" We know about the Boeing safety disasters that all of us have been concerned about. One of the issues identified was that for many years Boeing had a strong culture of safety and reliability. Most of us have grown up flying in Boeing planes, and if you and I go on a plane tomorrow, we would not care whether it is a Boeing plane or an Airbus plane. But at one point, we heard a story about a plane where a door, supposedly part of the fuselage, blew out. A teenage university student almost got sucked out and had his shirt ripped off because the air was rushing out. If he had not been wearing his seatbelt, he would have died after being pulled out of the plane. " - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast


" What’s interesting is that people rushing to get planes out on time at a cheaper budget ended up costing Boeing much more down the road, with recalls, grounded planes, and multiple investigations. A relatively small decision by the frontline manufacturer caused billions of dollars in damage to Boeing as a company because of this defect. The realization is that from a manufacturing perspective, it is important to be lean, to focus on small improvements, to let the frontline drive those improvements, and to allow production to stop when necessary. " - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast


" What’s important is that instead of just building, you build a minimum viable product, the easiest version to test your hypothesis. Then you measure the results—whether people like it, enjoy it, or whether it actually works. You look at the data, learn from it, change from it, get a better idea, and then build again to improve on it. That repeated loop is key, because when you do it faster than your enemy, you defeat your enemy. If another startup takes one month to learn and you take one day, by the end of that month you have learned 30 things more than your enemy. Your rate of learning is your ability to turn this crank over and over again. " - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast

Jeremy Au shared lessons from Toyota’s Kaizen model, Boeing’s safety lapses, and lean startup methods. He explained why small improvements, frontline empowerment, and rapid iteration matter for both manufacturing and startups. The discussion connected MVP thinking with divergence/convergence cycles and how faster learning beats the competition.

00:46 Kaizen as Learning Flywheel: Jeremy introduces Kaizen as a cycle of building, measuring, and improving that mirrors startup learning.

01:35 Toyota’s Frontline Empowerment: He highlights how Toyota empowered workers to suggest improvements and stop production to ensure quality.

03:10 Boeing’s Safety Failures: Jeremy shows how ignoring frontline mistakes caused costly recalls and damaged Boeing’s reputation.

05:00 Rate of Learning as Edge: He explains why startups win by learning faster than rivals, compounding insights into competitive advantage.

07:15 MVP by Stages: Jeremy uses the skateboard-to-car analogy and China’s EV path to show how MVPs accelerate learning at every stage.

Jeremy Au (01:06)

I want to talk to you about something called "Kaizen" and the secret of Toyota. What used to happen was that cars used to be produced in a very haphazard way. So obviously,

people were producing cars that were in different parts of the factory, there were different teams walking in and out, there was very bad version control, and basically what happened at that point of time was that there was a struggle to make cars that were cheap, reliable, and on time. And what happened as a result, obviously, is we know, Henry Ford created the assembly line that basically meant that cars would travel in a row and people would work on it in a sequence. And then,

more recently, we talked about Toyota, which is "Kaizen", and in Chinese, it would be the Japanese version of "kaishain", which means improvement for the better. And what's really important about changing for the better is the philosophy, was that Toyota basically said that our job is not to make this giant improvement. Our job is not necessarily to have a very top-down approach of how managers change the assembly line. We want to give the power to the front line, the workers who are touching

the metal every day, to have the ability to suggest improvements, but also have the ability to stop if there are errors or defects in a production. And this is an important piece for us to think about because what Toyota was able to do was because they say, our job is not to think about large changes or large insights. Our job is to accumulate and accelerate the number of small insights from the front line. And this is really important because we saw that this has allowed Toyota obviously to become

one of the number one car manufacturers in the world. But we also realized that the ability for us to learn and relearn this keeps happening over and over again. For example, we know about the Boeing safety disasters that all of us have been concerned about. And if you understand the news, one of the issues they identified was that for many years, basically Boeing had a strong culture of safety, reliability, and most of us

have grown up our lives flying in Boeing planes. And if you and I go on a plane tomorrow, you and I would not care about whether it's a Boeing plane or Airbus plane. But for time, we always hear about a story about, for example, on a plane, there was a door that was supposedly part of the fuselage, basically blew out, right? And this teenager university student almost got sucked out, right? And he was basically had his shirt ripped off

because the air was rushing out and if he wasn't wearing his seatbelt, he would have died because he had been pulled out of the plane, right? But what was the issue of that? The issue about that was that when people were rushing around in the Boeing plane trying to make things happen, there were mistakes that were being made and people were not checking each other and people saw the mistakes that were being made but did not choose to stop the production of the plane. And so,

what's interesting is that these people rushing to do stuff, rushing to get those planes out on time at a cheaper budget, basically cost Boeing a lot more money down the road because Boeing had to do a recall, you had to ground the planes, were multiple investigations. So a relatively small decision by the manufacturer at the front line caused this billions dollars of damage to Boeing as an overall company because of this defect. And so the realization that was there is that

from a manufacturing perspective, it's important for you to be lean, to be thoughtful about the perspective of being able to have small improvements and letting the front line be able to drive that improvement and having the ability to stop production of the Boeing plane when needs to be, because if that person felt safe and secure to stop the production of it and say, hey, we need an extra day or two to nail and properly nail down this door,

then the door wouldn't have blown out of the plane years later, right? And so that's important for us to think about it is that even though you and I think that Toyota inventing "Kaizen" is old news by now, it turns out that even today our manufacturers like Boeing have to relearn this attitude and approach. And the reason why I'm talking about this manufacturing approach is that it is

relevant for something that we call the lean startup. But what's really important for us to think about is that this is how the "Kaizen" model is very much about learning, improving, measuring. There's a similar approach of a flywheel where we're doing that, but for learning about the company that we're trying to build. So all of us understand the part about the learning loop.

And one them that I like, of course, and we all like is the building part. Because we judge our perspective about how things are done by building. I use somebody who builds or somebody who's just critiquing or observing or financing. So we respect builders. But what's important is that instead of building, you build a minimum viable product, which is what is the easiest version to test your hypothesis. Then you measure the results, whether people like it or not, whether people enjoy it or not,

whether it actually works or not. Then you look at that data, you learn from it, you change from it, you get a better idea, and then you build it again so you improve on it. So that repeated loop is really key because when you do that faster than your enemy, then you will kill your enemy. So if you are fighting for another startup and for you to learn this, it takes them one month.

And for you to learn it takes you one day. Then by the end of that one month, you have learned 30 things more than your enemy. And so that's really important for you to think about because your rate of learning is your ability to turn this crank over and over again. Just like how Toyota is able to turn the crank over and over again. Every time the production is happening, we're getting smarter and smarter. For founders, similarly, your job is to learn the insights that are needed over and over again.

So what do we mean by a minimum viable product? Most of us, when you are in school or in work, most of you will have to build memos or business plans. And so when you think about a minimum viable product, you feel like you have to design the whole car. So imagine you've been asked to say, my job is to build a company that does cars or travels from point A to point B. Some of you will say, month one, I'm going to build one wheel. Month two,

I'm going to build two wheels. Month three, I'll build the body of the car. And by the end of the month four, I'll build one car. But for a minimum viable product, what you try to achieve is that you're trying to build something that gets you the learning. So in month one, you might build a skateboard. In month two, you build kind of a kick scooter, right? In month three, you build an electric bicycle. Four is you build an electric motorbike. And five, you build a

full car, right? And so at each stage of it, you are not waiting for some product that's incomplete or waiting to happen because at each stage you're able to learn from it. And what's interesting is that if you look at it from electric vehicles versus electric turtle combustion cars, a lot of it has followed the same similar dynamic, especially from the Chinese ecosystem, right? Because obviously Tesla went off to build a car. They chose to build a very expensive car.

Then a bigger car, then a SUV, then a truck. Obviously for the Chinese ecosystem they chose it differently right they started building electric kick scooters I don't know if you remember electric small scooters, electric skateboards, electric bicycles, electric motorbikes and as they get better and better at it then they realized they can build eventually they can go all the way to building electric cars right and a lot of these components are similar the batteries are similar the wheels are similar the requirements are similar

So there is a component where a minimum viable product at each stage lets you learn faster at each stage instead of waiting for somebody to complete this deliverable that doesn't give you the customer insights because it's not live yet. And so what's also important as a result is that you as a group and you have already have started debating your startups, started debating what companies are gonna be unicorns or not. What's important is that as a group, whichever team you add is that you must always understand there are two versions

of how you think about life in your group conversations. There must be a time phase for divergence and there must be a time phase for convergence. And it's important for you to think about it. Divergence is what I call brainstorming. So five people at the start of the meeting might be like, everybody is, what do you like? I like this company, I like that company, I think about this, I think about that. So you want to hear everybody's opinions. You want to empathize, you want to understand it. And then you want to converge. You start to prune away

the stuff that's crazy, you prune away the stuff that doesn't really work, maybe you prune away things that are not viable at this point in time. So there's always that cycle where you're like, divergence, then convergence. And obviously some of you know that some groups are irritating, right? Because some people are, you're in the brainstorming phase and this person's being very irritating because he's saying, oh, we need to start pruning down the answer. Then this is very irritating because this person is not in synchronization with the rest of the group. But we also know the opposite of that.

Which is that we are down to the final five minutes, 10 minute decision that we need to make. And some boys still trying to brainstorm when people are trying to align on the final decision. So divergence and convergence is something that your teams will always need to do over time. And those who are good leaders will open up the space for different problems and then converge. Now in the context of the startup perspective, you always need to do it from the perspective of the problem and the solution.

So in this simplified version is you start out first by saying, hey, what problems does this person have? What promises the consumer have? Do I understand who they are? What's the education level? What's the income level? Who are they? Do I understand their problem? And then after that, I start converging. I start to define the problem from their perspective. And then at the end of that process of divergence and convergence, you kind of define the problem statement. And then,

you say, okay, the problem statement is that, for example, the problem is that the world doesn't have enough water. Somebody doesn't have enough water in the wild. Then you start diverging your idea. Do you want to build a desalination plant? Do you want to build a water purification system? Do you want to build an electric water purification system? Do you want to give them chemical water purification tablets? So there are different brainstorming of solutions.

And then after that, your group will converge on a new approach for that. And then you will loop that and iterate on that. So you bring out a solution, you bring out the customer, and you find out whether they like it or not. So that's something for you to do over and over again, which is, are you actually solving the right problem? Are you actually solving the right solution? Diverge, converge, diverge, converge, and get it done.