's blaming one another. It's also a lot of confidentiality, as you can imagine. The board has a fiduciary duty. People are often covered by nondisclosure agreements, so it takes time for the full truth to come out. It takes time for the business journalists to come in, dig through, and have a point of view. The truth is, we look at geopolitical events like the Vietnam War, or World War II, and even today, historians are looking at that set of data 50 years down the road, 100 years down the road, And we're still arguing about why World War II happened, World War I happened, because there's so many different factors and the question is, how do we weigh all of it?

(17:07) Jeremy Au:

So, what I'm trying to say here is that when we describe startup failure, we're really looking at this from the angle of being thoughtful about all the different parts, that it is due to multiple factors, not due to a single factor, that we have to be clear about what to attribute it to, which is both the environment and the individual actions of the team. And lastly, of course, is that it takes time and effort and judgment to be able to sort through the facts and the wreckage afterwards to figure out why failure happened.

So let's talk about the six types of startup failure that were identified by the professor. The first is good idea, bad fellows. And what that fundamentally means is that it was a decent idea, but the team didn't work out. And there's all kinds of reasons why the team doesn't work out because the truth of the matter is that it's again, hard to build a startup. But the thing is, even if the startup is a good idea, if the team is bad, then bad things happen. Now, bad teams can happen because one, the people are bad. Individually, they are poor performers. The other part that you can see is that you have high performers individually, but there's just not the right configuration or not the right teaming, or honestly, they don't get along. This doesn't just apply to co-founders, but also the early set of employees, the early set of early investors, that configuration on the early stage is really, really key to make happen.

So let's kind of write rather off a few of the errors that honestly, if you're a VC, it's kind of, you know, obvious, but let's just say them so that we're on the same page, right? One is you don't understand the market. So, you have two co-founders and you don't understand the market. And you're learning on a job, which is not a bad thing. But the truth is, if you were experienced in the industry, well, you already know the avoidable mistakes, you will know how to succeed. So experience is an advantage and inexperience is a disadvantage that can be overcome with the right attitude and right humility, but let's acknowledge it that there are some industries that also require much more of a learning curve. So, if you want to do engineering, and you don't have an engineering background, and both of you don't have an engineering background, then it's going to take a lot more work for you to catch up, versus other ideas, for example, consumer social, that could be a little bit easier for you, because you already have some prior experience, and some nuance and localization that you understand.

(19:11) Jeremy Au:

Other areas, as you can imagine, is who's the boss, right? So who's the CEO? Often, we see something called the co-CEO. So who's in charge? Who's making decisions? Who's calling the shots? It's surprising because obviously if you kind of don't really want to work for teams where there are two co-CEOs but we see a lot of that, right? We see that in series B, series C companies in Southeast Asia, and it's not resolved because oftentimes, the first set of founders will come in, they are both type A, they both want to be in charge. And so, you kick the can down the road, you call each other co-CEO. Often, very much a bad sign, because when push comes to shove, when a bad decision has to be made, when a good decision has to be made, neither is made, it ends up that nobody makes a decision, which is far worse, because even with a bad decision that's being made, you can try it. You can do something about it. You can execute it, and then you find out it doesn't work, then you learn from it, and then you make a good decision. Other examples be like a flexibility, like initiative, a poor investor fit.

So for example, we talked about the aspect, maybe your two co-founders who don't have engineering, but you bring in an investor who supposedly has an engineering background, but it turns out they don't, or they don't have time, or they don't have the time and energy to really want to support you on that. So even though technically, they have engineering background, they're not really a true complimentary fit. And so again, what I'm trying to say here is that these early stages, that if you are the CEO or founder of the startup, the question you're going to ask is, would you hire this person ? Would you hire this person to be a CEO? That's how you're going to be thinking about it. Would you want to hire this person to be a Chief Technical Officer? Would you hire this person to be your teammate? Would you hire this investor to be a board member? These are the things that you had to be thinking about because very much, when we are in charge of a single hiring employee, we're very clear about what kind of person that we need to have to fill this requirement.

But the question of course is, do they fill this requirement? And that question should still exist, whether they're a co-founder or investor or board member, because the truth is with an employee, you're paying them 50,000 per year, 100,000 per year, but for a co-founder, you're paying 50% of the company. For an investor, you're paying 20% of the company for exchange for a million dollars. These are the transactions that you're having. And so you should have the same hiring bar that you would have as an employer, as you would for this early team.

A good example of this would be Quincy Apparel, which raised about a million dollars to work on direct consumer work apparel. And these were both Harvard MBA graduates. And they very much identified something, which was they are both women. And they didn't feel like they had a good set of work apparel that fit them. At that time, direct consumer was really hot. And they were founded in 2011, and so, at that time, Bonobos had been around for four years already, and Bonobos was succeeding in direct to consumer menswear more on the casual side. And so, from their perspective is, well, you know, female work apparel, let's make it happen.

And so, very much, unfortunately, they fell victim to some of the traps that we talked about. They had very unclear decision making. It took a long time for them to make decisions. They didn't have experience in the apparel side. And when they brought on people to partner and to help them manufacture, they were unable to avoid the errors that it had. And so, in the first run of clothes, their designs didn't match up with the actual user experience that they had taught they were achieving. And so, when you have a million dollars capital, it seems like a lot of money, but the truth is that runs out pretty quick when you're making fundamental errors.

As a result the company closed and everybody moved on, but this story is not just about Quincy Apparel. And I really respect them for sharing that story because the truth is, we hear that story all the time for so many different startups that we see. And so, for me personally, I've seen hundreds of startups already fail at this stage because the team fundamentally was not a good fit for what they're trying to go after.

The second error is called the false start. And what that means is that you believe that you're product-market fit and then you start building and it turns out that you're building the wrong thing. This is often very much what happens when people are very focused on building a launching, but they're not clear about what a customer needs and requirements are. And so they are very much indexed often on themselves as customers, but it turns out that they themselves are also the builders and so they drink their own Kool Aid and they are building a product that themselves and a few friends want, but not actually what they really, really, really need out there in the market.

So an example for this would be Triangulate, founded in 2009 and they raised $1.5 billion and they did something called data-driven dating. And so from their perspective, it was quite straightforward, which was everyone wants a date. Obviously at that time, there was match.com and all this other stuff that were online classifieds that were basically a replacement of the old newspaper ads, which was like, hey, I'm looking for a date and I'm happy to write a newspaper. I'm happy to pay money for the newspaper ad, became online classifieds, which became match.com, online dating. So from this perspective, the Triangulate founder basically said, you know what, I'm an engineer. What I would really want to do is not just have to classify and look at people in terms of their bio and everything. We wanted to use data because I think that people really want to match deeply and find the one. And data is the best way to do that. We're going to match our hobbies, our interests, our dislikes, our likes and dislikes. And then everyone's kind of like, okay, this sounds a little bit like OkCupid, but basically the idea of using really deep data science to really help two people match together.

So in other words, supposedly a higher quality, higher hit rate matching algorithm dating. Of course, all of us today in the late 2020s know that that's not how the world went. In fact, the world went from, and I remember this time, people were using OkCupid and Match.com. They were writing a lot of profiles and doing quizzes and stuff like that to show the compatibility, but the world moved towards the number one player, which was Tinder, which was very much photo-based. And then, of course, it was very much a few bullets about what people self profiled themselves. And if you look at the generation of dating apps today, like Bumble and Hinge, they're all in contrast to Match.com but the truth of the matter is that actually they all have the same DNA, which is that their perspective is that people really want to have the sensation of matching, and so their job is to create lots of different matches and it's highly visual.

It's a big profile photo. It shows the age, it shows your job, and it shows your height, which, if you think about it is, as we know, not highly predictive of a relationship. If you look at Professor Gottman and what predicts for long term compatibility, it's about your communication style. It's about your sense of shared values. These are the things that the Triangulate founder was trying to go for, but it turns out that when you get a dating app, I'm here to play slots, right? I'm here to get lucky with the right match, the right person. And I want to see lots of options. And, if you are an app, that's a perfect match, but only as one person, it's not right for me, but I want to see hundreds, thousands of profiles and personas. And I want to have the sensation. I want to be picky. I want to be someone who feels desired. I want to be somebody who is feeling desire. And so I want to see lots of options and say yes and no. And the truth is bing, bing, bing. Lots of no's, some yes's, looks like slots, gamification, it's fun, you stay hooked.

And a lot of these dating apps, as we all know, is they make money while you are on the app. In other words, they don't really make money when you get married and stop using the app. So, those subscriptions, those 10, 20 per month, Hinge Plus, Tinder Plus, Tinder Premium, whatever you want to call it. All of that is based on a business model. It's about keeping you on the app, but not necessarily making money when you successfully match. And of course, I think it's gotten so funny. And for me personally, I find it hilarious because a lot of these brands are supposedly competitors to each other, all actually part of the same company. So you have Tinder, you have Hinge, which is supposed to be the anti-Tinder, Plenty of Fish, OkCupid, Match.com. These are all different competitors, but it turns out they're all under the same company called Match Group, which used to be part of IAC, and they all had the same monetization model, which is about keeping you on the app. And we just happened to sub brand private labels into sub communities so that you're able to match within sub pools. But from a broader perspective, there's a large pool. And you're just self selecting into your assorted niches. So again, what that is, is that the founders eager to build and it is basically overbuilt and they blow up the cash.

The third category is called false positives. And what that means is that it is actually something that has been validated across a small initial group, or even a medium sized group of early adopters. In other words, the error is incorrectly extrapolating the early adopter enthusiasm to the mass market. In other words, they're too early for the mass market in some ways, or they haven't adapted the product sufficiently for the mass mainstream. Again, this is also driven by founder eagerness and overconfidence. The big difference between a false start and a false positive, of course, is that a false start is that the founder has been indexing on themselves and they do a very small type, but it's built the wrong thing at a start. Whereas I think for the false positive, I think they very much often receive the seed funding. They receive that initial success, and then they are able to grow that success further. And so as a result, within the venture scale timeline, it becomes a failure, especially when it's not managed well.

So a good example of this is Baroo, which was founded in 2014. It raised $3.6 million and they were basically doing pet services. So the partnering of apartment buildings, as you imagine, which are high density and basically saying, Hey, can we service all of your pet services within this building? And so what was interesting about that experience, as you can imagine, was that they were successful in the early days, they were in Boston and they had several buildings that were doing well. They're profitable, and that gave them the confidence to expand to new markets. Unfortunately or fortunately, was that there were early adopters that bingoed and hit the right early adopters. So for example, they had launched their service. It was during winter time. So obviously when it's cold you don't want to walk the dogs or pet services are really popular, but also they pick the right block, the right demographic. And so everything kind of worked out in this block, but it wasn't replicable across more buildings in Boston, at least not without significant product change, and it wasn't replicable in expansion to a new city as well. And so, as a result for Baru, it eventually failed because they were unable to scale to a new city.

The next category is called a speed trap. And what that means is that it starts out with an early set where there's product market fit. There's an aggressive expansion opportunity. And so the founders and the team and the executive leadership really go for it and they saturate the early original markets and they start expanding to adjacent markets, either geographically or expanding it into different product categories in a way to kind of like kick it all. And this is very much driven by an understanding that for many types of businesses, there's a winner-takes-all all dynamic. There are significant network effects to the business. And so what that means is that for every additional user that we add, we make the product better for the earlier users. And as a result, we have that flywheel supposedly that's spinning very well, but the faster we grow, the faster we'll grow further.

Unfortunately, as you imagine, this can bring two things, which is that you have rapid growth and that attracts rivals, right? And then people start to compete, price competition happens, and margins fall. So we see that all the time. For example, we saw that in Quick Commerce across the world. People felt like it made a lot of sense during pandemic, lots of people crowded in, lots of competition, price war, nobody's making money. So that's one aspect about it. Of course, when you're going through the speed trap, you're also building a company that's very disorganized. It's really not a good culture because people are spending money for growth, but not necessarily focus on profitability. Nobody's really focused on doing what needs to be done.

And as a result, what happens is that a company ends up in a situation where they're effectively stalling. What I mean by that is that the top line could be growing, the GMV can be growing, but the profitability is honestly bad. They're bleeding a lot of money. And this is when, you know, things start to fall apart because when the funding markets say, hey, we're not going to give you more capital because we're worried about the deployment of this capital, when the CEO is unable to raise more capital as a result, you know, whatever the function is, well, the truth is that the CEO tends to slam the brakes and then ends up firing a lot of people. And we see this all the time in Southeast Asia where lots of people give feedback, like, hey, we're growing too fast. And the CEO says, we've got to grow quickly. It was a limited time, expect an opportunity, rivals, first come, first serve. We've got to win, dominate, conquer, expand, land, plant a new flag. All of that stuff happens. And things fall apart, right? And that is where the layoffs happen.

And so we saw that in Fab.com. Fab was founded in 2011 and was doing flash sales e-commerce. And this guy was on top of the mountain, right? He was on all of the various magazine covers. He was ahead of the curve. They raised $336 million and they effectively reach a $1 billion valuation. So there were unicorns and then the whole company imploded because again, people didn't want to provide more capital. There's way too much to believe in order to achieve profitability. It was too much work to achieve profitability once the money shot off and then their core markets was getting saturated. They had to pivot and so forth and everything kind of fell apart. Again, this is a function of the company growing too fast.

The fifth category is called help wanted. And what that basically means is that, in general, things just go wrong. So the company's at a growth stage, the company is scaled. They didn't grow too fast, but something still went wrong. So they were able to sustain product market fit while growing a customer base, but sometimes things can just be, you know, honestly a bit bad luck. And so what we saw is that sometimes economics and economies and macro economy happens. Biotech went through a huge downturn in terms of funding in the 1990s. Cleantech went through a huge downturn in the 2000s. And obviously, we're seeing in 2022, 2023, there's a huge inversion of the zero interest rate policy, which has caused interest rates to go up, but as a result, funding to dry up and be a huge drop, both on a growth stage and even to some extent, the early stage all across the world.

As a result, this down draft, this burst of downward momentum at a macro level, the industry and funding level kill startups because the startup, they often very much have 18 months to two years of timeline for funding and during this vulnerable period, they thought they're on track. They thought they're going to be able to raise money. And then the sector happens everything goes to shit. And then they just get killed because it was this bad timing. And I think we also saw that, for example, the pandemic was a great example of it. There are many companies that were focused on providing in home or personal care services. And the pandemic happened and you basically shut off revenue for everybody because everybody was quarantined. Everybody stayed home.

(32:37) Jeremy Au:

Everybody's fearful for the budget. And so it costs a lot of disruption for a lot of direct consumer services in the pandemic, especially in 2020 and 2021. This can be compounded obviously with the wrong hires, right? So yeah, the wrong management hire, dysfunctional team, executive leadership thing, but basically there's some sense of the company fundamentally sound, but just bad luck. It happens.

The sixth type is called cascading miracles. Cascading Miracles is when you have an incredibly large vision and we've seen this and they raise a lot of capital, hundreds of millions of dollars, but then the truth is when they die, they honestly had very little consumer traction at that point in time. We often see this with deep tech companies, companies that have a high amount of technical risk, especially for founders who are highly charismatic and very comfortable raising capital. What I mean by that is that these companies tend to have the same playbook is that they have a visionary way of changing the world. And so they want to basically, one, build that fundamental technology that changes the world. Two, is to persuade a critical mass of society to adopt that new way of the world. Thirdly, they often have to partner with corporations or partners in order to distribute and make that happen. Then fourthly is they need to get regulatory approval because of the large sense of that vision. Lastly, they raise a lot of capital.

When I say this, then you're kind of like, wait a moment, isn't that successful? Aren't there so many great companies that did that? Didn't we see that happen with Tesla? Again, electric vehicles, you know, convince people to buy electric cars. And then you had to get regulators to set up battery and regulate that and get the distribute and you had to raise a ton of money.

I mean, isn't that the same for SpaceX as well, which are both Elon Musk companies. I think he has the same set of playbook, the same set of understanding. But again, you're building this concept of a reusable rocket which you have to engineer. Then you had to convince the government to buy it. Then you had to convince people to put payloads in it. And then you have to avoid triggering competition from Lockheed Martin and the other competitors and honestly, you have to hope and eventually luck out with the Russians no longer being able to provide services because of the decoupling between the Russian space industry and the US space industry, which at that point of time, 10 years ago was the norm. And then you raise a ton of money again. So SpaceX is a success. Federal Express is another success at that point of time when they were founded. They also had, again, believed that people would pay for fast delivery, raise a ton of capital, break the local monopolies in delivery and package delivery and partner of a lot of folks. Lots of different wins, I think, obviously, come from this playbook. But the truth is there are lots of failures too.

We saw that for Iridium, for satellite communications. So, back then, it was a giant satellite phone. And basically, the concept was that if you are anywhere in the world, you can call and use data anywhere in the world, which honestly looks like Starlink, which is a subset of SpaceX. But it was way ahead of its curve. It raised hundreds of millions of dollars, and nobody bought it because it was a big, chunky thing. It was way ahead of its time, and nobody really needed it. And Iridium was eventually saved or salvaged from bankruptcy because it was bought out by the US government and the military requirements because imagine it's a useful satellite network to have. So those Iridium satellites are still traveling around the world, and now we have Starlink. So, you know, this happens to be like almost one to two decades ahead of the curve.

We saw that for Segway, for electrical mobility, as you can imagine, again, people have Segways, it's the future, everyone's gonna use a Segway, they're not gonna drive, they're not gonna cycle, we're gonna use a Segway. And nobody really uses Segways, except for those buddy movie comedies where cops are using Segways some people in the airports. It's not really used by anybody, but it was revolutionary for its time because it was batteries times personal mobility, times, you know, gyroscope, you know, all this stuff to be sent. And obviously today we're kind of saying like, okay, it's kind of coming back with electric bicycles there. Now we have self-driving cars, which are personal mobility. So there's a lot of stuff that looks like it.

We also have Webvan, it started almost 10 years before Amazon.com, but they wanted to do online groceries, but this was imagined even before Amazon. So it wasn't even doing books. It was doing groceries. People didn't have dial-up. People didn't have GPS. It was just really hard to build all of it. But as a result, Webvan failed, but then Amazon succeeded. And then fast forward another 10 years, RedMart has succeeded in Singapore, even though it takes a lot of capital. It's not the most profitable business. And now people are trying to make that work in Indonesia and Southeast Asia as well. So again, you can see these visions where these are really cascading miracles. You need multiple high-risk factors to multiply to each other and make it something that actually happens.

(36:44) Jeremy Au:

So the reason why I'm talking about all of these six things is that when we are as founders or as teammates or executive people or VCs, obviously you want companies to succeed, so we want to push them for the moonshot on top of the success factors and we want to study from success factors, but the truth is a lot of the success has been handicapped or cannot be achieved because failure is happening because of avoidable mistakes. And avoidable mistakes happen because we don't want to look at failure directly in the eye.

We don't want to spell out exactly what are the problems that we often see. And this is where we kind of enter that feedback bubble that founders can have because founders, often are very much facing the disapproval of family and friends saying like, Hey, why are you doing this crazy thing? And so, you and are getting a lot of criticism that is not well-founded But on the other side, you also have people who are in your camp and supporting you. They're supporting you because they're supporting you as a person. And so the truth is, there are very few people around the founders, where you have advisors who are supportive of you personally. They want the company to succeed, and they are spending the time and have the experience needed to fully understand the business. And if you don't have these things, then it kind of falls apart, right?

I mean, you can't imagine somebody who wants the company to succeed, but don't care about you personally. Well, that's a terrible advisor. There are people who care about you personally, but they don't have experience, and so they give you a bunch of bad advice, but in a genuine and authentic way. These are all terrible ways to fail, and we don't want to fail because of that. So what I'm really saying is that we need to focus on the specific example about why failure is happening. And what are the clusters of causes? And this is where Professor Tom Eisenmannn has done an incredible job listing out these clusters and just saying, Hey, if you avoid these six clusters of problems, at least you know that you're not going to fall into a hole and you can avoid these things and then you can also optimize for success.

So in conclusion, I'm sharing about failure because it's not just an economic thing, right? As a VC, it's like, oh, you know, 19 out of 20 fail, but only one out of 20 succeeded. Sure. I mean, that's a very nice portfolio way to think about it. But as a VC, obviously, economically, you want to maximize your success rate. Sure. So hopefully you're paying attention to this and being thoughtful about this. But I'm really talking about this for the founders and operators who are thinking about this and people who want to be founders, which is being very thoughtful about why executive failure is happening. When you open up the newspaper and you see something is failing, you have to say like, look, okay, this is one point of view. But what is the deeper point of view? What is the insider's point of view? What is the point of view of the executive team with the benefit of hindsight, with the benefit of experience, with the benefit of the knowledge of the counterfactual reality that time has passed? What are those lessons that we have?

And so, what I'm trying to say here is that the examples I gave were often much, five or ten years ago, which, as you think about it, quite like "donkey" years ago. But the truth is, because they had the data and the founders are now comfortable sharing those experiences. But the patterns of that failure still exist today. And I would say that these patterns of failure are more well-known, from my perspective, in more developed ecosystems. So what I mean by that is that when I was in San Francisco or New York, there were often these stories and these are all the ha-ha dinner party. These are the fairy tales or the morality stories that make it easier for me to viscerally feel it and know it and embed it. Even though I didn't know these six exact clusters, but you hear enough of these stories and you're like, okay, I kind of get it. And everybody can subdivide it into two or five or 10 different types of clusters, but you kind of know those mistakes.

But what's interesting is that for people in emerging markets like Southeast Asia or wherever you are, or if you are a first-time founder, or if you just happen to be a first-time founder in a new vertical, then what I'm trying to challenge you to do is, this is an opportunity to learn from the research that's been done by Harvard Business School on the US ecosystem. And these are not patterns of failure of the US side, but these are replicable, teeming, human, leadership, growth rate, organic problems, that would naturally come to existence. And if we can just look to avoid them, if we can look around the corner, then, honestly, we can avoid getting shot in the face.

(40:25) Jeremy Au:

So on that note, I want to share about these six types of startup failure. And I want to say that if you have the opportunity, go and buy this book. It's a hard book to read when you're winning because you've got so many things to read. It's a hard book to read when you're failing because it feels kind of like trying to catch a bullet after it's been fired, but I just recommend that you check out the book and inside the episode, there'll be a link to the book and you should check out the opportunity to buy this book if you can.

On that note, thank you so much and see you around.